[go: up one dir, main page]
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
5.3
2.5
Journals
Applied Sciences
Special Issues
Intelligent Control and Robotics II
applsci-logo
Submit to Special Issue Submit Abstract to Special Issue Review for Applied Sciences Propose a Special Issue

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Intelligent Control and Robotics II

A special issue of Applied Sciences (ISSN 2076-3417). This special issue belongs to the section "Robotics and Automation".

Deadline for manuscript submissions: 20 February 2025 | Viewed by 6037

Special Issue Editors

Dr. Minsung Kim

E-Mail Website
Guest Editor
Department of Electronic and Electrical Engineering, Dongguk University-Seoul Campus, Seoul 04620, Republic of Korea
Interests: highly efficient power conversion circuit design; intelligent controller design for industrial electronics; renewable energy and energy storage systems
Special Issues, Collections and Topics in MDPI journals
Prof. Dr. Tae-Yong Kuc

E-Mail Website
Guest Editor
College of Information and Communication Engineering, Sungkyunkwan University, 300 Cheoncheon-dong Jangan-gu, Suwon 440-746, Gyeonggi-do, Republic of Korea
Interests: autonomous navigation of mobile robots; VSLAM; 3D SLAM; semantic SLAM
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

Intelligent control and robotics are closely related to each other and have been the subject of many years of research. With the growing need of robots that can perform useful tasks for humans, human-like learning and cognitive skills are required for the upcoming intelligent robots that share common surroundings with humans, such as homes, offices, factories, and outdoor environments. In view of this, intelligent control and robotics have been evolving in such a way that the topics accommodate and take advantage of a large spectrum of convergence technologies developed from algorithmic research, symbolic AI, and computational AI using rule-based knowledge modeling, neural networks, fuzzy logic, GAs, and more recently, deep neural networks. Furthermore, since intelligent robots should support target tasks involving high-level planning and control strategies for manipulation, navigation, and interaction (human–robot interaction), a recent trend in robot intelligence research is to combine traditional data-driven approaches with the knowledge-driven approaches motivated by cognitive science and brain research. This extends the coverage of topics for this Special Issue from motor-level learning and trajectory control to semantic SLAM (simultaneous localization and mapping) and scene understanding for intelligent control. We feel that the timing of this Special Issue is favorable, given the recent major achievements in related research, such as the finding of brain GPS function in neuroscience and physiology and the high performance of state-of-the-art deep-learning-based recognition.

We encourage researchers in this field to contribute their original papers to share their technical achievements with the readers of this Special Issue.

Prof. Dr. Minsung Kim
Prof. Dr. Tae-Yong Kuc
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • deep learning and neural approaches for robotics
  • adaptive learning control for robotics
  • intelligent control of autonomous robots in dynamic environments
  • automated and intelligent path planning of mobile robots
  • cooperative robots and distributed control
  • semantic SLAM
  • 3D SLAM
  • visual SLAM
  • place recognition and scene understanding
  • fault detection and diagnosis of self-recovery robots

Benefits of Publishing in a Special Issue

  • Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
  • Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
  • Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
  • External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
  • e-Book format: Special Issues with more than 10 articles can be published as dedicated e-books, ensuring wide and rapid dissemination.

Further information on MDPI's Special Issue polices can be found here.

Related Special Issue

Published Papers (5 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

17 pages, 48887 KiB  
Article
Inverse Kinematics of Large Hydraulic Manipulator Arm Based on ASWO Optimized BP Neural Network
by Yansong Lin, Qiaoyu Xu, Wenhao Ju and Tianle Zhang
Appl. Sci. 2024, 14(13), 5551; https://doi.org/10.3390/app14135551 - 26 Jun 2024
Cited by 1 | Viewed by 862
Abstract
In order to solve the problem of insufficient end positioning accuracy due to factors such as gravity and material strength during the inverse solution process of a large hydraulic robotic arm, this paper proposes an inverse solution algorithm based on an adaptive spider [...] Read more.
In order to solve the problem of insufficient end positioning accuracy due to factors such as gravity and material strength during the inverse solution process of a large hydraulic robotic arm, this paper proposes an inverse solution algorithm based on an adaptive spider wasp optimization (ASWO) optimized back propagation (BP) neural network. Firstly, the adaptability of the SWO algorithm is enhanced by analyzing the phase change in population fitness and dynamically adjusting the trade-off rate, crossover rate, and population size in real time. Then, the ASWO algorithm is used to optimize the initial weights and biases of the BP neural network, effectively addressing the problem of the BP neural network falling into local optima. Finally, a neural network mapping relationship between the actual position of the robotic arm’s end-effector and the corresponding joint values is established to reduce the influence of forward kinematic errors on the accuracy of the inverse solution. Experimental results show that the average positioning error of the robotic arm in the XYZ direction is reduced from (91.3, 87.38, 117.31) mm to (18.16, 24.67, 27.21) mm, significantly improving positioning accuracy by 80.11%, 71.78%, and 76.81%, meeting project requirements. Full article
(This article belongs to the Special Issue Intelligent Control and Robotics II)
►▼ Show Figures

