Search Results (24)

Aiming at the problems of jittering waveforms and poor power quality caused by external disturbances during the operation of a dual-pulse-width-modulation (PWM) converter, an improved terminal sliding mode control and an improved active disturbance rejection control (ADRC) are investigated. The method is based on mathematical models of grid-side and machine-side converters to design the controllers separately, and the balance between the two sides is maintained by the capacitor voltage. An improved terminal fuzzy sliding mode control and equivalent input disturbance (EID)-error-estimation-based active disturbance rejection control are presented on the grid side to improve the voltage response rate, and an improved support vector modulation (SVM)–direct torque control (DTC)–ADRC method is developed on the motor side to improve the robustness against disturbances. Finally, theoretical simulation experiments are built in MATLAB R2023a/Simulink to verify the effectiveness and superiority of this method. Full article

The present paper proposes the use of fractional derivatives in the definition of sliding function, giving a new mode control applied to induction motor drives in electric vehicle (EV) applications. The proposed Fractional-Order Sliding Mode Direct Torque Control-Space Vector Modulation (FOSM-DTC-SVM) strategy aims to address the limitations of conventional control techniques and mitigate torque and flux ripples in induction motor systems. The paper first introduces the motivation for using fractional-order control methods to handle the nonlinear and fractional characteristics inherent in induction motor systems. The core describes the proposed FOSM-DTC-SVM control strategy, which leverages a fractional sliding function and the associated Lyapunov stability analysis. The efficiency of the proposed strategy is validated via three scenarios. (i) The first scenario, where the acceleration of the desired speed is defined by pulses, leading to Dirac impulses in its second derivative, demonstrates the advantage of the proposed control approach in tracking the desired speed while minimizing flux ripples and generating pulses in the rotor pulsation. (ii) The second scenario demonstrates the effectiveness of filtering the desired speed to eliminate Dirac impulses, resulting in smoother rotor pulsation variations and a slightly slower speed response while maintaining similar flux ripples and stator current characteristics. (iii) The third scenario consists of eliminating the fractional derivatives of the pulses existing in the expression of the control, leading to the elimination of Dirac impulses. These results demonstrate the potential of the FOSM-DTC-SVM to revolutionize the performance and efficiency of EVs. By incorporating fractional control in the control scheme for PV-powered EVs, the paper showcases a promising avenue for sustainable transportation. Full article

This paper evaluates the potential of using artificial intelligence (AI) and machine learning (ML) approaches for classification of Landsat satellite imagery for environmental coastal mapping. The aim is to identify changes in patterns of land cover types in a coastal area around Cheetham Wetlands, Port Phillip Bay, Australia. The scripting approach of the Geographic Resources Analysis Support System (GRASS) geographic information system (GIS) uses AI-based methods of image analysis to accurately discriminate land cover types. Four ML algorithms are applied, tested and compared for supervised classification. Technical approaches are based on using the ‘r.learn.train’ module, which employs the scikit-learn library of Python. The methodology includes the following algorithms: (1) random forest (RF), (2) support vector machine (SVM), (3) an ANN-based approach using a multi-layer perceptron (MLP) classifier, and (4) a decision tree classifier (DTC). The tested methods using AI demonstrated robust results for image classification, with the highest overall accuracy exceeding 98% and reached by the SVM and RF models. The presented scripting approach for GRASS GIS accurately detected changes in land cover types in southern Victoria over the period of 2013–2024. From our findings, the use of AI and ML algorithms offers effective solutions for coastal monitoring by analysis of change detection using multi-temporal RS data. The demonstrated methods have potential applications in coastal and wetland monitoring, environmental analysis and urban planning based on Earth observation data. Full article

