[go: up one dir, main page]
Next Issue
Volume 13, June-2
Previous Issue
Volume 13, May-2
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
5.3
2.6
Journals
Electronics
Volume 13
Issue 11
share announcement

Electronics, Volume 13, Issue 11 (June-1 2024) – 230 articles

Cover Story (view full-size image): In the last decade, radar sensors have gained a lot of interest as key technological devices for the implementation of next-generation advanced driver-assistance systems (ADASs). This paper presents the latest achievements regarding CMOS-integrated circuits for automotive radar sensors in the 77 GHz frequency band. The work mainly focuses on the most challenging part of implementing radar sensors, which concerns the RF front end; its performance must simultaneously satisfy the stringent requirements of short-, medium-, and long-range radar applications. Innovative circuit topologies in 28 nm FDSOI CMOS technology are discussed, pertaining to mixers, variable-gain amplifiers, and filters in the receiver and voltage-controlled oscillators, frequency doublers and power amplifiers in the transmitter. View this paper
  • Issues are regarded as officially published after their release is announced to the table of contents alert mailing list.
  • You may sign up for e-mail alerts to receive table of contents of newly released issues.
  • PDF is the official format for papers published in both, html and pdf forms. To view the papers in pdf format, click on the "PDF Full-text" link, and use the free Adobe Reader to open them.
Order results
Result details
Section
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
17 pages, 6402 KiB  
Article
Capacitive and Non-Contact Liquid Level Detection Sensor Based on Interdigitated Electrodes with Flexible Substrate
by Yong Ren, Bin Luo, Xueyu Feng, Zihao Feng, Yanyi Song and Fang Yan
Electronics 2024, 13(11), 2228; https://doi.org/10.3390/electronics13112228 - 6 Jun 2024
Cited by 1 | Viewed by 1800
Abstract
Achieving accurate and high-sensitivity liquid level detection in medical instruments has always been a knotty task. In this paper, a high-precision, non-contact, flexible capacitive liquid level sensor is proposed, aiming to apply capacitive sensors in test tube liquid level measurement and improving the [...] Read more.
Achieving accurate and high-sensitivity liquid level detection in medical instruments has always been a knotty task. In this paper, a high-precision, non-contact, flexible capacitive liquid level sensor is proposed, aiming to apply capacitive sensors in test tube liquid level measurement and improving the sensitivity of real-time liquid level sensors. The simulation study is conducted using ANSYS Maxwell and demonstrates the correlation between test tube thickness and sensitivity. A geometric model of the test container and sensing electrodes is established to optimize the design strategy for the physical dimensions of the sensor’s interdigitated (IDT) electrodes based on a flexible printed circuit (FPC). The hardware and software designs are completed based on the FDC2214 capacitive-to-digital converter to collect the capacitance variation data of the sensing electrodes accurately. To assess the system’s performance, an experimental platform for a liquid level sensor system has been constructed, facilitating the measurement, communication, processing, and visualization of liquid levels. The performance results demonstrate that the system is capable of accurately measuring the effective liquid level range within a standard 5 mL test tube with a resolution of up to 1 mm, as well as a sensitivity of 78.68 fF/mm, verifying the simulation results and exhibiting excellent linearity. Full article
(This article belongs to the Section Flexible Electronics)
►▼ Show Figures

Figure 1

Figure 1
<p>Transition of a parallel to a coplanar copper electrode capacitor (the colors in the figure reflect the electric field intensity and follow the distribution of a rainbow): (<b>a</b>) ideal parallel electrode capacitor; (<b>b</b>) e-vector in open transition with copper electrodes opening angle of 30 degrees; (<b>c</b>) e-vector in open transition with copper electrodes opening angle of 90 degrees; (<b>d</b>) Mag-E in coplanar electrode capacitor.</p> Full article ">Figure 2
<p>Diagram of IDT electrode structure.</p> Full article ">Figure 3
<p>Top view and cross-sectional of IDT electrode.</p> Full article ">Figure 4
<p>Model of IDT electrode with two kinds of structure in plane: (<b>a</b>) <math display="inline"><semantics> <mrow> <msub> <mi>x</mi> <mn>0</mn> </msub> <mo>=</mo> <mi>w</mi> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math> mm, <math display="inline"><semantics> <mrow> <msub> <mi>y</mi> <mn>0</mn> </msub> <mo>=</mo> <mi>l</mi> <mo>=</mo> <mn>39</mn> </mrow> </semantics></math> mm; (<b>b</b>) <math display="inline"><semantics> <mrow> <msub> <mi>x</mi> <mn>0</mn> </msub> <mo>=</mo> <mi>w</mi> <mo>=</mo> <mn>1.5</mn> </mrow> </semantics></math> mm, <math display="inline"><semantics> <mrow> <msub> <mi>y</mi> <mn>0</mn> </msub> <mo>=</mo> <mi>l</mi> <mo>=</mo> <mn>39.5</mn> </mrow> </semantics></math> mm; (<b>c</b>) glass tube model with IDT size in (<b>a</b>); (<b>d</b>) glass tube model with IDT size in (<b>b</b>).</p> Full article ">Figure 5
<p>The influence of IDT electrodes’ lateral and longitudinal metalization on capacitance.</p> Full article ">Figure 6
<p>Capacitance values in the form of simulation point data and fitting lines with respect to the liquid level for six kinds of IDT sensors.</p> Full article ">Figure 7
<p>Electrical field distribution of the sensor in the glass tube with water (<b>a</b>) from a cross-sectional (liquid level above 1 mm) perspective; (<b>b</b>) from a cross-sectional (liquid level) perspective; (<b>c</b>) from a cross-sectional (liquid level below 1 mm) perspective; (<b>d</b>) from a frontal perspective; (<b>e</b>) from behind.</p> Full article ">Figure 8
<p>Capacitance values in the form of simulation point data and a fitting line with respect to the liquid level for three kinds of IDT wavelength (1 mm, 2 mm, and 4 mm).</p> Full article ">Figure 9
<p>(<b>a</b>) Simulation capacitance with respect to liquid level for different thicknesses of the test tube; (<b>b</b>) sensitivity.</p> Full article ">Figure 10
<p>Fabricated IDT electrode with flexible substrate.</p> Full article ">Figure 11
<p>Diagram of RLC series circuit.</p> Full article ">Figure 12
<p>PCBA of the main control board in (<b>a</b>) 3D preview and (<b>b</b>) simplified diagram view.</p> Full article ">Figure 13
<p>Experimental setup for non-contact capacitive sensing liquid level detection.</p> Full article ">Figure 14
<p>Comparison of the change in sensor capacitance relative to liquid level variation between simulated and measured values.</p> Full article ">Figure 15
<p>Comparison of measurement and experimental results with maximum error 1.23% full-scale span (FSS).</p> Full article ">Figure 16
<p>The standard deviations of the calibration curve.</p> Full article ">
18 pages, 6958 KiB  
Article
An Improved Current Signal Extraction-Based High-Frequency Pulsating Square-Wave Voltage Injection Method for Interior Permanent-Magnet Synchronous Motor Position-Sensorless Control
by Dongyi Meng, Qiya Wu, Jia Zhang and Lijun Diao
Electronics 2024, 13(11), 2227; https://doi.org/10.3390/electronics13112227 - 6 Jun 2024
Cited by 1 | Viewed by 1312
Abstract
The high-frequency (HF) voltage injection method is widely applied in achieving position-sensorless control for interior permanent-magnet synchronous motors (IPMSMs). This method necessitates precise and rapid extraction of the current signal for accurate position estimation and field-oriented control (FOC). In the traditional methods, the [...] Read more.
The high-frequency (HF) voltage injection method is widely applied in achieving position-sensorless control for interior permanent-magnet synchronous motors (IPMSMs). This method necessitates precise and rapid extraction of the current signal for accurate position estimation and field-oriented control (FOC). In the traditional methods, the position error signal and fundamental current are extracted from the current signal using band-pass filters (BPFs) and low-pass filters (LPFs), or a method based on time-delay filters. However, the traditional extraction method falls short in ensuring simultaneous dynamic performance and accuracy, particularly when the switching frequency is limited or when encountering harmonic and noise interference. In this article, a novel HF pulsating square-wave voltage injection method based on an improved current signal-extraction strategy is proposed to improve the extraction accuracy while maintaining good dynamic performance. The newly devised current signal-extraction method is crafted upon a notch filter (NF). Through harnessing NF’s effective separation characteristics of specific frequency signals, the current signal is meticulously processed. This process yields the extraction of the position error signal and fundamental-current component, crucial for accurate position estimation and motor FOC. Simulation and hardware-in-the-loop (HIL) testing are conducted to validate the effectiveness of the proposed approach. Full article
(This article belongs to the Special Issue Recent Advances in Electrified Vehicles and Transportation Electrification)
►▼ Show Figures

Figure 1

Figure 1
<p>Coordinate reference frames of IPMSM.</p> Full article ">Figure 2
<p>PSWVI method with BPF + LPF.</p> Full article ">Figure 3
<p>PSWVI method with TDFSE strategy.</p> Full article ">Figure 4
<p>Bode plot of NF.</p> Full article ">Figure 5
<p>Block diagram of improved NF-based signal-extraction (NFSE) method.</p> Full article ">Figure 6
<p>Block Diagram of HF PSWVI strategy with improved signal-extraction method.</p> Full article ">Figure 7
<p>Steady−state simulation results at the speed of 20 Hz. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">Figure 8
<p>Steady−state simulation results at the speed of 10 Hz. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">Figure 9
<p>Simulation results of the IPMSM traction process ranging from 2 Hz to 20 Hz. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">Figure 10
<p>Simulation results of the IPMSM braking process ranging from 20 Hz to 2 Hz. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">Figure 11
<p>The artificial noise signals added to the sampled three-phase currents.</p> Full article ">Figure 12
<p>Simulation results at the speed of 20 Hz with artificial noise. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">Figure 13
<p>Simulation results at the speed of 10 Hz with artificial noise. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">Figure 14
<p>RT-LAB experimental platform.</p> Full article ">Figure 15
<p>Steady−state HIL testing results at the speed of 20 Hz. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">Figure 16
<p>Steady−state HIL testing results at the speed of 10 Hz. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">Figure 17
<p>HIL testing results of the IPMSM traction and braking process. (<b>a</b>) With the conventional TDFSE. (<b>b</b>) With the proposed NFSE.</p> Full article ">
33 pages, 5572 KiB  
Article
Performance Evaluation of Mobile RPL-Based IoT Networks under Hello Flood Attack
by Amal Hkiri, Sami Alqurashi, Omar Ben Bahri, Mouna Karmani, Hamzah Faraj and Mohsen Machhout
Electronics 2024, 13(11), 2226; https://doi.org/10.3390/electronics13112226 - 6 Jun 2024
Viewed by 866
Abstract
The RPL protocol is essential for efficient communication within the Internet of Things (IoT) ecosystem, yet it remains vulnerable to various attacks, particularly in dense and mobile environments where it shows certain limitations and susceptibilities. This paper presents a comprehensive simulation-based analysis of [...] Read more.
The RPL protocol is essential for efficient communication within the Internet of Things (IoT) ecosystem, yet it remains vulnerable to various attacks, particularly in dense and mobile environments where it shows certain limitations and susceptibilities. This paper presents a comprehensive simulation-based analysis of the RPL protocol’s vulnerability to the Hello Flood attack in mobile environments. Using four different group mobility models—the Column Mobility Model (CMM), Reference Point Group Mobility Model (RPGM), Nomadic Community Mobility Model (NCM), and Pursue Mobility Model (PMM)—within the Cooja simulator, this study uniquely investigates the Hello Flood attack in mobile settings, an area previously overlooked. Our systematic evaluation focuses on critical performance metrics, including the Packet Delivery Ratio (PDR), End-to-End Delay (E2ED), throughput, Expected Transmission Count (ETX), and Average Power Consumption (APC). The findings reveal several key insights: PDR decreases significantly, indicating increased packet loss or delivery failures; ETX values rise, necessitating more packet retransmissions and routing hops; E2ED increases, introducing delays in routing decisions and data transmission times; throughput declines as the attack disrupts data flow; and APC escalates due to higher energy usage on packet transmissions, especially over extended paths. These results underscore the urgent need for robust security measures to protect RPL-based IoT networks in mobile environments. Furthermore, our work emphasizes the exacerbated impact of the attack in mobile scenarios, highlighting the evolving security requirements of IoT networks. Full article
(This article belongs to the Section Networks)
►▼ Show Figures