Figure 1

Figure 1
<p>G3Zi rock drilling rig.</p> Full article ">Figure 2
<p>The DH model structure of the robotic arm.</p> Full article ">Figure 3
<p>The process of establishing the forward kinematics error model.</p> Full article ">Figure 4
<p>Robotic arm inverse kinematics error.</p> Full article ">Figure 5
<p>SWO algorithm judgement process.</p> Full article ">Figure 6
<p>Flowchart of the BP neural network algorithm optimized by ASWO.</p> Full article ">Figure 7
<p>Topology of the BP neural network.</p> Full article ">Figure 8
<p>Simulation verification process.</p> Full article ">Figure 9
<p>Optimal fitness curves of four algorithms.</p> Full article ">Figure 10
<p>Comparison of end positioning accuracy of four inverse solution methods for robotic arms.</p> Full article ">Figure 11
<p>Comparison of errors for the two positions as inputs to the data set.</p> Full article ">Figure 12
<p>Comparison of end position error between ASWO-BP method and Jacobi iterative method.</p> Full article ">
21 pages, 36012 KiB  
Article
DFD-SLAM: Visual SLAM with Deep Features in Dynamic Environment
by Wei Qian, Jiansheng Peng and Hongyu Zhang
Appl. Sci. 2024, 14(11), 4949; https://doi.org/10.3390/app14114949 - 6 Jun 2024
Viewed by 1091
Abstract
Visual SLAM technology is one of the important technologies for mobile robots. Existing feature-based visual SLAM techniques suffer from tracking and loop closure performance degradation in complex environments. We propose the DFD-SLAM system to ensure outstanding accuracy and robustness across diverse environments. Initially, [...] Read more.
Visual SLAM technology is one of the important technologies for mobile robots. Existing feature-based visual SLAM techniques suffer from tracking and loop closure performance degradation in complex environments. We propose the DFD-SLAM system to ensure outstanding accuracy and robustness across diverse environments. Initially, building on the ORB-SLAM3 system, we replace the original feature extraction component with the HFNet network and introduce a frame rotation estimation method. This method determines the rotation angles between consecutive frames to select superior local descriptors. Furthermore, we utilize CNN-extracted global descriptors to replace the bag-of-words approach. Subsequently, we develop a precise removal strategy, combining semantic information from YOLOv8 to accurately eliminate dynamic feature points. In the TUM-VI dataset, DFD-SLAM shows an improvement over ORB-SLAM3 of 29.24% in the corridor sequences, 40.07% in the magistrale sequences, 28.75% in the room sequences, and 35.26% in the slides sequences. In the TUM-RGBD dataset, DFD-SLAM demonstrates a 91.57% improvement over ORB-SLAM3 in highly dynamic scenarios. This demonstrates the effectiveness of our approach. Full article
(This article belongs to the Special Issue Intelligent Control and Robotics II)
►▼ Show Figures