Energy efficiency is a critical aspect of engineering due to the associated monetary and environmental benefits it can bring. One aspect in particular, namely, the prediction of heating and cooling loads, plays a significant role in reducing energy use costs and in minimising the risks associated with climate change. Recently, data-driven approaches, such as artificial intelligence (AI) and machine learning (ML), have provided cost-effective and high-quality solutions for the prediction of heating and cooling loads. However, few studies have focused on interpretable classifiers that can generate not only reliable predictive systems but are also easy to understand for the stakeholders. This research investigates the applicability of ML techniques (classification) in the prediction of the heating and cooling loads of residential buildings using a dataset consisting of various variables such as roof area, building height, orientation, surface area, wall area, and glassing area distribution. Specifically, we sought to determine whether models that derive rules are competitive in terms of performance when compared with other classification techniques for assessing the energy efficiency of buildings, in particular the associated heating and cooling loads. To achieve this aim, several ML techniques including k-nearest neighbor (kNN), Decision Tree (DT)-C4.5, naive Bayes (NB), Neural Network (Nnet), Support Vector Machine (SVM), and Rule Induction (RI)- Repeated Incremental Pruning to Produce Error (RIPPER) were modelled and then evaluated based on residential data using a range of model evaluation parameters such as recall, precision, and accuracy. The results show that most classification techniques generate models with good predictive power with respect to the heating or cooling loads, with better results achieved with interpretable classifiers such as Rule Induction (RI), and Decision Trees (DT). Full article

Breast cancer results from a disruption of certain cells in breast tissue that undergo uncontrolled growth and cell division. These cells most often accumulate and form a lump called a tumor, which may be benign (non-cancerous) or malignant (cancerous). Malignant tumors can spread quickly throughout the body, forming tumors in other areas, which is called metastasis. Standard screening techniques are insufficient in the case of metastasis; therefore, new and advanced techniques based on artificial intelligence (AI), machine learning, and regression models have been introduced, the primary aim of which is to automatically diagnose breast cancer through the use of advanced techniques, classifiers, and real images. Real fine-needle aspiration (FNA) images were collected from Wisconsin, and four classifiers were used, including three machine learning models and one regression model: the support vector machine (SVM), naive Bayes (NB), k-nearest neighbors (k-NN), and decision tree (DT)-C4.5. According to the accuracy, sensitivity, and specificity results, the SVM algorithm had the best performance; it was the most powerful computational classifier with a 97.13% accuracy and 97.5% specificity. It also had around a 96% sensitivity for the diagnosis of breast cancer, unlike the models used for comparison, thereby providing an exact diagnosis on the one hand and a clear classification between benign and malignant tumors on the other hand. As a future research prospect, more algorithms and combinations of features can be considered for the precise, rapid, and effective classification and diagnosis of breast cancer images for imperative decisions. Full article

Accurate classification of forest stand is crucial for protection and management needs. However, forest stand classification remains a great challenge because of the high spectral and textural similarity of different tree species. Although existing studies have used multiple remote sensing data for forest identification, the effects of different spatial resolutions and combining multi-source remote sensing data for automatic complex forest stand identification using deep learning methods still require further exploration. Therefore, this study proposed an object-oriented convolutional neural network (OCNN) classification method, leveraging data from Sentinel-2, RapidEye, and LiDAR to explore classification accuracy of using OCNN to identify complex forest stands. The two red edge bands of Sentinel-2 were fused with RapidEye, and canopy height information provided by LiDAR point cloud was added. The results showed that increasing the red edge bands and canopy height information were effective in improving forest stand classification accuracy, and OCNN performed better in feature extraction than traditional object-oriented classification methods, including SVM, DTC, MLC, and KNN. The evaluation indicators show that ResNet_18 convolutional neural network model in the OCNN performed the best, with a forest stand classification accuracy of up to 85.68%. Full article

In this study, the analysis and control of a multi-phase linear induction motor loaded with a variable mechanical system are carried out. Mathematical models are established, and simulation results are analyzed for an improved proportional–integral–derivative controller with closed-loop vector control for PLIM. To make the PID controller more responsive to load thrust disturbances, a fuzzy PID load thrust observer was developed. The FPID is similarly based on space-vector modulation DTC technology to regulate the PLIM’s speed, flux, and thrust. The FPID output is used to calculate the reference thrust force, which is compared to the actual thrust value to calculate the second error. To maintain the linear speed of the PLIM at the specified reference values and at different load values, the FPID controller settings are adjusted. Four indicators were used to compare the capabilities of the FPID controller with those of the conventional PID controller in order to evaluate the performance of PLIM in both cases. These indices represent the individual SSE for each operational phase and the total SSE for the entire loading period. According to the simulation results, the FPID works better than a regular PID when used to adjust the operation of DTC-SVM to drive a PLIM to improve the overall system performance. The simulation results using MATLAB Simulink for a PLIM-drive system show that the proposed FPID control provides improved control behavior and operating performance with fast and accurate speed tracking. Full article