Figure 1

Figure 1
<p>The Hello Flood attack flowchart.</p> Full article ">Figure 2
<p>PDR without and with attack using CMM.</p> Full article ">Figure 3
<p>E2ED without and with attack using CMM.</p> Full article ">Figure 4
<p>Throughput without and with attack using CMM.</p> Full article ">Figure 5
<p>ETX without and with attack using CMM.</p> Full article ">Figure 6
<p>APC without and with attack using CMM.</p> Full article ">Figure 7
<p>PDR without and with attack using RPGM mobility model.</p> Full article ">Figure 8
<p>E2ED without and with attack using RPGM mobility model.</p> Full article ">Figure 9
<p>Throughput without and with attack using RPGM mobility model.</p> Full article ">Figure 10
<p>ETX without and with attack using RPGM mobility model.</p> Full article ">Figure 11
<p>APC without and with attack using RPGM mobility model.</p> Full article ">Figure 12
<p>PDR without and with attack using NCM.</p> Full article ">Figure 13
<p>E2ED without and with attack using NCM.</p> Full article ">Figure 14
<p>Throughput without and with attack using NCM.</p> Full article ">Figure 15
<p>ETX without and with attack using NCM.</p> Full article ">Figure 16
<p>APC without and with attack using NCM.</p> Full article ">Figure 17
<p>PDR without and with attack using PMM.</p> Full article ">Figure 18
<p>E2ED without and with attack using PMM.</p> Full article ">Figure 19
<p>Throughput without and with attack using PMM.</p> Full article ">Figure 20
<p>ETX without and with attack using PMM.</p> Full article ">Figure 21
<p>APC without and with attack using PMM.</p> Full article ">
39 pages, 24746 KiB  
Review
Review of DC Motor Modeling and Linear Control: Theory with Laboratory Tests
by Miklós Kuczmann
Electronics 2024, 13(11), 2225; https://doi.org/10.3390/electronics13112225 - 6 Jun 2024
Viewed by 1197
Abstract
This review paper introduces the modeling, measurement, identification and control of direct current motors based on the state space modeling and the transfer function representation. These models are identified by real laboratory measurements, and the simulated results are compared with the measurements. Continuous-time [...] Read more.
This review paper introduces the modeling, measurement, identification and control of direct current motors based on the state space modeling and the transfer function representation. These models are identified by real laboratory measurements, and the simulated results are compared with the measurements. Continuous-time and discrete-time PID (Proportional-Integral-Derivative) controllers, discrete-time state feedback and linear quadratic controllers are designed mathematically. The designed controllers are realized by the microcontroller Arduino UNO, and the behavior of the controllers is compared and analyzed. The noisy current signal has been measured by a discrete-time observer, steady-state Kalman filtering is also studied. The practical results of the implemented controllers support the theoretical results very well. Full article
(This article belongs to the Special Issue Advances in Power Converter Design, Control and Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>The Lorentz force acting on a current carrying frame.</p> Full article ">Figure 2
<p>Equipment for testing and controlling the DC motor.</p> Full article ">Figure 3
<p>Block diagram of the open loop measurement system.</p> Full article ">Figure 4
<p>Block diagram of the closed loop measurement system.</p> Full article ">Figure 5
<p>Equivalent circuit model of the DC motor.</p> Full article ">Figure 6
<p>Block diagram model of the DC motor.</p> Full article ">Figure 7
<p>Block diagram of the closed loop system with PID type controllers.</p> Full article ">Figure 8
<p>The discrete-time observer/estimator.</p> Full article ">Figure 9
<p>DC motor control by state feedback controller extended by integrator and state observer/estimator.</p> Full article ">Figure 10
<p>The characteristic <math display="inline"><semantics> <mrow> <mi>n</mi> <mrow> <mo>[</mo> <mi>RPM</mi> <mo>]</mo> </mrow> <mo>−</mo> <msub> <mi>u</mi> <mi mathvariant="normal">a</mi> </msub> <mrow> <mo>[</mo> <mi>PWM</mi> <mo>]</mo> </mrow> </mrow> </semantics></math> is nonlinear.</p> Full article ">Figure 11
<p>The characteristic <math display="inline"><semantics> <mrow> <mi>n</mi> <mrow> <mo>[</mo> <mi>RPM</mi> <mo>]</mo> </mrow> <mo>−</mo> <msub> <mi>u</mi> <mi mathvariant="normal">a</mi> </msub> <mrow> <mo>[</mo> <mi mathvariant="normal">V</mi> <mo>]</mo> </mrow> </mrow> </semantics></math> is linear.</p> Full article ">Figure 12
<p>The characteristic <math display="inline"><semantics> <mrow> <msub> <mi>i</mi> <mi mathvariant="normal">a</mi> </msub> <mrow> <mo>[</mo> <mi mathvariant="normal">A</mi> <mo>]</mo> </mrow> <mo>−</mo> <msub> <mi>u</mi> <mi mathvariant="normal">a</mi> </msub> <mrow> <mo>[</mo> <mi mathvariant="normal">V</mi> <mo>]</mo> </mrow> </mrow> </semantics></math> is approximately linear.</p> Full article ">Figure 13
<p>Electric and mechanical parts of the block diagram.</p> Full article ">Figure 14
<p>Switch-on transient of motor speed and armature current, comparison between measured and simulated data.</p> Full article ">Figure 15
<p>The realized closed loop system with PID controller.</p> Full article ">Figure 16
<p>Design of P controller for phase margin <math display="inline"><semantics> <mrow> <msub> <mi>φ</mi> <mi mathvariant="normal">m</mi> </msub> <mo>=</mo> <msup> <mn>60</mn> <mo>∘</mo> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 17
<p>Comparison of closed loop step response data for the phase margin <math display="inline"><semantics> <mrow> <msub> <mi>φ</mi> <mi mathvariant="normal">m</mi> </msub> <mo>=</mo> <msup> <mn>60</mn> <mo>∘</mo> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 18
<p>The reference signal tracking of the P controller with phase margins <math display="inline"><semantics> <mrow> <msub> <mi>φ</mi> <mi mathvariant="normal">m</mi> </msub> <mo>=</mo> <msup> <mn>60</mn> <mo>∘</mo> </msup> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mi>φ</mi> <mi mathvariant="normal">m</mi> </msub> <mo>=</mo> <msup> <mn>40</mn> <mo>∘</mo> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 19
<p>The reference signal tracking of the PI controller with phase margin <math display="inline"><semantics> <mrow> <msub> <mi>φ</mi> <mi mathvariant="normal">m</mi> </msub> <mo>=</mo> <msup> <mn>60</mn> <mo>∘</mo> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 20
<p>The reference signal tracking of the PD controller with phase margin <math display="inline"><semantics> <mrow> <msub> <mi>φ</mi> <mi mathvariant="normal">m</mi> </msub> <mo>=</mo> <msup> <mn>60</mn> <mo>∘</mo> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 21
<p>Comparison of closed-loop step response data for the phase margin <math display="inline"><semantics> <mrow> <msub> <mi>φ</mi> <mi mathvariant="normal">m</mi> </msub> <mo>=</mo> <msup> <mn>60</mn> <mo>∘</mo> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 22
<p>Comparison of measured and observed motor current and speed.</p> Full article ">Figure 23
<p>Measured data to set up the Kalman filter parameters.</p> Full article ">Figure 24
<p>Closed loop system behavior designed by pole placement, case 1.</p> Full article ">Figure 25
<p>Closed loop system behavior designed by pole placement, case 2.</p> Full article ">
15 pages, 1798 KiB  
Article
A Novel Paradigm for Controlling Navigation and Walking in Biped Robotics
by Giuseppe Menga
Electronics 2024, 13(11), 2224; https://doi.org/10.3390/electronics13112224 - 6 Jun 2024
Viewed by 631
Abstract
This paper extends the three-dimensional inverted pendulum (spherical inverted pendulum or SIP) in a polar coordinate system to simulate human walking in free fall and the energy recovery when the foot collides with the ground. The purpose is to propose a general model [...] Read more.
This paper extends the three-dimensional inverted pendulum (spherical inverted pendulum or SIP) in a polar coordinate system to simulate human walking in free fall and the energy recovery when the foot collides with the ground. The purpose is to propose a general model to account for all characteristics of the biped and of the gait, while adding minimal dynamical complexity with respect to the SIP. This model allows for both walking omnidirectionally on a flat surface and going up and down staircases. The technique does not use torque control. However, for the gait, the only action is the change in angular velocity at the start of a new step with respect to those given after the collision (emulating the torque action in the brief double stance period) to recover from the losses, as well as the preparation of the position in the frontal and sagittal planes of the swing foot for the next collision for balance and maneuvering. Moreover, in climbing or descending staircases, during the step, the length of the supporting leg is modified for the height of the step of the staircase. Simulation examples are offered for a rectilinear walk, ascending and descending rectilinear or spiral staircases, showing stability of the walk, and the expenditure of energy. Full article
(This article belongs to the Special Issue Advances in Mobile Robots: Navigation, Motion Planning and Control)
►▼ Show Figures

Figure 1

Figure 1
<p>The spherical inverted pendulum. (<b>a</b>) The kinematics of the model. (<b>b</b>) The control of the length <math display="inline"><semantics> <msub> <mi>L</mi> <mn>1</mn> </msub> </semantics></math> by the force <math display="inline"><semantics> <msub> <mi>F</mi> <mn>1</mn> </msub> </semantics></math>.</p> Full article ">Figure 2
<p>The controllers of the six objectives of the gait.</p> Full article ">Figure 3
<p>The COG behavior. (<b>a</b>) COG along the <span class="html-italic">y</span> and <span class="html-italic">z</span> axes. (<b>b</b>) Details of COG and pivot foot position.</p> Full article ">Figure 4
<p>Angle position and velocity behaviors. (<b>a</b>) The angles <math display="inline"><semantics> <msub> <mi>θ</mi> <mi>z</mi> </msub> </semantics></math>, <math display="inline"><semantics> <mi>θ</mi> </semantics></math>, and <math display="inline"><semantics> <msub> <mi>θ</mi> <mi>x</mi> </msub> </semantics></math>. (<b>b</b>) The velocities <math display="inline"><semantics> <msub> <mover accent="true"> <mi>θ</mi> <mo>˙</mo> </mover> <mi>z</mi> </msub> </semantics></math>, <math display="inline"><semantics> <mover accent="true"> <mi>θ</mi> <mo>˙</mo> </mover> </semantics></math>, and <math display="inline"><semantics> <msub> <mover accent="true"> <mi>θ</mi> <mo>˙</mo> </mover> <mi>x</mi> </msub> </semantics></math>.</p> Full article ">Figure 5
<p>Kinetic energy on a straight path on a horizontal surface. (<b>a</b>) Kinetic energy full view. (<b>b</b>) Details.</p> Full article ">Figure 6
<p>Total energy and angular momentum about the vertical axis. (<b>a</b>) Total energy. (<b>b</b>) Angular momentum about the <span class="html-italic">z</span> axis.</p> Full article ">Figure 7
<p>Length, position, and velocity of the supporting leg on a staircase with steps of 15 cm. (<b>a</b>) Up the stair. (<b>b</b>) Down the stair.</p> Full article ">Figure 8
<p>Different staircases. (<b>a</b>) COG on a spiral staircase. (<b>b</b>) COG going up and down a straight staircase.</p> Full article ">Figure 9
<p>Going up and down a rectilinear staircase. (<b>a</b>) Kinetic energy—total view. (<b>b</b>) Going down the staircase. (<b>c</b>) Going up the staircase.</p> Full article ">Figure 10
<p>Kinetic energy in ascending a spiral staircase.</p> Full article ">Figure 11
<p>Details of the total energy in ascending a spiral staircase.</p> Full article ">Figure 12
<p>FPE upon arrival. (<b>a</b>) Angular position. (<b>b</b>) Angular velocities. (<b>c</b>) Total and potential energy. (<b>d</b>) Kinetic energy and momentum.</p> Full article ">
20 pages, 16671 KiB  
Article
A Light-Field Video Dataset of Scenes with Moving Objects Captured with a Plenoptic Video Camera
by Kamran Javidi and Maria G. Martini
Electronics 2024, 13(11), 2223; https://doi.org/10.3390/electronics13112223 - 6 Jun 2024
Viewed by 776
Abstract
Light-field video provides a detailed representation of scenes captured from different perspectives. This results in a visualisation modality that enhances the immersion and engagement of the viewers with the depicted environment. In order to perform research on compression, transmission and signal processing of [...] Read more.
Light-field video provides a detailed representation of scenes captured from different perspectives. This results in a visualisation modality that enhances the immersion and engagement of the viewers with the depicted environment. In order to perform research on compression, transmission and signal processing of light field data, datasets with light-field contents of different categories and acquired with different modalities are required. In particular, the development of machine learning models for quality assessment and for light-field processing, including the generation of new views, require large amounts of data. Most existing datasets consist of static scenes and, in many cases, synthetic contents. This paper presents a novel light-field plenoptic video dataset, KULFR8, involving six real-world scenes with moving objects and 336 distorted light-field videos derived from the original contents; in total, the original scenes in the dataset contain 1800 distinctive frames, with angular resolution of 5×5 with and total spatial resolution of 9600×5400 pixels (considering all the views); overall, the dataset consists of 45,000 different views with spatial resolution of 1920×1080 pixels. We analyse the content characteristics based on the dimensions of the captured objects and via the acquired videos using the central views extracted from each quilted frame. Additionally, we encode and decode the contents using various video encoders across different bitrate ranges. For quality assessments, we consider all the views, utilising frames measuring 9600×5400 pixels, and employ two objective quality metrics: PSNR and SSIM. Full article
(This article belongs to the Special Issue Advances in Human-Centered Digital Systems and Services)
►▼ Show Figures