Figure 1

Figure 1
<p>System architecture.</p> Full article ">Figure 2
<p>The complete process of precise elimination. (<b>a</b>) The segmentation results of YOLOv8. (<b>b</b>) The results of optical flow tracking on the extracted feature points. (<b>c</b>) The results of epipolar constraints. (<b>d</b>,<b>e</b>) The system dividing the detected potential dynamic objects into sub-frames and identifying the dynamic regions within them. In (<b>e</b>), the red boxes indicate dynamic regions, while the green areas indicate static regions. (<b>f</b>) The final retained segmentation results after dilation processing.</p> Full article ">Figure 3
<p>Filter out optical flow vectors that do not meet the requirements.</p> Full article ">Figure 4
<p>In a rotating scene, detect the matching situation before and after frame rotation. The first row uses HFNet descriptors. The second row is the frame identified as rotating, with red points indicating the optimized rotation center. The third row uses ORB descriptors instead.</p> Full article ">Figure 5
<p>Matching performance of DFD-SLAM and ORB-SLAM3 under varying lighting and scene conditions. The first row shows the matching performance of ORB-SLAM3 using its strategies. The second row illustrates the matching performance of DFD-SLAM using HFNet for feature point extraction and descriptor matching. In most cases, the deep-features-based extraction method still holds advantages.</p> Full article ">Figure 6
<p>Comparison of loop closure detection in monocular mode. The final trajectory maps are shown in (<b>h</b>,<b>i</b>). The numbers annotated above indicate the positions where loop closure detection occurred in each system. Scenes (<b>a</b>–<b>g</b>) correspond to the occurrences of loop closure detection, where the second row indicates the frames where the systems correctly detected loop closures relative to the first row.</p> Full article ">Figure 7
<p>Comparison of trajectories between outstanding dynamic SLAM systems and our method in highly dynamic environments. The first row shows the trajectory map for the <math display="inline"><semantics> <mrow> <mi mathvariant="italic">W</mi> <mo>/</mo> <mi mathvariant="italic">static</mi> </mrow> </semantics></math> sequence, the second row for the <math display="inline"><semantics> <mrow> <mi mathvariant="italic">W</mi> <mo>/</mo> <mi mathvariant="italic">xyz</mi> </mrow> </semantics></math> sequence, the third row for the <math display="inline"><semantics> <mrow> <mi mathvariant="italic">W</mi> <mo>/</mo> <mi mathvariant="italic">rpy</mi> </mrow> </semantics></math> sequence, and the fourth row for the W/half sequence. The blue lines represent the system’s result trajectory, the black lines indicate the ground truth, and the red lines show the difference between the two. More prominent and numerous red lines indicate a higher absolute trajectory error, signifying lower tracking accuracy of the system.</p> Full article ">Figure 8
<p>Dynamic point culling flowchart in <math display="inline"><semantics> <mrow> <mi>W</mi> <mo>/</mo> <mi>r</mi> <mi>p</mi> <mi>y</mi> </mrow> </semantics></math> sequence. Each of these lines represents a complete culling process. Each column represents a cull step.</p> Full article ">
16 pages, 1675 KiB  
Article
Mobile Robot Path Planning Algorithm Based on NSGA-II
by Sitong Liu, Qichuan Tian and Chaolin Tang
Appl. Sci. 2024, 14(10), 4305; https://doi.org/10.3390/app14104305 - 19 May 2024
Viewed by 895
Abstract
Path planning for mobile robots is a key technology in robotics. To address the issues of local optima trapping and non-smooth paths in mobile robot path planning, a novel algorithm based on the NSGA-II (Non-dominated Sorting Genetic Algorithm II) is proposed. The algorithm [...] Read more.
Path planning for mobile robots is a key technology in robotics. To address the issues of local optima trapping and non-smooth paths in mobile robot path planning, a novel algorithm based on the NSGA-II (Non-dominated Sorting Genetic Algorithm II) is proposed. The algorithm utilizes a search window approach for population initialization, which improves the quality of the initial population. An innovative fitness function is designed as the objective function for optimization iterations. A probability-based selection strategy is employed for population selection and optimization, enhancing the algorithm’s ability to escape local minima and preventing premature convergence to suboptimal solutions. Furthermore, a path smoothing algorithm is developed by incorporating Bézier curves. By connecting multiple segments of Bézier curves, the problem of the high computational complexity associated with high-degree Bézier curves is addressed, while simultaneously achieving smooth paths. Simulation results demonstrated that the proposed path planning algorithm exhibited fewer iterations, superior path quality, and path smoothness. Compared to other methods, the proposed approach demonstrated better overall performance and practical applicability. Full article
(This article belongs to the Special Issue Intelligent Control and Robotics II)
►▼ Show Figures

Figure 1

Figure 1
<p>Search window generation diagram.</p> Full article ">Figure 2
<p>Path angle.</p> Full article ">Figure 3
<p>The planned path coincides with the obstacle.</p> Full article ">Figure 4
<p>Path smoothing algorithm flow chart.</p> Full article ">Figure 5
<p>Grid map.</p> Full article ">Figure 6
<p>Two types of grid maps. (<b>a</b>) complex map; (<b>b</b>) simple map.</p> Full article ">Figure 7
<p>Planned path and convergence curve of NSGA-II. (<b>a</b>) Planned path of simple map; (<b>b</b>) convergence curve of simple map; (<b>c</b>) planned path of complex map; (<b>d</b>) convergence curve of complex map.</p> Full article ">Figure 7 Cont.
<p>Planned path and convergence curve of NSGA-II. (<b>a</b>) Planned path of simple map; (<b>b</b>) convergence curve of simple map; (<b>c</b>) planned path of complex map; (<b>d</b>) convergence curve of complex map.</p> Full article ">Figure 8
<p>Planned path and convergence curve of GA. (<b>a</b>) Planned path of simple map; (<b>b</b>) convergence curve of simple map; (<b>c</b>) planned path of complex map; (<b>d</b>) convergence curve of complex map.</p> Full article ">Figure 9
<p>Planned path and convergence curve in this paper. (<b>a</b>) Planned path of simple map; (<b>b</b>) convergence curve of simple map; (<b>c</b>) planned path of complex map; (<b>d</b>) convergence curve of complex map.</p> Full article ">Figure 10
<p>Path of smoothing algorithm. (<b>a</b>) Smoothed path of simple map; (<b>b</b>) smoothed path of complex map.</p> Full article ">
17 pages, 1212 KiB  
Article
A Novel Zeroing Neural Network Control Scheme for Tracked Mobile Robot Based on an Extended State Observer
by Yuxuan Cao and Jinyun Pu
Appl. Sci. 2024, 14(1), 303; https://doi.org/10.3390/app14010303 - 29 Dec 2023
Viewed by 849
Abstract
A novel zeroing neural network control scheme based on an extended state observer is proposed for the trajectory tracking of a tracked mobile robot which is subject to unknown external disturbances and uncertainties. To estimate unknown lumped disturbances and unmeasured velocities, a third-order [...] Read more.
A novel zeroing neural network control scheme based on an extended state observer is proposed for the trajectory tracking of a tracked mobile robot which is subject to unknown external disturbances and uncertainties. To estimate unknown lumped disturbances and unmeasured velocities, a third-order fixed-time extended state observer is proposed, and the observation errors converge to zero in fixed time. Based on the estimated values, the zeroing neural network controller is designed for a tracked mobile robot to track an eight shape. The stability of the system is analyzed based on Lyapunov theory. Simulation results are illustrated to show the effectiveness of the proposed control scheme. Full article
(This article belongs to the Special Issue Intelligent Control and Robotics II)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic diagram of TMR motion.</p> Full article ">Figure 2
<p>Schematic of the fixed-time control system for a TMR.</p> Full article ">Figure 3
<p>Comparison results of the FTESO, the FTESO proposed in [<a href="#B26-applsci-14-00303" class="html-bibr">26</a>], and LESO proposed in [<a href="#B41-applsci-14-00303" class="html-bibr">41</a>].</p> Full article ">Figure 4
<p>Tracking performance of the proposed model and control inputs in noise-free environment.</p> Full article ">Figure 5
<p>Evolution of TMR’s position (<span class="html-italic">x</span>,<span class="html-italic">y</span>,<math display="inline"><semantics> <mi>θ</mi> </semantics></math>).</p> Full article ">Figure 6
<p>Evolution of TMR linear velocity <span class="html-italic">v</span> and angular velocity <math display="inline"><semantics> <mi>ω</mi> </semantics></math> in noise-free environment.</p> Full article ">Figure 7
<p>The lumped disturbances and their observations.</p> Full article ">Figure 8
<p>Tracking performance of the proposed model and tracking error considering noise.</p> Full article ">Figure 9
<p>Evolution of TMR linear velocity <span class="html-italic">v</span> and angular velocity <math display="inline"><semantics> <mi>ω</mi> </semantics></math> considering noise.</p> Full article ">Figure 10
<p>Tracking control signals <math display="inline"><semantics> <msub> <mi>τ</mi> <mi>v</mi> </msub> </semantics></math> and <math display="inline"><semantics> <msub> <mi>τ</mi> <mi>ω</mi> </msub> </semantics></math> considering noise.</p> Full article ">
17 pages, 16909 KiB  
Article
Adaptive Locomotion Learning for Quadruped Robots by Combining DRL with a Cosine Oscillator Based Rhythm Controller
by Xiaoping Zhang, Yitong Wu, Huijiang Wang, Fumiya Iida and Li Wang
Appl. Sci. 2023, 13(19), 11045; https://doi.org/10.3390/app131911045 - 7 Oct 2023
Cited by 1 | Viewed by 1840
Abstract
Animals have evolved to adapt to complex and uncertain environments, acquiring locomotion skills for diverse surroundings. To endow a robot’s animal-like locomotion ability, in this paper, we propose a learning algorithm for quadruped robots based on deep reinforcement learning (DRL) and a rhythm [...] Read more.
Animals have evolved to adapt to complex and uncertain environments, acquiring locomotion skills for diverse surroundings. To endow a robot’s animal-like locomotion ability, in this paper, we propose a learning algorithm for quadruped robots based on deep reinforcement learning (DRL) and a rhythm controller that is based on a cosine oscillator. For a quadruped robot, two cosine oscillators are utilized at the hip joint and the knee joint of one leg, respectively, and, finally, eight oscillators form the controller to realize the quadruped robot’s locomotion rhythm during moving. The coupling between the cosine oscillators of the rhythm controller is realized by the phase difference, which is simpler and easier to realize when dealing with the complex coupling relationship between different joints. DRL is used to help learn the controller parameters and, in the reward function design, we address the challenge of terrain adaptation without relying on the complex camera-based vision processing but based on the proprioceptive information, where a state estimator is introduced to achieve the robot’s posture and help finally utilize the food-end coordinate. Experiments are carried out in CoppeliaSim, and all of the flat, uphill and downhill conditions are considered. The results show that the robot can successfully accomplish all the above skills and, at the same time, with the reward function designed, the robot’s pitch angle, yaw angle and roll angle are very small, which means that the robot is relatively stable during walking. Then, the robot is transplanted to a new scene; the results show that although the environment is previously unencountered, the robot can still fulfill the task, which demonstrates the effectiveness and robustness of this proposed method. Full article
(This article belongs to the Special Issue Intelligent Control and Robotics II)
►▼ Show Figures