Figure 1

Figure 1
<p>Raytrix R8 view samples. (<b>a</b>–<b>f</b>) subfigures illustrate the view samples of Bee, Crab, Dinosaur, Magician, Mouse, and Water contents, respectively.</p> Full article ">Figure 2
<p>Raytrix R8 central view depth samples with the scale bar in (cm). (<b>a</b>–<b>f</b>) subfigures illustrate the depth maps of Bee, Crab, Dinosaur, Magician, Mouse, and Water contents, respectively.</p> Full article ">Figure 3
<p>The geometry details of each of the used models are as follows: (<b>a</b>) Bee, (<b>b</b>) Crab, (<b>c</b>) Dinosaur, (<b>d</b>) Magician, (<b>e</b>) Mouse, and two models are used for the Water content, which are (<b>f</b>) Fish and (<b>g</b>) Turtle.</p> Full article ">Figure 4
<p>Content characterisation evaluation flows for (<b>a</b>) motion displacement characterisation flow, and (<b>b</b>) <math display="inline"><semantics> <mrow> <mi>S</mi> <msub> <mi>I</mi> <mrow> <mi>c</mi> <mi>v</mi> </mrow> </msub> </mrow> </semantics></math>, <math display="inline"><semantics> <mrow> <mi>T</mi> <msub> <mi>I</mi> <mrow> <mi>c</mi> <mi>v</mi> </mrow> </msub> </mrow> </semantics></math>, and <math display="inline"><semantics> <mrow> <mi>C</mi> <msub> <mi>F</mi> <mrow> <mi>c</mi> <mi>v</mi> </mrow> </msub> </mrow> </semantics></math> characterisation flow.</p> Full article ">Figure 4 Cont.
<p>Content characterisation evaluation flows for (<b>a</b>) motion displacement characterisation flow, and (<b>b</b>) <math display="inline"><semantics> <mrow> <mi>S</mi> <msub> <mi>I</mi> <mrow> <mi>c</mi> <mi>v</mi> </mrow> </msub> </mrow> </semantics></math>, <math display="inline"><semantics> <mrow> <mi>T</mi> <msub> <mi>I</mi> <mrow> <mi>c</mi> <mi>v</mi> </mrow> </msub> </mrow> </semantics></math>, and <math display="inline"><semantics> <mrow> <mi>C</mi> <msub> <mi>F</mi> <mrow> <mi>c</mi> <mi>v</mi> </mrow> </msub> </mrow> </semantics></math> characterisation flow.</p> Full article ">Figure 5
<p>Each row of images represents one content in the different color spaces HSL, HSV, LAB, and YUV (in order from left to right for the first four sample images) followed by the (L) component for HSL, (V) component for HSV, (L) component for LAB, and (Y) component for YUV (sorted in each row from five to eight). The view samples reported in the six rows are from the contents ‘Bee’, ‘Crab’, ‘Dinosaur’, ‘Magician’, ‘Mouse’, and ‘Water’.</p> Full article ">Figure 6
<p>Detected motion vectors; (<b>a</b>) ‘L’ component in HSL, (<b>b</b>) ‘V’ component in HSV, (<b>c</b>) ‘L’ component in LAB, and (<b>d</b>) ‘Y’ component in YUV.</p> Full article ">Figure 7
<p>Overall motion vectors added on a sample view of ‘V’ component in HSV color space. (<b>a</b>–<b>f</b>) subfigures illustrate the motion vectors on Bee, Crab, Dinosaur, Magician, Mouse, and water contents, respectively.</p> Full article ">Figure 8
<p>Content characterisation values in (<b>a</b>,<b>b</b>) present motion displacement values for HSV colour space in the vertical axis versus two of SI and TI, respectively, in the horizontal axis; (<b>c</b>,<b>d</b>) present colourfulness values in the vertical axis and SI and TI, respectively, in the horizontal axis. (<b>e</b>–<b>g</b>) represent SI versus TI values for three colour spaces of HSL, HSV, and YUV, respectively.</p> Full article ">Figure 9
<p>Quality assessment encompasses two main components: (<b>a</b>) the encode–decode procedure applied to the light-field video contents, and (<b>b</b>) the evaluation of quality metrics such as PSNR and SSIM.</p> Full article ">Figure 10
<p>Quality metric plots <math display="inline"><semantics> <mrow> <mi>P</mi> <mi>S</mi> <mi>N</mi> <msub> <mi>R</mi> <mrow> <mi>Y</mi> <mi>U</mi> <mi>V</mi> </mrow> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>S</mi> <mi>S</mi> <mi>I</mi> <msub> <mi>M</mi> <mrow> <mi>Y</mi> <mi>U</mi> <mi>V</mi> </mrow> </msub> </mrow> </semantics></math> for ‘Bee’, ‘Crab’, ‘Dinosaur’, ‘Magician’, ‘Mouse’, and ‘Water’ light-field video contents.</p> Full article ">
16 pages, 6121 KiB  
Article
Prediction of Machine-Generated Financial Tweets Using Advanced Bidirectional Encoder Representations from Transformers
by Muhammad Asad Arshed, Ștefan Cristian Gherghina, Dur-E-Zahra and Mahnoor Manzoor
Electronics 2024, 13(11), 2222; https://doi.org/10.3390/electronics13112222 - 6 Jun 2024
Viewed by 883
Abstract
With the rise of Large Language Models (LLMs), distinguishing between genuine and AI-generated content, particularly in finance, has become challenging. Previous studies have focused on binary identification of ChatGPT-generated content, overlooking other AI tools used for text regeneration. This study addresses this gap [...] Read more.
With the rise of Large Language Models (LLMs), distinguishing between genuine and AI-generated content, particularly in finance, has become challenging. Previous studies have focused on binary identification of ChatGPT-generated content, overlooking other AI tools used for text regeneration. This study addresses this gap by examining various AI-regenerated content types in the finance domain. Objective: The study aims to differentiate between human-generated financial content and AI-regenerated content, specifically focusing on ChatGPT, QuillBot, and SpinBot. It constructs a dataset comprising real text and AI-regenerated text for this purpose. Contribution: This research contributes to the field by providing a dataset that includes various types of AI-regenerated financial content. It also evaluates the performance of different models, particularly highlighting the effectiveness of the Bidirectional Encoder Representations from the Transformers Base Cased model in distinguishing between these content types. Methods: The dataset is meticulously preprocessed to ensure quality and reliability. Various models, including Bidirectional Encoder Representations Base Cased, are fine-tuned and compared with traditional machine learning models using TFIDF and Word2Vec approaches. Results: The Bidirectional Encoder Representations Base Cased model outperforms other models, achieving an accuracy, precision, recall, and F1 score of 0.73, 0.73, 0.73, and 0.72 respectively, in distinguishing between real and AI-regenerated financial content. Conclusions: This study demonstrates the effectiveness of the Bidirectional Encoder Representations base model in differentiating between human-generated financial content and AI-regenerated content. It highlights the importance of considering various AI tools in identifying synthetic content, particularly in the finance domain in Pakistan. Full article
(This article belongs to the Special Issue Advanced Machine Learning, Pattern Recognition, and Deep Learning Technologies: Methodologies and Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Abstract diagram for proposed study.</p> Full article ">Figure 2
<p>WordCloud of prepared and processed dataset.</p> Full article ">Figure 3
<p>Prepared dataset text length.</p> Full article ">Figure 4
<p>Abstract diagram of model architecture.</p> Full article ">Figure 5
<p>BERT base model, BERT large model, and TFIDF and Word2Vec approach-based best ML models’ scores.</p> Full article ">
18 pages, 10177 KiB  
Article
Distributed High-Density Anchor (Cable) Support Force Monitoring System Research
by Lei Wang, Kai Sun, Junyan Qi and Ruifu Yuan
Electronics 2024, 13(11), 2221; https://doi.org/10.3390/electronics13112221 - 6 Jun 2024
Cited by 1 | Viewed by 614
Abstract
In shaft mining, monitoring the deformation of the roadway due to mining pressure is of great significance to the safe production of coal mines. For this reason, a distributed high-density anchor (cable) support force monitoring system was designed by developing a low-cost anchor [...] Read more.
In shaft mining, monitoring the deformation of the roadway due to mining pressure is of great significance to the safe production of coal mines. For this reason, a distributed high-density anchor (cable) support force monitoring system was designed by developing a low-cost anchor (cable) stress monitoring device, which consists of an anchor (cable) stress sensor and a data acquisition device. The whole system consists of an anchor bar (cable) stress monitoring device and a mine roadway deformation monitoring substation. The signals collected by the anchor force sensors are processed by the data acquisition device and sent to the self-developed mine roadway deformation monitoring substation through Long Range Radio (LoRa) wireless communication. All data from the monitoring substation are transmitted to the ground control center in real time via the Message Queuing Telemetry Transport (MQTT) network transmission protocol. The distributed high-density arrangement of monitoring nodes reflects the deformation trend of the whole section of the roadway by monitoring the anchor bar (cable) support force data of multiple sections, which effectively ensures the safety of the roadway. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Circuit structure of data acquisition device.</p> Full article ">Figure 2
<p>Distributed connection relationship of anchor rod (cable) stress monitoring device.</p> Full article ">Figure 3
<p>Installation diagram of anchor (cable) stress monitoring device.</p> Full article ">Figure 4
<p>Schematic diagram of strain gauges for anchor rod (cable) stress sensors.</p> Full article ">Figure 5
<p>Structure of anchor rod (cable) stress sensor.</p> Full article ">Figure 6
<p>Sensor on-site testing diagram.</p> Full article ">Figure 7
<p>Test values of anchor rod (cable) stress sensor.</p> Full article ">Figure 8
<p>Schematic diagram of signal amplification and filtering circuit for data acquisition device.</p> Full article ">Figure 9
<p>Data after amplification circuit processing.</p> Full article ">Figure 10
<p>Comparison chart before and after Kalman filtering.</p> Full article ">Figure 11
<p>The conversion relationship between pressure and voltage.</p> Full article ">Figure 12
<p>Mine roadway deformation monitoring substation structure.</p> Full article ">Figure 13
<p>Monitoring substation workflow diagram.</p> Full article ">Figure 14
<p>Schematic diagram for the installation of the cross-section of the return airway.</p> Full article ">Figure 15
<p>Section monitoring interval diagram.</p> Full article ">Figure 16
<p>Stress monitoring of anchor rods (cables) 600 m away from the working face.</p> Full article ">Figure 17
<p>Stress monitoring of anchor rods (cables) 300 m away from the working face.</p> Full article ">Figure 18
<p>Stress variation in anchor cables at the top and bottom of small coal pillars.</p> Full article ">Figure 19
<p>Stress variation in the middle anchor rod on the side of the small coal pillar.</p> Full article ">
22 pages, 6722 KiB  
Article
A Cloud-Based WEB Platform for Fall Risk Assessment Using a Therapist-Centered User Interface Which Enables Patients’ Tracking Remotely
by Jesús Damián Blasco-García, Nieves Pavón-Pulido, Juan Antonio López-Riquelme, Ana María Roldán-Ruiz and Jorge Juan Feliu-Batlle
Electronics 2024, 13(11), 2220; https://doi.org/10.3390/electronics13112220 - 6 Jun 2024
Viewed by 628
Abstract
This work describes a system to help in the remote assessment of fall risk in elderly people. A portable hardware system equipped with an RGB-D sensor is used for motion capture. A set of anonymous frames, representing the process of skeleton tracking, and [...] Read more.
This work describes a system to help in the remote assessment of fall risk in elderly people. A portable hardware system equipped with an RGB-D sensor is used for motion capture. A set of anonymous frames, representing the process of skeleton tracking, and a collection of sequences of interesting features, obtained from body landmark evaluations through time, are stored in the Cloud for each patient. A WEB dashboard allows for tailored tests to be designed, which include the typical items within well-known fall risk evaluation tests in the literature. Such a dashboard helps therapists to evaluate each item from the analysis and observation of the sequences and the 3D representation of the body through time, and to compare the results of tests carried out in different moments, checking on the evolution of the fall risk. The software architecture that implements the system allows the information to be stored in a safe manner and preserves patients’ privacy. The paper shows the obtained results after testing an early prototype of the system, a discussion about its advantages, and the current limitations from the Human–Computer Interaction point of view, and a plan to deploy and evaluate the system from the usability perspective in the near future. Full article
(This article belongs to the Special Issue Human-Computer Interactions in E-health)
►▼ Show Figures

Figure 1

Figure 1
<p>Mapping from MediaPipe Pose landmarks to the items of a specific test from the standard of care. In the example, the Tinetti test consists of 16 items, from 1 to 9 for measuring balance and from 10 to 16 for measuring gait. Therapists select those parameters, which are automatically calculated by the system and stored as sequences through time, useful for balance and gait assessment, according to the items to be observed.</p> Full article ">Figure 2
<p>Application of the UCD lifecycle for designing and developing the system for automatic fall risk assessment.</p> Full article ">Figure 3
<p>Back and front views of the portable device for helping caregivers/relatives to apply the tests. Such a device could be easily installed on top of a table or on any similar piece of furniture.</p> Full article ">Figure 4
<p>Distributed software architecture that allows the implementation and application of the “STEADI algorithm for Fall Risk Screening, Assessment and Intervention” remotely. Caregivers and relatives access the software architecture through the portable device, and therapists could use the system on their own personal computers.</p> Full article ">Figure 5
<p>Simplified E–R model for storing all the relevant information related to the STEADI algorithm, including all the needed entities which are transformed into Datastore’s entities for storing timestamped sequences of parameters and landmarks calculated from the MediaPipe Pose solution. Only very relevant attributes are shown for simplicity. Ts is timestamp and N is package’s number.</p> Full article ">Figure 6
<p>Example that illustrates how the E–R entities are implemented as Datastore entities and attributes for storing all the parameters and landmarks associated to different items of a test executed by a patient.</p> Full article ">Figure 7
<p>UI mockup for the application executed in the portable device. (<b>a</b>) Login screenshot; (<b>b</b>) List of patients (elderly people) attached to a caregiver; (<b>c</b>) List of tests prescribed to a patient; (<b>d</b>) List of items included in a selected test; (<b>e</b>) Screenshot of a selected item ready to be executed; (<b>f</b>) Results of the executed item.</p> Full article ">Figure 8
<p>UI mockup for the WEB application used by therapists. (<b>a</b>) List of patients (elderly people) attached to a therapist; (<b>b</b>) Patient’s history; (<b>c</b>) Layout for designing tests and items; (<b>d</b>) Screenshot of the results’ analyzer.</p> Full article ">Figure 9
<p>Screenshot of the “Design Activity” layout in the WEB application, which allows therapists to see the list of designed items.</p> Full article ">Figure 10
<p>Screenshot of the “Test results” layout in the WEB application for a specific patient.</p> Full article ">Figure 11
<p>Screenshot of the “Test Comparison” layout in the WEB application for a specific patient.</p> Full article ">Figure 12
<p>Screenshot of the layout for the “STEADI-Rx”, which implements the checklist for finding correlation between certain medication classes with fall risk.</p> Full article ">Figure 13
<p>Results of the activities carried out by real elderly users in a nursing home. (<b>a</b>) Trunk swing measured for Patient 1 in a “standing balance” activity; (<b>b</b>) Trunk swing measured for Patient 3 in a “standing balance” activity; (<b>c</b>) Separation of the heels measured for Patient 2 in a “standing balance” activity; (<b>d</b>) Separation of the heels measured for Patient 4 in a “standing balance” activity; (<b>e</b>) Results for Patient 2 while walking straight for 3 m; (<b>f</b>) Results for Patient 4 while walking straight for 3 m.</p> Full article ">Figure 13 Cont.
<p>Results of the activities carried out by real elderly users in a nursing home. (<b>a</b>) Trunk swing measured for Patient 1 in a “standing balance” activity; (<b>b</b>) Trunk swing measured for Patient 3 in a “standing balance” activity; (<b>c</b>) Separation of the heels measured for Patient 2 in a “standing balance” activity; (<b>d</b>) Separation of the heels measured for Patient 4 in a “standing balance” activity; (<b>e</b>) Results for Patient 2 while walking straight for 3 m; (<b>f</b>) Results for Patient 4 while walking straight for 3 m.</p> Full article ">
12 pages, 5521 KiB  
Article
Improving the Efficiency of the Axial Flux Machine with Hybrid Excitation
by Pawel Prajzendanc, Ryszard Palka, Piotr Paplicki, Marcin Wardach, Michal Cichowicz, Kamil Cierzniewski, Lech Dorobczynski and Edison Gundabattini
Electronics 2024, 13(11), 2219; https://doi.org/10.3390/electronics13112219 - 6 Jun 2024
Viewed by 593
Abstract
This paper discusses the construction and operating principle of an axial flux electric machine with hybrid excitation. Based on computer simulations using the Finite Element Method, an analysis was conducted with changes in the geometry of the magnetic circuit, which involves the rotation [...] Read more.
This paper discusses the construction and operating principle of an axial flux electric machine with hybrid excitation. Based on computer simulations using the Finite Element Method, an analysis was conducted with changes in the geometry of the magnetic circuit, which involves the rotation of the rotor disks relative to each other on the operating parameters of the machine. Both the generator state of operation, in the meaning of analyzing the induced voltage (adjustment at −11% ÷ +64%) and the cogging torque, and the motor state of operation, in the meaning of analyzing the ripple of the electromagnetic torque (possible reduction by almost 30%), were examined. The article concludes with observations on how the change in the angle of the rotor disks affects the efficiency of the disk machine with axial flux and hybrid excitation. Full article
(This article belongs to the Special Issue The Latest Progress in Computational Electromagnetics and Beyond)
►▼ Show Figures