Figure 1

Figure 1
<p>The quadruped robot, where LF represents the left-front leg, RF represents the right-front leg, LB represents the left-back leg, and RB represents the right-back leg. The direction of the red arrow is straight ahead.</p> Full article ">Figure 2
<p>Trot gait pattern. The black squares indicate that the robot’s legs are in the swing phase, and the white squares indicate that the robot’s legs are in the support phase.</p> Full article ">Figure 3
<p>Quadruped robot adaptive locomotion learning algorithm architecture.The blue line output by the control unit represents the hip signal, and the red line represents the knee signal.</p> Full article ">Figure 4
<p>Output of the single-legged control unit.</p> Full article ">Figure 5
<p>Four-legged rhythm controller structure, where the hip joints are represented by h, the knee joints are represented by k, B represents the body, and the dashed lines represent the coupling between two joints, represented by <math display="inline"><semantics> <msub> <mi>φ</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> </semantics></math>.</p> Full article ">Figure 6
<p>The robot is on the slope. The red and blue dots indicate the position of the foot endpoints.</p> Full article ">Figure 7
<p>Reward.</p> Full article ">Figure 8
<p>Curves of the robot’s center of gravity position during walking on flat ground.</p> Full article ">Figure 9
<p>Attitude curves of the robot during walking.</p> Full article ">Figure 10
<p>Slope of 5°. The yellow slope has three sections, which are 5° uphill, 0° slope and 5° downhill, and each section is 3m as shown in the black line segment.</p> Full article ">Figure 11
<p>Robot uphill and downhill processes.</p> Full article ">Figure 12
<p>The robot’s center of gravity position during movement in the scene in <a href="#applsci-13-11045-f010" class="html-fig">Figure 10</a>.</p> Full article ">Figure 13
<p>The orientation of the robot during movement in the scene in <a href="#applsci-13-11045-f010" class="html-fig">Figure 10</a>.</p> Full article ">Figure 14
<p>The output of the cosine rhythm controller during movement in the scene in <a href="#applsci-13-11045-f010" class="html-fig">Figure 10</a>.</p> Full article ">Figure 15
<p>The center of gravity curve of the robot when using fixed cosine rhythm controller parameters.</p> Full article ">Figure 16
<p>The slope called S-4, which has 4 sections. The first section is 5° uphill slope which is 1.4 m, as shown by the black line segment with arrows, the second section is a 0° slope which is 1.4 m, the third section is a 9° uphill slope which is 1.2 m, and the fourth section is a 0° slope which is 1.4 m.</p> Full article ">Figure 17
<p>The robot’s center of gravity position when climbing S-4.</p> Full article ">Figure 18
<p>The orientation of the robot when climbing S-4.</p> Full article ">Figure 19
<p>The output of the cosine rhythm controller when the robot climbed S-4.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-5
Back to TopTop