Figure 1

Figure 1
<p>Basic designs of electromechanical energy converters: (<b>a</b>) radial flux, (<b>b</b>) axial flux.</p> Full article ">Figure 2
<p>Mechanical characteristics of the machine with hybrid excitation.</p> Full article ">Figure 3
<p>Design of machine with axial flux and hybrid excitation.</p> Full article ">Figure 4
<p>Distribution of magnetic fluxes in the machine for different currents in an additional electromagnetic coil: (<b>a</b>) <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi mathvariant="normal">D</mi> <mi mathvariant="normal">C</mi> </mrow> </msub> <mo>=</mo> <mn>0</mn> <mo> </mo> <mi mathvariant="normal">A</mi> </mrow> </semantics></math>, (<b>b</b>) <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi mathvariant="normal">D</mi> <mi mathvariant="normal">C</mi> </mrow> </msub> <mo>=</mo> <mo>−</mo> <mn>5</mn> <mo> </mo> <mi mathvariant="normal">A</mi> </mrow> </semantics></math>, (<b>c</b>) <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi mathvariant="normal">D</mi> <mi mathvariant="normal">C</mi> </mrow> </msub> <mo>=</mo> <mn>5</mn> <mo> </mo> <mi mathvariant="normal">A</mi> </mrow> </semantics></math>.</p> Full article ">Figure 5
<p>Changing the angle between the rotor discs.</p> Full article ">Figure 6
<p>Distribution of magnetic field induction within the machine.</p> Full article ">Figure 7
<p>The computational mesh generated by the FEM method.</p> Full article ">Figure 8
<p>The relationship between the induced voltage and the angle <math display="inline"><semantics> <mrow> <mi>α</mi> </mrow> </semantics></math>.</p> Full article ">Figure 9
<p>Stator connection diagrams: (<b>a</b>) basic series connection; (<b>b</b>) inverted stator connection.</p> Full article ">Figure 10
<p>The relationship between the induced voltage and the angle <math display="inline"><semantics> <mrow> <mi>α</mi> </mrow> </semantics></math> for basic and inverted stator connections.</p> Full article ">Figure 11
<p>The relationship between the maximum value of the cogging torque and the angle <math display="inline"><semantics> <mrow> <mi>α</mi> </mrow> </semantics></math>.</p> Full article ">Figure 12
<p>Electromagnetic torque for two angles: (<b>a</b>) <math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>0</mn> <mo>°</mo> </mrow> </semantics></math>, (<b>b</b>) <math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>4</mn> <mo>°</mo> </mrow> </semantics></math>.</p> Full article ">
22 pages, 6864 KiB  
Article
Position Estimation Method for Small Drones Based on the Fusion of Multisource, Multimodal Data and Digital Twins
by Shaochun Qu, Jian Cui, Zijian Cao, Yongxing Qiao, Xuemeng Men and Yanfang Fu
Electronics 2024, 13(11), 2218; https://doi.org/10.3390/electronics13112218 - 6 Jun 2024
Cited by 1 | Viewed by 1076
Abstract
In response to the issue of low positioning accuracy and insufficient robustness in small UAVs (unmanned aerial vehicle) caused by sensor noise and cumulative motion errors during flight in complex environments, this paper proposes a multisource, multimodal data fusion method. Initially, it employs [...] Read more.
In response to the issue of low positioning accuracy and insufficient robustness in small UAVs (unmanned aerial vehicle) caused by sensor noise and cumulative motion errors during flight in complex environments, this paper proposes a multisource, multimodal data fusion method. Initially, it employs a multimodal data fusion of various sensors, including GPS (global positioning system), an IMU (inertial measurement unit), and visual sensors, to complement the strengths and weaknesses of each hardware component, thereby mitigating motion errors to enhance accuracy. To mitigate the impact of sudden changes in sensor data, a high-fidelity UAV model is established in the digital twin based on the real UAV parameters, providing a robust reference for data fusion. By utilizing the extended Kalman filter algorithm, it fuses data from both the real UAV and its digital twin, and the filtered positional information is fed back into the control system of the real UAV. This enables the real-time correction of UAV positional deviations caused by sensor noise and environmental disturbances. The multisource, multimodal fusion Kalman filter method proposed in this paper significantly improves the positioning accuracy of UAVs in complex scenarios and the overall stability of the system. This method holds significant value in maintaining high-precision positioning in variable environments and has important practical implications for enhancing UAV navigation and application efficiency. Full article
(This article belongs to the Section Computer Science & Engineering)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic diagram of the multisource, multimodal data fusion positioning method.</p> Full article ">Figure 2
<p>Architecture of the multisource and multimodal data fusion system.</p> Full article ">Figure 3
<p>Real drone and geometric model of the drone.</p> Full article ">Figure 4
<p>Digital twin drone model.</p> Full article ">Figure 5
<p>Double closed-loop cascade PID control loop.</p> Full article ">Figure 6
<p>IMU pre-integration schematic diagram.</p> Full article ">Figure 7
<p>Visual–inertial fusion process.</p> Full article ">Figure 8
<p>Multisensor data fusion time alignment.</p> Full article ">Figure 9
<p>Schematic diagram of the pose graph.</p> Full article ">Figure 10
<p>UAV hardware system.</p> Full article ">Figure 11
<p>Realistic image (<b>left</b>). Twin scene image (<b>right</b>).</p> Full article ">Figure 12
<p>Data fusion experiment frame diagram.</p> Full article ">Figure 13
<p>The error curves of integrated navigation in the x, y, and z directions. The blue line represents the navigation error of the visual and IMU fusion, the green line represents the navigation error of the GPS and IMU fusion, and the red line represents the navigation error of the GPS, IMU, and visual fusion.</p> Full article ">Figure 14
<p>Visual positioning error and covariance line chart. The red line represents the covariance curve, and the blue line represents the error line.</p> Full article ">Figure 15
<p>Multimodal data fusion navigation error and covariance line. The red line represents the covariance curve, and the blue line represents the error line.</p> Full article ">Figure 16
<p>The error line graph comparing multimodal with multisource, multimodal (incorporating digital twin drone sources).</p> Full article ">Figure 17
<p>Actual flight trajectories of the drone near tall buildings. (<b>a</b>) Comparison of the three-dimensional flight trajectory lines of the drone. The red dashed line represents the trajectory data from the PX4 flight control system, while the green line represents the trajectory data from the fusion-based system. (<b>b</b>) Comparison of the two-dimensional planar trajectories on the ground station. The blue line represents the trajectory data from the fusion-based system, and the red line represents the flight trajectory data provided by the PX4 flight control system, the area marked with an orange ellipse indicates the trajectory when the drone was near tall buildings and the number of satellites was fewer than six.</p> Full article ">
18 pages, 1943 KiB  
Article
Matter Protocol Integration Using Espressif’s Solutions to Achieve Smart Home Interoperability
by Afonso Mota, Carlos Serôdio and António Valente
Electronics 2024, 13(11), 2217; https://doi.org/10.3390/electronics13112217 - 6 Jun 2024
Viewed by 1269
Abstract
Smart home devices are becoming more popular over the years. A diverse range of appliances is being created, and Ambient Intelligence is growing in homes. However, there are various producers of these gadgets, different kinds of protocols, and diverse environments. The lack of [...] Read more.
Smart home devices are becoming more popular over the years. A diverse range of appliances is being created, and Ambient Intelligence is growing in homes. However, there are various producers of these gadgets, different kinds of protocols, and diverse environments. The lack of interoperability reduces comfort of the user and turns into a barrier to smart home adoption. Matter is growing by constructing an open-source application layer protocol that can be compatible with all smart home ecosystems. In this article, a Matter overview is provided (namely, of the Commissioning stage), and a Matter Accessory using ESP32-S3 is developed referring to the manufacturer’s SDKs and is inserted into an existent household ecosystem. Its behavior on the network is briefly analyzed, and interactions with the device are carried out. The simplicity of these tasks demonstrates accessibility for developers to create products, especially when it comes to firmware. Additionally, device commissioning and control are straightforward for the consumer. This capacity of gadget incorporation into diverse ecosystems using Matter is already on the market and might result in higher device production and enhanced smart home adoption. Full article
(This article belongs to the Special Issue Advanced Wireless Sensor Networks: Applications, Challenges and Research Trends, 2nd Edition)
►▼ Show Figures

Figure 1

Figure 1
<p>Normal operation mode of Matter TCP/IP stack [<a href="#B29-electronics-13-02217" class="html-bibr">29</a>].</p> Full article ">Figure 2
<p>Matter-enabled firmware development stack for Espressif devices. Image derived from [<a href="#B40-electronics-13-02217" class="html-bibr">40</a>].</p> Full article ">Figure 3
<p>Mobile application auto-detecting Matter device.</p> Full article ">Figure 4
<p>Scanning QR code to establish secure BLE connection.</p> Full article ">Figure 5
<p>Provisioning device through BLE. (<b>a</b>) Wi-Fi selection. (<b>b</b>) Uncertified device. (<b>c</b>) Uploading information. (<b>d</b>) Connecting to Wi-Fi.</p> Full article ">Figure 6
<p>Obtained UI. (<b>a</b>) Application UI. (<b>b</b>) Device UI.</p> Full article ">Figure 7
<p>Wireshark logs.</p> Full article ">Figure 8
<p>Obtained Matter ecosystem.</p> Full article ">
20 pages, 500 KiB  
Article
Distributed Group Key Management Based on Blockchain
by Jia Ni, Guowei Fang, Yekang Zhao, Jingjing Ren, Long Chen and Yongjun Ren
Electronics 2024, 13(11), 2216; https://doi.org/10.3390/electronics13112216 - 6 Jun 2024
Viewed by 744
Abstract
Against the backdrop of rapidly advancing cloud storage technology, as well as 5G and 6G communication technologies, group key management faces increasingly daunting challenges. Traditional key management encounters difficulties in key distribution, security threats, management complexity, and issues of trustworthiness. Particularly in scenarios [...] Read more.
Against the backdrop of rapidly advancing cloud storage technology, as well as 5G and 6G communication technologies, group key management faces increasingly daunting challenges. Traditional key management encounters difficulties in key distribution, security threats, management complexity, and issues of trustworthiness. Particularly in scenarios with a large number of members or frequent member turnover within groups, this may lead to security vulnerabilities such as permission confusion, exacerbating the security risks and management complexity faced by the system. To address these issues, this paper utilizes blockchain technology to achieve distributed storage and management of group keys. This solution combines key management with the distributed characteristics of blockchain, enhancing scalability, and enabling tracking of malicious members. Simultaneously, by integrating intelligent authentication mechanisms and lightweight data update mechanisms, it effectively enhances the security, trustworthiness, and scalability of the key management system. This provides important technical support for constructing a more secure and reliable network environment. Full article
(This article belongs to the Topic Recent Advances in Security, Privacy, and Trust)
►▼ Show Figures

Figure 1

Figure 1
<p>Key management life cycle.</p> Full article ">Figure 2
<p>DGKM framework.</p> Full article ">Figure 3
<p>Node network model within the group.</p> Full article ">Figure 4
<p>The process diagram for node joining and leaving.</p> Full article ">Figure 5
<p>Comparison of the number of group nodes and computational overhead in different schemes.</p> Full article ">Figure 6
<p>Comparison results of the number of group nodes and communication costs in different schemes.</p> Full article ">Figure 7
<p>Comparison of the number of group nodes and storage overhead in different schemes.</p> Full article ">
10 pages, 1476 KiB  
Article
Temperature Dependence of Total Ionizing Dose Effects of β-Ga2O3 Schottky Barrier Diodes
by Weili Fu, Teng Ma, Zhifeng Lei, Chao Peng, Hong Zhang, Zhangang Zhang, Tao Xiao, Hongjia Song, Yuangang Wang, Jinbin Wang, Zhao Fu and Xiangli Zhong
Electronics 2024, 13(11), 2215; https://doi.org/10.3390/electronics13112215 - 6 Jun 2024
Viewed by 665
Abstract
This paper investigates the temperature-dependent effects of gamma-ray irradiation on β-Ga2O3 vertical Schottky barrier diodes (SBDs) under a 100 V reverse bias condition at a total dose of 1 Mrad(Si). As the irradiation dose increased, the radiation damage became more [...] Read more.
This paper investigates the temperature-dependent effects of gamma-ray irradiation on β-Ga2O3 vertical Schottky barrier diodes (SBDs) under a 100 V reverse bias condition at a total dose of 1 Mrad(Si). As the irradiation dose increased, the radiation damage became more severe. The total ionizing dose (TID) degradation behavior and mechanisms were evaluated through DC, capacitance–voltage (C-V), and low-frequency noise (LFN) measurements by varying irradiation, and the test results indicated that TID effects introduced interface defects and altered the carrier concentration within the material. The impact of TID effects was more pronounced at lower temperatures compared to higher temperatures. Additionally, the annealing effect in the high-temperature experimental conditions ameliorated the growth of interface trap defects caused by irradiation. These results suggest that compared to low-temperature testing, the device exhibits higher TID tolerance after high-temperature exposure, providing valuable insights for in-depth radiation reliability studies on subsequent related devices. Full article
(This article belongs to the Special Issue Radiation Effects of Advanced Electronic Devices and Circuits, 2nd Edition)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic cross-section of β-Ga<sub>2</sub>O<sub>3</sub> SBD structure.</p> Full article ">Figure 2
<p>(<b>a</b>) Forward current density (J-V) in semi-log-scale at different doses (0 krad(Si), 300 krad(Si), 500 krad(Si), and 1 Mrad(Si)) at a temperature of room temperature and (<b>b</b>) forward current density (lgJ-V) in semi-log-scale at different temperature levels (−25 °C, 0 °C, 25 °C, 50 °C, 75 °C, and 100 °C) with a cumulative dose of 1 Mrad(Si).</p> Full article ">Figure 3
<p>(<b>a</b>) Reverse current density (J-V) in semi-log-scale after different doses of gamma-ray radiation at room temperature; (<b>b</b>) reverse current density (J-V) in the semi-log-scale plot at different temperatures (−25 °C, 0 °C, 25 °C, 50 °C, 75 °C, and 100 °C) after 1 Mrad(Si) doses of gamma-ray radiation.</p> Full article ">Figure 4
<p>β-Ga<sub>2</sub>O<sub>3</sub> SBD C-V plot after 1 Mrad(Si) dose of gamma-ray radiation at different temperatures (−25 °C, 0 °C, 25 °C, 50 °C, 75 °C, and 100 °C) while the frequency is 1 MHz.</p> Full article ">Figure 5
<p>β-Ga<sub>2</sub>O<sub>3</sub> SBD LFN plot after 1 Mrad(Si) dose of gamma-ray radiation at different temperatures (−25 °C, 0 °C, 25 °C, 50 °C, 75 °C, and 100 °C) while the frequency is 100 Hz. Before is the device without radiation.</p> Full article ">
15 pages, 4633 KiB  
Article
Faster Intra-Prediction of Versatile Video Coding Using a Concatenate-Designed CNN via DCT Coefficients
by Sio-Kei Im and Ka-Hou Chan
Electronics 2024, 13(11), 2214; https://doi.org/10.3390/electronics13112214 - 6 Jun 2024
Viewed by 525
Abstract
As the next generation video coding standard, Versatile Video Coding (VVC) significantly improves coding efficiency over the current High-Efficiency Video Coding (HEVC) standard. In practice, this improvement comes at the cost of increased pre-processing complexity. This increased complexity faces the challenge of implementing [...] Read more.
As the next generation video coding standard, Versatile Video Coding (VVC) significantly improves coding efficiency over the current High-Efficiency Video Coding (HEVC) standard. In practice, this improvement comes at the cost of increased pre-processing complexity. This increased complexity faces the challenge of implementing VVC for time-consuming encoding. This work presents a technique to simplify VVC intra-prediction using Discrete Cosine Transform (DCT) feature analysis and a concatenate-designed CNN. The coefficients of the (DTC-)transformed CUs reflect the complexity of the original texture, and the proposed CNN employs multiple classifiers to predict whether they should be split. This approach can determine whether to split Coding Units (CUs) of different sizes according to the Versatile Video Coding (VVC) standard. This helps to simplify the intra-prediction process. The experimental results indicate that our approach can reduce the encoding time by 52.77% with a minimal increase of 1.48%. We use the Bjøntegaard Delta Bit Rate (BDBR) compared to the original algorithm, demonstrating a competitive result with other state-of-the-art methods in terms of coding efficiency with video quality. Full article
(This article belongs to the Special Issue Image and Video Processing Based on Deep Learning)
►▼ Show Figures

Figure 1

Figure 1
<p>The complexity of the CU structure is highly dependent on the <math display="inline"><semantics> <mi mathvariant="italic">QP</mi> </semantics></math> values that control the video quality, so more complex parts are split into smaller CUs.</p> Full article ">Figure 2
<p>Distribution of various CU sizes with respect to different <math display="inline"><semantics> <mi mathvariant="italic">QP</mi> </semantics></math>s.</p> Full article ">Figure 3
<p>The computation of encoding and decoding features. The original colour is projected into the frequency domain using DCT transformation. The zero, low, medium, and high frequencies are, respectively, filled by yellow, blue, green, and orange colours. The resulting DCT coefficients are then passed to quantisation to reduce the higher-frequency domain, resulting in a simplified feature for optimal CU determination.</p> Full article ">Figure 4
<p>The main structure of the proposed model.</p> Full article ">Figure 5
<p>An illustration of feature extraction by VGGreNet, and the process of collecting these features using the concatenate-designed CNN with multiple classifiers.</p> Full article ">
13 pages, 19038 KiB  
Article
Multi-Scale Feature Fusion Point Cloud Object Detection Based on Original Point Cloud and Projection
by Zhikang Zhang, Zhongjie Zhu, Yongqiang Bai, Yiwen Jin and Ming Wang
Electronics 2024, 13(11), 2213; https://doi.org/10.3390/electronics13112213 - 6 Jun 2024
Viewed by 916
Abstract
Existing point cloud object detection algorithms struggle to effectively capture spatial features across different scales, often resulting in inadequate responses to changes in object size and limited feature extraction capabilities, thereby affecting detection accuracy. To solve this problem, we present a point cloud [...] Read more.
Existing point cloud object detection algorithms struggle to effectively capture spatial features across different scales, often resulting in inadequate responses to changes in object size and limited feature extraction capabilities, thereby affecting detection accuracy. To solve this problem, we present a point cloud object detection method based on multi-scale feature fusion of the original point cloud and projection, which aims to improve the multi-scale performance and completeness of feature extraction in point cloud object detection. First, we designed a 3D feature extraction module based on the 3D Swin Transformer. This module pre-processes the point cloud using a 3D Patch Partition approach and employs a self-attention mechanism within a 3D sliding window, along with a downsampling strategy, to effectively extract features at different scales. At the same time, we convert the 3D point cloud to a 2D image using projection technology and extract 2D features using the Swin Transformer. A 2D/3D feature fusion module is then built to integrate 2D and 3D features at the channel level through point-by-point addition and vector concatenation to improve feature completeness. Finally, the integrated feature maps are fed into the detection head to facilitate efficient object detection. Experimental results show that our method has improved the average precision of vehicle detection by 1.01% on the KITTI dataset over three levels of difficulty compared to Voxel-RCNN. In addition, visualization analyses show that our proposed algorithm also exhibits superior performance in object detection. Full article
(This article belongs to the Topic Computer Vision and Image Processing, 2nd Edition)
►▼ Show Figures

Figure 1

Figure 1
<p>An overview of our methodology.</p> Full article ">Figure 2
<p>Spatial Swin Transformer module.</p> Full article ">Figure 3
<p>Feature fusion module.</p> Full article ">Figure 4
<p>Depth-separable convolution.</p> Full article ">Figure 5
<p>Comparison of visualisation results. The three figures (<b>a</b>–<b>c</b>) are the visualization of different scenarios, from which we can see that our proposed method has the best detection performance.</p> Full article ">
20 pages, 8417 KiB  
Article
How to Circumvent and Beat the Ransomware in Android Operating System—A Case Study of Locker.CB!tr
by Kornel Drabent, Robert Janowski and Jordi Mongay Batalla
Electronics 2024, 13(11), 2212; https://doi.org/10.3390/electronics13112212 - 6 Jun 2024
Viewed by 1080
Abstract
Ransomware is one of the most extended cyberattacks. It consists of encrypting a user’s files or locking the smartphone in order to blackmail a victim. The attacking software is ordered on the infected device from the attacker’s remote server, known as command and [...] Read more.
Ransomware is one of the most extended cyberattacks. It consists of encrypting a user’s files or locking the smartphone in order to blackmail a victim. The attacking software is ordered on the infected device from the attacker’s remote server, known as command and control. In this work, we propose a method to recover from a Locker.CB!tr ransomware attack after it has infected and hit a smartphone. The novelty of our approach lies on exploiting the communication between the ransomware on the infected device and the attacker’s command and control server as a point to reverse disruptive actions like screen locking or file encryption. For this purpose, we carried out both a dynamic and a static analysis of decompiled Locker.CB!tr ransomware source code to understand its operation principles and exploited communication patterns from the IP layer to the application layer to fully impersonate the command and control server. This way, we gained full control over the Locker.CB!tr ransomware instance. From that moment, we were able to command the Locker.CB!tr ransomware instance on the infected device to unlock the smartphone or decrypt the files. The contributions of this work are a novel method to recover the mobile phone after ransomware attack based on the analysis of the ransomware communication with the C&C server; and a mechanism for impersonating the ransomware C&C server and thus gaining full control over the ransomware instance. Full article
(This article belongs to the Special Issue Intelligent Solutions for Network and Cyber Security)
►▼ Show Figures

Figure 1

Figure 1
<p>The content of AndroidManifest.xml file from the VideoPlayer application with Locker.CB!tr hidden inside.</p> Full article ">Figure 2
<p>The result of the decompilation: the readable part of the Java code of the AbsRequest class.</p> Full article ">Figure 3
<p>The result of the decompilation: parseNetworkResponse method from the AbsRequest class, which is responsible for parsing the response from the C&amp;C server.</p> Full article ">Figure 4
<p>The code of the SmsToSendData class (on the left) and the definition of the SmsToSendInnerData type used (on the right).</p> Full article ">Figure 5
<p>Set of commands recognizable by Locker.CB!tr and the corresponding actions undertaken through the invocation of appropriate methods.</p> Full article ">Figure 6
<p>Scheme of taking control over Locker.CB!tr on an infected device by replacing the C&amp;C server with our own.</p> Full article ">Figure 7
<p>Capturing the android.util.Log class and all its methods using the Frida tool by writing the hooks in JavaScript.</p> Full article ">Figure 8
<p>LogCat messages showing the initial communication between Locker.CB!tr and its C&amp;C server just after it has infected a device.</p> Full article ">Figure 9
<p>A fragment of message exchange between Locker.CB!tr on the infected device and the C&amp;C server.</p> Full article ">Figure 10
<p>Consecutive actions of the reverse-engineering process carried by means of static and dynamic analyses.</p> Full article ">Figure 11
<p>Fragment of the code intercepting the constructor of the Java.Net.Url class and replacing the address of the C&amp;C server.</p> Full article ">Figure 12
<p>Captured logs showing an http request from Locker.CB!tr and the evidence of replacing the C&amp;C server’s IP address on the fly.</p> Full article ">Figure 13
<p>A fragment of the our C&amp;C server Go code responsible for sending commands to unlock the phone and decrypt the file system after Locker.CB!tr attack.</p> Full article ">Figure 14
<p>Implementation details of the commandRequest subroutine.</p> Full article ">Figure 15
<p>The view of the encrypted file (on the <b>left</b>) and its content after decryption (on the <b>right</b>) in a hex-viewer program (imhex).</p> Full article ">Figure 16
<p>View of a locked screen (on the <b>left</b>) after Locker.CB!tr attack and then, 2 min later, unlocked (on the <b>right</b>) thanks to the use of our fake C&amp;C server.</p> Full article ">Figure 17
<p>Logs displayed in the console of our fake C&amp;C server.</p> Full article ">
13 pages, 2655 KiB  
Article
Multi-Point Sensing via Organic Optical Fibres for FLASH Proton Therapy
by Crystal Penner, Samuel Usherovich, Sophia Andru, Camille Bélanger-Champagne, Janina Hohnholz, Boris Stoeber, Cheryl Duzenli and Cornelia Hoehr
Electronics 2024, 13(11), 2211; https://doi.org/10.3390/electronics13112211 - 6 Jun 2024
Viewed by 803
Abstract
Optical fibres are gaining popularity for relative dosimetry in proton therapy due to their spatial resolution and ability for near real-time acquisition. For FLASH proton therapy, these fibres need to handle higher dose rates and larger doses than for conventional proton dose rates. [...] Read more.
Optical fibres are gaining popularity for relative dosimetry in proton therapy due to their spatial resolution and ability for near real-time acquisition. For FLASH proton therapy, these fibres need to handle higher dose rates and larger doses than for conventional proton dose rates. We developed a multi-point fibre sensor embedded in a 3D-printed phantom which can measure the profile of a FLASH proton beam. Seven PMMA fibres of 1 mm diameter were embedded in a custom 3D-printed plastic phantom of the same density as the fibres. The phantom was placed in a proton beam with FLASH dose rates at the TRIUMF Proton Therapy Research Centre (PTRC). The sensor was exposed to different proton energies, 13.5 MeV, 19 MeV and 40.4 MeV, achieved by adding PMMA bolus in front of the phantom and three different beam currents, varying the dose rates from 7.5 to 101 Gy/s. The array was able to record beam profiles in both transverse and axial directions in relative agreement with measurements from EBT-XD radiochromic films (transverse) and Monte Carlo simulations (axial). A decrease in light output over time was observed, which might be caused by radiation damage in the matrix of the fibre and characterised by an exponential decay function. Full article
(This article belongs to the Special Issue Applications of Optical Fiber Sensors)
►▼ Show Figures

Figure 1

Figure 1
<p>Light output from PMMA measured by the MPPC photodetector as a function of average dose rate over 4–6 trials at three different energies. Good linearity was established [<a href="#B18-electronics-13-02211" class="html-bibr">18</a>].</p> Full article ">Figure 2
<p>Schematic showing one channel of the seven channels used in this study, including sensing fibre, fibre extensions and proton irradiation direction.</p> Full article ">Figure 3
<p>(<b>a</b>) Beamline showing transverse phantom attached to the 2D stage, 60 mm downstream from the secondary collimator. (<b>b</b>) Axial phantom design with fibre channels at the leading edge.</p> Full article ">Figure 4
<p>Film showing beam transverse profile and fibre placed over the beam central region.</p> Full article ">Figure 5
<p>(<b>a</b>–<b>c</b>) Each current for the three energies tested, with the individual measurements at each energy shown as 1, 2, and 3. There is a noticeable decrease in signal from one series of irradiations to the next. Position 0 mm is the most central fibre position; however, the fact that the beam itself is not perfectly symmetrical is evident in all profiles. (<b>d</b>–<b>f</b>) Comparison between film and fibre response at 2 nA for all three energies. Results for the film and fibres are normalised to 1 at their peaks of the second irradiation. Film profiles appear to widen relative to the fibres as energy decreases below 40.4 MeV in (<b>e</b>) and even more so in (<b>f</b>). This could be a result of saturation in the films in the peaks at the lower energies—an expected result considering the dose levels received at these energies.</p> Full article ">Figure 6
<p>(<b>a</b>) Six irradiations of a pristine portion of PMMA optical fibre with a current of 9 nA at 19 MeV. The Figure includes recovery time between irradiations. (<b>b</b>) The beginning and end signal/nA for the five irradiations with virtually identical irradiation durations. The fits shown are polynomial functions to show the consistency of the attenuation from subsequent 1,000,000 MC (around 21,000 Gy) irradiations. Data are normalised to the current-per-second during each irradiation.</p> Full article ">Figure 7
<p>Depth dose comparison of five fibres at three energies of protons to the FLUKA simulation at 73 MeV. The overall beam shape and Bragg peak position are in very good agreement.</p> Full article ">
15 pages, 2252 KiB  
Article
Artificial Intelligence Implementation in Internet of Things Embedded System for Real-Time Person Presence in Bed Detection and Sleep Behaviour Monitor
by Minh Long Hoang, Guido Matrella and Paolo Ciampolini
Electronics 2024, 13(11), 2210; https://doi.org/10.3390/electronics13112210 - 6 Jun 2024
Cited by 1 | Viewed by 1223
Abstract
This paper works on detecting a person in bed for sleep routine and sleep pattern monitoring based on the Micro-Electro-Mechanical Systems (MEMS) accelerometer and Internet of Things (IoT) embedded system board. This work provides sleep information, patient assessment, and elderly care for patients [...] Read more.
This paper works on detecting a person in bed for sleep routine and sleep pattern monitoring based on the Micro-Electro-Mechanical Systems (MEMS) accelerometer and Internet of Things (IoT) embedded system board. This work provides sleep information, patient assessment, and elderly care for patients who live alone via tele-distance to doctors or family members. About 216,000 pieces of acceleration data were collected, including three classes: no person in bed, a static laying position, and a moving state for Artificial Intelligence (AI) application. Six well-known Machine-Learning (ML) algorithms were evaluated with precision, recall, F1-score, and accuracy in the workstation before implementing in the STM32-microcontroller for real-time state classification. The four best algorithms were selected to be programmed into the IoT board and applied for real-time testing. The results demonstrate the high accuracy of the ML performance, more than 99%, and the Classification and Regression Tree algorithm is among the best models with a light code size of 1583 bytes. The smart bed information is sent to the IoT dashboard of Node-RED via a Message Queuing Telemetry broker (MQTT). Full article
(This article belongs to the Special Issue Internet of Things, Embedded Solutions, and Edge Intelligence for Smart Health)
►▼ Show Figures

Figure 1

Figure 1
<p>Data acquisition diagram.</p> Full article ">Figure 2
<p>AI implementation in microcontroller (MCU) under IoT communication.</p> Full article ">Figure 3
<p>Encapsulated accelerometer and MCU platform under the bed frame.</p> Full article ">Figure 4
<p>Smart bed under test.</p> Full article ">Figure 5
<p>Raw data in 3 states.</p> Full article ">Figure 6
<p>K-fold cross-validation for model evaluations.</p> Full article ">Figure 7
<p>K-Node-RED block connection and IoT dashboard.</p> Full article ">
26 pages, 1958 KiB  
Review
Digital Twins in Agriculture: A Review of Recent Progress and Open Issues
by Li Wang
Electronics 2024, 13(11), 2209; https://doi.org/10.3390/electronics13112209 - 5 Jun 2024
Cited by 3 | Viewed by 2001
Abstract
Digital twin technology is expected to transform agriculture. By creating the virtual representation of a physical entity, it assists food producers in monitoring, predicting, and optimizing the production process remotely and even autonomously. However, the progress in this area is relatively slower than [...] Read more.
Digital twin technology is expected to transform agriculture. By creating the virtual representation of a physical entity, it assists food producers in monitoring, predicting, and optimizing the production process remotely and even autonomously. However, the progress in this area is relatively slower than in industries like manufacturing. A systematic investigation of agricultural digital twins’ current status and progress is imperative. With seventy published papers, this work elaborated on the studies targeting agricultural digital twins from overall trends, focused areas (including domains, processes, and topics), reference architectures, and open questions, which could help scholars examine their research agenda and support the further development of digital twins in agriculture. Full article
(This article belongs to the Special Issue Digital Twins in Industry 4.0)
►▼ Show Figures

Figure 1

Figure 1
<p>Types of published works in this study. Note: Those published at the conference first, but compiled as a book chapter afterward, were categorized as the conference paper.</p> Full article ">Figure 2
<p>Number of published works per year after 2020.</p> Full article ">Figure 3
<p>Distribution of publication types.</p> Full article ">Figure 4
<p>Agricultural domains mentioned in 2021–2023 studies.</p> Full article ">Figure 5
<p>Types of publications in different agricultural domains.</p> Full article ">Figure 6
<p>Designed digital twins for different crop production processes.</p> Full article ">Figure 7
<p>Architectural framework of agricultural digital twins.</p> Full article ">
23 pages, 1627 KiB  
Article
Hybrid State Estimation: Integrating Physics-Informed Neural Networks with Adaptive UKF for Dynamic Systems
by J. de Curtò and I. de Zarzà
Electronics 2024, 13(11), 2208; https://doi.org/10.3390/electronics13112208 - 5 Jun 2024
Cited by 1 | Viewed by 920
Abstract
In this paper, we present a novel approach to state estimation in dynamic systems by combining Physics-Informed Neural Networks (PINNs) with an adaptive Unscented Kalman Filter (UKF). Recognizing the limitations of traditional state estimation methods, we refine the PINN architecture with hybrid loss [...] Read more.
In this paper, we present a novel approach to state estimation in dynamic systems by combining Physics-Informed Neural Networks (PINNs) with an adaptive Unscented Kalman Filter (UKF). Recognizing the limitations of traditional state estimation methods, we refine the PINN architecture with hybrid loss functions and Monte Carlo Dropout for enhanced uncertainty estimation. The Unscented Kalman Filter is augmented with an adaptive noise covariance mechanism and incorporates model parameters into the state vector to improve adaptability. We further validate this hybrid framework by integrating the enhanced PINN with the UKF for a seamless state prediction pipeline, demonstrating significant improvements in accuracy and robustness. Our experimental results show a marked enhancement in state estimation fidelity for both position and velocity tracking, supported by uncertainty quantification via Bayesian inference and Monte Carlo Dropout. We further extend the simulation and present evaluations on a double pendulum system and state estimation on a quadcopter drone. This comprehensive solution is poised to advance the state-of-the-art in dynamic system estimation, providing unparalleled performance across control theory, machine learning, and numerical optimization domains. Full article
(This article belongs to the Special Issue Cyber-Physical Systems: Recent Developments and Emerging Trends)
►▼ Show Figures

Figure 1

Figure 1
<p>Training loss curves for the PINN model with Monte Carlo Dropout.</p> Full article ">Figure 2
<p>True vs. estimated positions using the hybrid framework.</p> Full article ">Figure 3
<p>True vs. estimated velocities using the hybrid framework.</p> Full article ">Figure 4
<p>Training loss curves for the improved PINN model with Monte Carlo Dropout.</p> Full article ">Figure 5
<p>True vs. estimated positions using the improved hybrid framework.</p> Full article ">Figure 6
<p>True vs. estimated velocities using the improved hybrid framework.</p> Full article ">Figure 7
<p>True vs. estimated positions using EKF.</p> Full article ">Figure 8
<p>True vs. estimated velocities using EKF.</p> Full article ">Figure 9
<p>Training loss history with the enhanced physics-informed loss function.</p> Full article ">Figure 10
<p>True vs. estimated angles (<math display="inline"><semantics> <msub> <mi>θ</mi> <mn>1</mn> </msub> </semantics></math> and <math display="inline"><semantics> <msub> <mi>θ</mi> <mn>2</mn> </msub> </semantics></math>) with the enhanced model.</p> Full article ">Figure 11
<p>True vs. estimated angular velocities (<math display="inline"><semantics> <mrow> <mi>d</mi> <msub> <mi>θ</mi> <mn>1</mn> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>d</mi> <msub> <mi>θ</mi> <mn>2</mn> </msub> </mrow> </semantics></math>) with the enhanced model.</p> Full article ">Figure 12
<p>Training loss curves for the PINN model with Monte Carlo Dropout.</p> Full article ">Figure 13
<p>True vs. estimated position over time for the quadcopter simulation.</p> Full article ">Figure 14
<p>True vs. estimated orientation over time for the quadcopter simulation.</p> Full article ">Figure 15
<p>True vs. estimated linear velocities over time for the quadcopter simulation.</p> Full article ">Figure 16
<p>True vs. estimated angular velocities over time for the quadcopter simulation.</p> Full article ">
20 pages, 6916 KiB  
Article
An Improved YOLOv5 Algorithm for Tyre Defect Detection
by Mujun Xie, Heyu Bian, Changhong Jiang, Zhong Zheng and Wei Wang
Electronics 2024, 13(11), 2207; https://doi.org/10.3390/electronics13112207 - 5 Jun 2024
Viewed by 846
Abstract
In this study, a tyre defect detection model is improved and optimized under the YOLOv5 framework, aiming at radial tyre defects with characteristics such as an elongated shape and various target sizes and defect types. The DySneakConv module is introduced to replace the [...] Read more.
In this study, a tyre defect detection model is improved and optimized under the YOLOv5 framework, aiming at radial tyre defects with characteristics such as an elongated shape and various target sizes and defect types. The DySneakConv module is introduced to replace the first BotteneckCSP in the Backbone network. The deformation offset of the DySneakConv module is used to make the convolutional energy freely adapt to the structure to improve the recognition rate of tyre defects with elongated features; the AIFI module is introduced to replace the fourth BotteneckCSP, and the self-attention mechanism and the processing of large-scale features are used to improve the recognition rate of tyre defects with elongated features using the AIFI module. This latter module has a self-attention mechanism and the ability to handle large-scale features to solve the problems of diverse tyre defects and different sizes. Secondly, the CARAFE up-sampling operator is introduced to replace the up-sampling operator in the Neck network. The up-sampling kernel prediction module in the CARAFE operator is used to increase the receptive field and allow the feature reorganization module to capture more semantic information to overcome the information loss problem of the up-sampling operator. Finally, based on the improved YOLOv5 detection algorithm, the Channel-wise Knowledge Distillation algorithm lightens the model, reducing its computational requirements and size while ensuring detection accuracy. Experimental studies were conducted on a dataset containing four types of tyre defects. Experimental results for the training set show that the improved algorithm improves the mAP0.5 by 4.6 pp, reduces the model size by 25.6 MB, reduces the computational complexity of the model by 31.3 GFLOPs, and reduces the number of parameters by 12.7 × 106 compared to the original YOLOv5m algorithm. Experimental results for the test set show that the improved algorithm improves the mAP0.5 by 2.6 pp compared to the original YOLOv5m algorithm. This suggests that the improved algorithm is more suitable for tyre defect detection than the original YOLOv5. Full article
(This article belongs to the Special Issue Fault Detection Technology Based on Deep Learning)
►▼ Show Figures

Figure 1

Figure 1
<p>YOLOv5 network framework.</p> Full article ">Figure 2
<p>Calculation of DySneakConv coordinates.</p> Full article ">Figure 3
<p>True feeling range of DySneakConv.</p> Full article ">Figure 4
<p>General structure of AIFI.</p> Full article ">Figure 5
<p>Working principle of AIFI.</p> Full article ">Figure 6
<p>General structure of CARAFE.</p> Full article ">Figure 7
<p>Improved YOLOv5 network framework.</p> Full article ">Figure 8
<p>Working principle of knowledge distillation.</p> Full article ">Figure 9
<p>Overall structure of Channel-wise Knowledge Distillation.</p> Full article ">Figure 10
<p>Tyre defect characteristics.</p> Full article ">Figure 11
<p>Detection results for the four defect types.</p> Full article ">Figure 12
<p>Training results of the two algorithms.</p> Full article ">Figure 13
<p>Comparison of the two algorithms on a test set.</p> Full article ">Figure 14
<p>Training results of other comparison algorithms.</p> Full article ">Figure 14 Cont.
<p>Training results of other comparison algorithms.</p> Full article ">
18 pages, 730 KiB  
Article
SURE: Structure for Unambiguous Requirement Expression in Natural Language
by Franklin Parrales-Bravo, Rosangela Caicedo-Quiroz, Julio Barzola-Monteses, Leonel Vasquez-Cevallos, María Isabel Galarza-Soledispa and Manuel Reyes-Wagnio
Electronics 2024, 13(11), 2206; https://doi.org/10.3390/electronics13112206 - 5 Jun 2024
Cited by 1 | Viewed by 624
Abstract
This study presents three structures for clearly expressing functional requirements (FRs) and quantitative non-functional requirements (qt-NFRs). Expressing requirements with these structures will allow the understanding of requirements by stakeholders and software developers. The first structure is the SURE format, which is composed of [...] Read more.
This study presents three structures for clearly expressing functional requirements (FRs) and quantitative non-functional requirements (qt-NFRs). Expressing requirements with these structures will allow the understanding of requirements by stakeholders and software developers. The first structure is the SURE format, which is composed of three main sections: a title, a short definition, and a detailed description. The second proposed structure is a template to facilitate the definition of the title and description of unambiguous FRs. It is based on the application of CRUD operations on a certain entity, calling it the “CRUDE” structure. Finally, the third structure serves as a template to make it easier to clearly define the description and title of qt-NFRs. It is based on the application of system properties to computer events or actions, calling it the “PROSE” structure. In this, it is very important to specify those metric values that are desired or expected by the stakeholder. To know how much the definition of FRs and qt-NFRs improved when the proposed structures were used, 46 requirement specification documents elaborated as homework by students of the “Requirement Engineering” course offered at the University of Guayaquil between 2020 and 2022 were evaluated by five experts with more than 10 years of experience in software development for Ecuadorian companies. The findings showed that students reduced the percentage of unambiguous FRs and qt-NFRs from over 80% to about 10%. In conclusion, the findings demonstrate how crucial the three structures proposed in this paper are to helping students develop the ability to clearly express requirements. Full article
(This article belongs to the Special Issue Advances in Software Engineering and Programming Languages)
►▼ Show Figures

Figure 1

Figure 1
<p>Structure of the SURE format, used to specify both user and system requirements in software projects.</p> Full article ">Figure 2
<p>Steps to specify both user and system requirements in software projects through the SURE format.</p> Full article ">Figure 3
<p>Example of a FR expressed in the SURE format.</p> Full article ">Figure 4
<p>Example of a qt-NFR expressed in the SURE format. Technical words are marked with blue and red text.</p> Full article ">Figure 5
<p>Evaluation of FRs collected before explaining the CRUDE structure to students during the period 2020–2022.</p> Full article ">Figure 6
<p>Evaluation of FRs collected after explaining the CRUDE structure to students during the period 2020–2022.</p> Full article ">Figure 7
<p>Evaluation of the SURE format sections on those FRs indicated as ambiguous before explaining the CRUDE structure.</p> Full article ">Figure 8
<p>Evaluation of the SURE format sections on those FRs indicated as ambiguous after explaining the CRUDE structure.</p> Full article ">Figure 9
<p>Evaluation of qt-NFRs collected before explaining the PROSE structure during the period 2020–2022.</p> Full article ">Figure 10
<p>Evaluation of qt-NFRs collected after explaining the PROSE structure during the period 2020–2022.</p> Full article ">Figure 11
<p>Evaluation of the SURE format sections on those qt-NFRs indicated as ambiguous before explaining the PROSE structure.</p> Full article ">Figure 12
<p>Evaluation of the SURE format sections on those requirements indicated as ambiguous after explaining the PROSE structure.</p> Full article ">
17 pages, 1373 KiB  
Article
Chip-Level Defect Analysis with Virtual Bad Wafers Based on Huge Big Data Handling for Semiconductor Production
by Jinsik Kim and Inwhee Joe
Electronics 2024, 13(11), 2205; https://doi.org/10.3390/electronics13112205 - 5 Jun 2024
Viewed by 896
Abstract
Semiconductors continue to shrink in die size because of benefits like cost savings, lower power consumption, and improved performance. However, this reduction leads to more defects due to increased inter-cell interference. Among the various defect types, customer-found defects are the most costly. Thus, [...] Read more.
Semiconductors continue to shrink in die size because of benefits like cost savings, lower power consumption, and improved performance. However, this reduction leads to more defects due to increased inter-cell interference. Among the various defect types, customer-found defects are the most costly. Thus, finding the root cause of customer-found defects has become crucial to the quality of semiconductors. Traditional methods involve analyzing the pathways of many low-yield wafers. Yet, because of the extremely limited number of customer-found defects, obtaining significant results is difficult. After the products are provided to customers, they undergo rigorous testing and selection, leading to a very low defect rate. However, since the timing of defect occurrence varies depending on the environment in which the product is used, the quantity of defective samples is often quite small. Unfortunately, with such a low number of samples, typically 10 or fewer, it becomes impossible to investigate the root cause of wafer-level defects using conventional methods. This paper introduces a novel approach to finding the root cause of these rare defective chips for the first time in the semiconductor industry. Defective wafers are identified using rare customer-found chips and chip-level EDS (Electrical Die Sorting) data, and these newly identified defective wafers are termed vBADs (virtual bad wafers). The performance of root cause analysis is dramatically improved with vBADs. However, the chip-level analysis presented here demands substantial computing power. Therefore, MPP (Massive Parallel Processing) architecture is implemented and optimized to handle large volumes of chip-level data within a large architecture infrastructure that can manage big data. This allows for a chip-level defect analysis system that can recommend the relevant EDS test and identify the root cause in real time even with a single defective chip. The experimental results demonstrate that the proposed root cause search can reveal the hidden cause of a single defective chip by amplifying it with 90 vBADs, and system performance improves by a factor of 61. Full article
(This article belongs to the Section Industrial Electronics)
►▼ Show Figures

Figure 1

Figure 1
<p>Overall procedure of semiconductor production. FAB (fabrication) and EDS (Electrical Die Sorting) are the most important steps in the semiconductor field. Customer-found defective chips are analyzed to find the root cause and prevent recurrence.</p> Full article ">Figure 2
<p>Potential latent reliability defect and killer defect. Potential latent defect (middle) is not detected at t = 0, but it will fail after being embedded in the customer’s electronic device.</p> Full article ">Figure 3
<p>Potential latent reliability defect in EDS wafer map. Red chips are defective chips. The normal white chip surrounded by defective chips can be a reliability defective chip.</p> Full article ">Figure 4
<p>Functional scheme for selection of a critical EDS test item and latent defect discovery.</p> Full article ">Figure 5
<p>High-level system architecture.</p> Full article ">Figure 6
<p>Distributed and asynchronous architecture. (<b>a</b>) Synchronous job execution. (<b>b</b>) Distributed and asynchronous job execution.</p> Full article ">Figure 7
<p>Visualizations for defective chip analysis. (<b>a</b>) The scatter chart of chips shows that a defective chip’s measurement value is in the outlier range. (<b>b</b>) The wafer map chart shows that the defective chip is in the line patterns.</p> Full article ">Figure 8
<p>Similar defective wafers discovery. Using the selected test item and the defective chip measurement value, similar wafers can be found.</p> Full article ">Figure 9
<p>Root cause analysis with vBADs (virtual bad wafers).</p> Full article ">Figure 10
<p>Root cause search performance increased by vBADs.</p> Full article ">Figure 11
<p>System parameters affecting the performance. (<b>a</b>) Data size affects the performance. 50 GB on average, 200 s elapsed. (<b>b</b>) The number of test items affects the performance. Test item counts are 2200, and 120 s elapsed.</p> Full article ">Figure 12
<p>Performance improvement by MPP nodes counts.</p> Full article ">
23 pages, 1927 KiB  
Article
D2Former: Dual-Domain Transformer for Change Detection in VHR Remote Sensing Images
by Huanhuan Zheng, Hui Liu, Lei Lu, Shiyin Li and Jiyan Lin
Electronics 2024, 13(11), 2204; https://doi.org/10.3390/electronics13112204 - 5 Jun 2024
Viewed by 575
Abstract
Computational intelligence technologies have been extensively applied for the interpretation of remote sensing imagery. Recently, the computational-intelligence-based Transformer change detection (CD) approach has attracted increasing attention. However, the current Transformer-based CD method can better capture global features, but there is no good solution [...] Read more.
Computational intelligence technologies have been extensively applied for the interpretation of remote sensing imagery. Recently, the computational-intelligence-based Transformer change detection (CD) approach has attracted increasing attention. However, the current Transformer-based CD method can better capture global features, but there is no good solution for the loss of local detail information. For this reason, introducing semantic and frequency information from the perspective of a dual-domain can be beneficial for improving the representation of detailed features to improve CD performance. To overcome this limitation, a dual-domain Transformer (D2Former) is proposed for CD. Firstly, we adopt a semantic tokenizer to capture the semantic information, which promotes the enrichment and refinement of semantic change information in the Transformer. Secondly, a frequency tokenizer is introduced to acquire the frequency information of the features, which offers the proposed D2Former another aspect and dimension to enhance the ability to detect change information. Therefore, the proposed D2Former employs dual-domain tokenizers to acquire and fuse the feature representation with rich semantic and frequency information, which can refine the features to acquire more fine-grained CD ability. Extensive experiments on three CD benchmark datasets demonstrate that the proposed D2Former obviously outperforms some other existing approaches. The results present the competitive performance of our method on the WHU-CD, LEVIR-CD, and GZ-CD datasets, for which it achieved F1-Score metrics of 92.85%, 90.60%, and 87.02%, respectively. Full article
(This article belongs to the Topic Computational Intelligence in Remote Sensing: 2nd Edition)
►▼ Show Figures

Figure 1

Figure 1
<p>Overview of the proposed D<sup>2</sup>Former. The pre-change image and post-change image denote the T1 and T2 images, respectively.</p> Full article ">Figure 2
<p>The structure of the semantic tokenizer. <math display="inline"><semantics> <msup> <mi>T</mi> <mi>L</mi> </msup> </semantics></math> indicates the token length.</p> Full article ">Figure 3
<p>Illustration of the proposed frequency tokenizer. <math display="inline"><semantics> <msup> <mi>T</mi> <mi>L</mi> </msup> </semantics></math> indicates the token length.</p> Full article ">Figure 4
<p>Some examples from the three experimental datasets: (<b>a</b>) WHU-CD dataset, (<b>b</b>) LEVIR-CD dataset, and (<b>c</b>) GZ-CD dataset.</p> Full article ">Figure 5
<p>The visual results for different methods on WHU-CD dataset: (<b>a</b>) T1-time image, (<b>b</b>) T2-time image, (<b>c</b>) ground truth image, (<b>d</b>) FC-EF [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>e</b>) FC-Siam-C [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>f</b>) FC-Siam-D [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>g</b>) STANet [<a href="#B30-electronics-13-02204" class="html-bibr">30</a>], (<b>h</b>) BIT [<a href="#B36-electronics-13-02204" class="html-bibr">36</a>], (<b>i</b>) ChangeFormer [<a href="#B37-electronics-13-02204" class="html-bibr">37</a>], and (<b>j</b>) proposed D<sup>2</sup>Former. Note: white, black, red, and green denote true positive pixels, true negative pixels, false positive pixels, and false negative pixels, respectively.</p> Full article ">Figure 6
<p>The visual results for different methods on the LEVIR-CD dataset: (<b>a</b>) T1-time image, (<b>b</b>) T2-time image, (<b>c</b>) ground truth image, (<b>d</b>) FC-EF [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>e</b>) FC-Siam-C [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>f</b>) FC-Siam-D [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>g</b>) STANet [<a href="#B30-electronics-13-02204" class="html-bibr">30</a>], (<b>h</b>) BIT [<a href="#B36-electronics-13-02204" class="html-bibr">36</a>], (<b>i</b>) ChangeFormer [<a href="#B37-electronics-13-02204" class="html-bibr">37</a>], and (<b>j</b>) proposed D<sup>2</sup>Former. Note: white, black, red, and green denote true positive pixels, true negative pixels, false positive pixels, and false negative pixels, respectively.</p> Full article ">Figure 7
<p>The visual results of different methods on the GZ-CD dataset: (<b>a</b>) T1-time image, (<b>b</b>) T2-time image, (<b>c</b>) ground truth image, (<b>d</b>) FC-EF [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>e</b>) FC-Siam-C [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>f</b>) FC-Siam-D [<a href="#B29-electronics-13-02204" class="html-bibr">29</a>], (<b>g</b>) STANet [<a href="#B30-electronics-13-02204" class="html-bibr">30</a>], (<b>h</b>) BIT [<a href="#B36-electronics-13-02204" class="html-bibr">36</a>], (<b>i</b>) ChangeFormer [<a href="#B37-electronics-13-02204" class="html-bibr">37</a>], and (<b>j</b>) proposed D<sup>2</sup>Former. Note: white, black, red, and green denote true positive pixels, true negative pixels, false positive pixels, and false negative pixels, respectively.</p> Full article ">Figure 8
<p>The results on different datasets for ablation studies: (<b>a</b>) T1-time image, (<b>b</b>) T2-time image, (<b>c</b>) ground truth image, (<b>d</b>) backbone, (<b>e</b>) w/semantic tokenizer, (<b>f</b>) w/frequency tokenizer, (<b>g</b>) full (proposed D<sup>2</sup>Former). Note: white, black, red, and green denote true positive pixels, true negative pixels, false positive pixels, and false negative pixels, respectively.</p> Full article ">Figure 9
<p>The feature maps for the combinations of different components: (<b>a</b>) T1-time image, (<b>b</b>) T2-time image, (<b>c</b>) ground truth image, (<b>d</b>) backbone, (<b>e</b>) w/semantic tokenizer, (<b>f</b>) w/frequency tokenizer, and (<b>g</b>) full (proposed D<sup>2</sup>Former).</p> Full article ">
16 pages, 1680 KiB  
Article
Short-Term Flood Prediction Model Based on Pre-Training Enhancement
by Yang Xia and Jiamin Lu
Electronics 2024, 13(11), 2203; https://doi.org/10.3390/electronics13112203 - 5 Jun 2024
Cited by 1 | Viewed by 624
Abstract
With the rapid advancement of deep learning techniques, deep learning-based flood prediction models have drawn significant attention. However, for short-term prediction in small- and medium-sized river basins, models typically rely on hydrological data spanning from the past several hours to one day, utilizing [...] Read more.
With the rapid advancement of deep learning techniques, deep learning-based flood prediction models have drawn significant attention. However, for short-term prediction in small- and medium-sized river basins, models typically rely on hydrological data spanning from the past several hours to one day, utilizing a fixed-length input window. Such input limits the models’ adaptability to diverse rainfall events and restricts their capability to comprehend historical temporal patterns. To address the underutilization of historical information by existing models, we introduce a Pre-training Enhanced Short-term Flood Prediction Method (PE-SFPM) to enrich the model’s temporal understanding without necessitating changes to the input window size. In the pre-training stage, the model uses a random masking and prediction strategy to learn segment features, capturing a more comprehensive evolutionary trend of historical floods. In the flow forecasting stage, temporal features and spatial features are captured and fused using the temporal attention, spatial attention, and gated fusion. Features are further enhanced by integrating segment features using a feed-forward network. Experimental results demonstrate that the proposed PE-SFPM model achieves excellent performance in short-term flood prediction tasks. Full article
(This article belongs to the Special Issue AI in Disaster, Crisis, and Emergency Management)
►▼ Show Figures

Figure 1

Figure 1
<p>Overall architecture of the proposed method PE-SFPM.</p> Full article ">Figure 2
<p>MAE error of the prediction results of each model at multiple time steps.</p> Full article ">Figure 3
<p>RMSE error of the prediction results of each model at multiple time steps.</p> Full article ">Figure 4
<p>MAPE error of the prediction results of each model at multiple time steps.</p> Full article ">Figure 5
<p>Actual forecasting performance of flood prediction models on the ChangHua dataset.</p> Full article ">Figure 6
<p>Actual forecasting performance of flood prediction models on the HeiHe Dataset.</p> Full article ">Figure 7
<p>Actual forecasting performance of flood prediction models on the TunXi dataset.</p> Full article ">
17 pages, 6521 KiB  
Article
Enhancing Signal Recognition Accuracy in Delay-Based Optical Reservoir Computing: A Comparative Analysis of Training Algorithms
by Ruibo Zhang, Tianxiang Luan, Shuo Li, Chao Wang and Ailing Zhang
Electronics 2024, 13(11), 2202; https://doi.org/10.3390/electronics13112202 - 5 Jun 2024
Viewed by 706
Abstract
To improve the accuracy of signal recognition in delay-based optical reservoir computing (RC) systems, this paper proposes the use of nonlinear algorithms at the output layer to replace traditional linear algorithms for training and testing datasets and apply them to the identification of [...] Read more.
To improve the accuracy of signal recognition in delay-based optical reservoir computing (RC) systems, this paper proposes the use of nonlinear algorithms at the output layer to replace traditional linear algorithms for training and testing datasets and apply them to the identification of frequency-modulated continuous wave (FMCW) LiDAR signals. This marks the inaugural use of the system for the identification of FMCW LiDAR signals. We elaborate on the fundamental principles of a delay-based optical RC system using an optical-injected distributed feedback laser (DFB) laser and discriminate four FMCW LiDAR signals through this setup. In the output layer, three distinct training algorithms—namely linear regression, support vector machine (SVM), and random forest—were employed to train the optical reservoir. Upon analyzing the experimental results, it was found that regardless of the size of the dataset, the recognition accuracy of the two nonlinear training algorithms was superior to that of the linear regression algorithm. Among the two nonlinear algorithms, the Random Forest algorithm had a higher recognition accuracy than SVM when the sample size was relatively small. Full article
(This article belongs to the Special Issue Artificial Intelligence and Signal Processing: Circuits and Systems)
►▼ Show Figures

Figure 1

Figure 1
<p>Experimental structure of the delay-based optical reservoir computing. TL: tunable laser, AWG: arbitrary waveform generator. MZM: Mach-Zehnder modulator, VA: variable attenuator, DSA: digital signal oscilloscope, DFB: distributed feedback laser, ISO: isolator, PC: polarization controller, PD: Photodetector, OSA: optical spectrum analyzer, EYDFA: erbium/ytterbium co-doped fiber amplifier.</p> Full article ">Figure 2
<p>Applying a mask in the input layer.</p> Full article ">Figure 3
<p>Distribution map of injection locking and unlocking regions in DFB.</p> Full article ">Figure 4
<p>The working condition of the DFB laser captured by the OSA.</p> Full article ">Figure 5
<p>Measurement feedback delay schematic diagram.</p> Full article ">Figure 6
<p>Pulse interval. The dashed box indicates the result of zooming in on a single cycle.</p> Full article ">Figure 7
<p>Classification principle of random forest.</p> Full article ">Figure 8
<p>Schematic of the FMCW LiDAR system. (<b>a</b>) is a nonlinear sweep frequency light source, (<b>b-i</b>) and (<b>b-ii</b>) is two different radar signals at different distances.</p> Full article ">Figure 9
<p>Four types of input signals. (<b>a</b>–<b>d</b>) are the FMCW LiDAR signals at four different distances generated by simulating the experimental principle in <a href="#electronics-13-02202-f008" class="html-fig">Figure 8</a> using MatlabR2021b.</p> Full article ">Figure 10
<p>Four types of pre-modulated signals. (<b>a</b>–<b>d</b>) represent the pre modulated signal formed by adding masks to the four signals in <a href="#electronics-13-02202-f009" class="html-fig">Figure 9</a>.</p> Full article ">Figure 11
<p>Frequency of relaxation oscillations in the DFB laser.</p> Full article ">Figure 12
<p>DSA collection results. (<b>a</b>–<b>d</b>) represent the results collected on the DSA after the four signals shown in <a href="#electronics-13-02202-f010" class="html-fig">Figure 10</a> pass through the delay optics RC.</p> Full article ">Figure 13
<p>The virtual node status of four types of modulated signals. (<b>a</b>–<b>d</b>) represent the results after downsampling. This is the dataset that we need to train on the computer.</p> Full article ">Figure 14
<p>Classification results between true values and predicted values, based on linear regression training algorithm.</p> Full article ">Figure 15
<p>SVM recognition results.</p> Full article ">Figure 16
<p>Random Forest recognition results.</p> Full article ">Figure 17
<p>(<b>a</b>) Linear regression algorithm test results, (<b>b</b>) SVM algorithm test results, (<b>c</b>) Random forest algorithm test results.</p> Full article ">Figure 18
<p>The recognition accuracy of three algorithms under different dataset sizes.</p> Full article ">
11 pages, 1571 KiB  
Article
Detection of Train Wheelset Tread Defects with Small Samples Based on Local Inference Constraint Network
by Jianhua Liu, Shiyi Jiang, Zhongmei Wang and Jiahao Liu
Electronics 2024, 13(11), 2201; https://doi.org/10.3390/electronics13112201 - 5 Jun 2024
Viewed by 727
Abstract
Due to the long-term service through wheel-rail rolling contact, the train wheelset tread will inevitably suffer from different types of defects, such as wear, cracks, and scratches. The effective detection of wheelset tread defects can provide critical support for the operation and maintenance [...] Read more.
Due to the long-term service through wheel-rail rolling contact, the train wheelset tread will inevitably suffer from different types of defects, such as wear, cracks, and scratches. The effective detection of wheelset tread defects can provide critical support for the operation and maintenance of trains. In this paper, a new method based on a local inference constraint network is proposed to detect wheelset tread defects, and the main purpose is to address the issue of insufficient feature spaces caused by small samples. First, a generative adversarial network is applied to generate diverse samples with semantic consistency. An attention mechanism module is introduced into the feature extraction network to increase the importance of defect features. Then, the residual spine network for local input decisions is constructed to establish an association between sample features and defect types. Furthermore, the network’s activation function is improved to obtain higher learning speed and accuracy with fewer parameters. Finally, the validity and feasibility of the proposed method are verified using experimental data. Full article
(This article belongs to the Special Issue Machine Vision in Industrial Systems)
►▼ Show Figures

Figure 1

Figure 1
<p>Local inference constraint network-based detection method of train wheelset tread defect.</p> Full article ">Figure 2
<p>Defect feature enhancement with attention mechanism.</p> Full article ">Figure 3
<p>Residual spinal fully connected layer.</p> Full article ">Figure 4
<p>The data generated by Non, GT, PT, and Gan.</p> Full article ">
16 pages, 4054 KiB  
Article
Noise-like-Signal-Based Sub-Synchronous Oscillation Prediction for a Wind Farm with Doubly-Fed Induction Generators
by Junjie Ma, Linxing Lyu, Junfeng Man, Mengqi Chen and Yijun Cheng
Electronics 2024, 13(11), 2200; https://doi.org/10.3390/electronics13112200 - 5 Jun 2024
Viewed by 629
Abstract
The DFIG-based wind farm faces sub-synchronous oscillation (SSO) when it is integrated with a series-compensated transmission system. The equivalent SSO damping is influenced by both wind speed and compensation level. However, it is hard for the wind farm to obtain a compensation level [...] Read more.
The DFIG-based wind farm faces sub-synchronous oscillation (SSO) when it is integrated with a series-compensated transmission system. The equivalent SSO damping is influenced by both wind speed and compensation level. However, it is hard for the wind farm to obtain a compensation level in time to predict the SSO risk. In this paper, an SSO risk prediction method for a DFIG wind farm is proposed based on the characteristics identified from noise-like signals. First, SSO-related parameters are analyzed. Then, the potential SSO frequency and damping are identified from signals at normal working points by integration using variational mode decomposition and Prony analysis. Finally, a fuzzy inference system is established to predict the SSO risk of a DFIG wind farm. The effectiveness of the proposed method is verified by simulation. The proposed prediction method can predict SSO risks caused by the variation in wind speed, while the transmission line parameters are undetectable for the wind farm. Full article
(This article belongs to the Special Issue Recent Advances in Smart Grid)
►▼ Show Figures

Figure 1

Figure 1
<p>Relationship between wind speed and wind turbine generator rotating speed.</p> Full article ">Figure 2
<p>Typical SSO noise-like signals that respond (<b>a</b>) to Case 1; (<b>b</b>) to Case 2.</p> Full article ">Figure 3
<p>The framework of the SSO prediction method.</p> Full article ">Figure 4
<p>The BPNN-based wind speed prediction model.</p> Full article ">Figure 5
<p>Structure of the FIS-based wind farm SSO risk prediction model.</p> Full article ">Figure 6
<p>Membership functions and fuzzy sets of (<b>a</b>) SSO damping ratio; (<b>b</b>) wind speed differences; (<b>c</b>) membership functions of SSO risk.</p> Full article ">Figure 7
<p>Fuzzy rule surface.</p> Full article ">Figure 8
<p>Prediction results of (<b>a</b>) wind speed; (<b>b</b>) wind speed trend.</p> Full article ">Figure 9
<p>Constructed test signals. (<b>a</b>) <span class="html-italic">S<sub>test</sub></span>; (<b>b</b>) <span class="html-italic">S<sub>SSO1</sub></span>; (<b>c</b>) <span class="html-italic">S<sub>SSO2</sub></span>; (<b>d</b>) <span class="html-italic">S<sub>noise</sub></span>.</p> Full article ">Figure 10
<p>SSO mode decomposition results. (<b>a</b>) <span class="html-italic">S<sub>SSO1</sub></span>; (<b>b</b>) <span class="html-italic">S<sub>SSO2</sub></span>.</p> Full article ">Figure 11
<p>De-noised SSO response signals (<b>a</b>) to Case 1; (<b>b</b>) to Case 2.</p> Full article ">Figure 12
<p>VMD-based mode decomposition results of the de-noised SSO response signal to Case 2.</p> Full article ">Figure 13
<p>Studied system.</p> Full article ">Figure 14
<p>The change in damping ratio according to wind speed.</p> Full article ">Figure 15
<p>The inferenced SSO risk under wind speed variations.</p> Full article ">Figure 16
<p>Wind speed and corresponding compensation level.</p> Full article ">Figure 17
<p>Predicted SSO risk by FIS under the change in compensation level.</p> Full article ">Figure 18
<p>The time-domain responses of SSO in wind farm at different time points. (<b>a</b>) At 1:45; (<b>b</b>) At 2:00; (<b>c</b>) At 4:15; (<b>d</b>) At 9:00; (<b>e</b>) At 11:00.</p> Full article ">
17 pages, 1820 KiB  
Article
MEDAL: A Multimodality-Based Effective Data Augmentation Framework for Illegal Website Identification
by Li Wen, Min Zhang, Chenyang Wang, Bingyang Guo, Huimin Ma, Pengfei Xue, Wanmeng Ding and Jinghua Zheng
Electronics 2024, 13(11), 2199; https://doi.org/10.3390/electronics13112199 - 5 Jun 2024
Viewed by 925
Abstract
The emergence of illegal (gambling, pornography, and attraction) websites seriously threatens the security of society. Due to the concealment of illegal websites, it is difficult to obtain labeled data with high quantity. Meanwhile, most illegal websites usually disguise themselves to avoid detection; for [...] Read more.
The emergence of illegal (gambling, pornography, and attraction) websites seriously threatens the security of society. Due to the concealment of illegal websites, it is difficult to obtain labeled data with high quantity. Meanwhile, most illegal websites usually disguise themselves to avoid detection; for example, some gambling websites may visually resemble gaming websites. However, existing methods ignore the means of camouflage in a single modality. To address the above problems, this paper proposes MEDAL, a multimodality-based effective data augmentation framework for illegal website identification. First, we established an illegal website identification framework based on tri-training that combines information from different modalities (including image, text, and HTML) while making full use of numerous unlabeled data. Then, we designed a multimodal mutual assistance module that is integrated with the tri-training framework to mitigate the introduction of error information resulting from an unbalanced single-modal classifier performance in the tri-training process. Finally, the experimental results on the self-developed dataset demonstrate the effectiveness of the proposed framework, performing well on accuracy, precision, recall, and F1 metrics. Full article
(This article belongs to the Section Artificial Intelligence)
►▼ Show Figures

Figure 1

Figure 1
<p>A web page screenshot of an illegal website and the extracted OCR text and HTML text. The Chinese sentence in OCR text means high commissions are settled daily, the Chinese sentence in OCR text means we promise to provide every customer with the safest, fairest, and most equitable gambling games. Classification from different modalities may yield inconsistent classification results.</p> Full article ">Figure 2
<p>Multimodality-based Effective Data Augmentation Framework for Illegal Website Identification.</p> Full article ">Figure 3
<p>Basic image classifier structure.</p> Full article ">Figure 4
<p>Basic text classifier structure.</p> Full article ">Figure 5
<p>The multimodal mutual assistance module attached to the framework.</p> Full article ">Figure 6
<p>Basic classifier selection.</p> Full article ">Figure 7
<p>Decreasing trends in accuracy under the misleading of the worst model during basic tri-training.</p> Full article ">Figure 8
<p>Final performance improvement results.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-230
Previous Issue
Next Issue
Back to TopTop