This article presents the DTC-SVM approach for controlling a sensorless speed induction motor. To implement this approach, a practical prototype is built using a microcontroller, an embedded GPS module, and a memory card to collect real-time data during the driving route, such as road geographical data, speed, and time. These data are then utilized in the laboratory to implement the control law (DTC-SVM) on the electric vehicle. The d-q model of the induction motor is first presented to explain the requirements for calculating the rotor speed. Then, an adaptive model reference system speed estimator is developed based on the rotor flux, along with a controller and DTC-SVM strategy, which are implemented using the dSpace 1104 board to achieve the desired performance. The simulation results demonstrate satisfactory speed regulation with the proposed system. In this study too, an electronic differential system is modeled for the four wheels of an electric vehicle equipped with an integrated motor, all controlled by the DTC-SVM strategy. Vehicle speed and electrical vehicle steering angle variations, as well as wheel speeds estimated by code system, are verified using MATLAB/Simulink simulations. Full article

The development of hybrid electric vehicles (HEVs) is rapidly gaining traction as a viable solution for reducing carbon emissions and improving fuel efficiency. One type of HEV that is gaining significant interest is the fuel cell/battery/supercapacitor HEV (FC/Bat/SC HEV), which combines fuel cell, battery, supercapacitor, AC, and DC generators. These FC/B/SC HEVs are particularly appealing because they excel at efficiently managing energy and cater to a wide range of driving requirements. This study presents a novel approach for exploiting the kinetic energy of a sensorless HEV. The vehicle has a primary fuel cell resource, a supercapacitor, and lithium-ion battery energy storage banks, where each source is connected to a special converter. The obtained hybrid system allows the vehicle to enhance autonomy, support the fuel cell during low production moments, and improve transient and steady-state load requirements. The exploitation of kinetic energy is performed by the DC and AC generators that are linked to the electric vehicle front wheels to transfer the HEV’s wheel rotation into power, contributing to the overall power balance of the vehicle. The energy management system for electric vehicles determines the FC setpoint power through the classical state machine method. At the same time, a robust speed controller-based artificial intelligence algorithm reduces power losses and enhances the supply efficiency for the vehicle. Furthermore, we evaluate the performance of a robust controller with a speed estimator, specifically using the adaptive neuro-fuzzy inference system (ANFIS) and the model reference adaptive system (MRAS) estimator in conjunction with the direct torque control-support vector machine (DTC-SVM), to enhance the torque and speed performance of HEVs. The results demonstrate the feasibility and reliability of the vehicle while utilizing the additional DC and AC generators to extract free kinetic energy, both of which contributed to 28% and 24% of the total power for the vehicle, respectively. This approach leads to a vehicle supply efficiency exceeding 96%, reducing the burden on fuel cells and batteries and resulting in a significant reduction in fuel consumption, which is estimated to range from 25% to 35%. Full article

One innovative composite fault-tolerant control tactic is presented for the reliable operation of a power transmission system, which consists of both an asymmetric six-phase permanent magnet synchronous motor (PMSM) and a T-type mid-point clamp type (T-NPC) three-level inverter. First, in order to inherit the better harmonic property of simplified space vector modulation (SVM) and the rapid dynamic capability of direct torque control (DTC), the SVM-DTC control scheme was determined, and the harmonic electric current suppression unit was added to the basic control scheme to obtain good harmonic electric current suppression. In addition, a strategy for open-circuit fault-tolerant control under the SVM-DTC scheme was designed by analyzing the mutual influence between the stator flux linkage and the stator voltage of each phase under an open-circuit fault. Finally, the PMSM drive system principle prototype was tested. By comparing the waveforms of output torque and current of each phase before and after fault tolerance, it shows that the large torque fluctuation (±5%) before fault tolerance was suppressed to ±2% and smoothed out, verifying the effectiveness of fault tolerance control. Full article

Two algorithms have been extensively studied for motor control: Field Oriented Control (FOC) and Direct Torque Control (DTC). Both control algorithms use a Voltage Source Inverter (VSI) to drive a Permanent Magnet Synchronous Motor (PMSM). To prevent short-arm short-circuit accidents when driving PMSM using VSI, a dead time is used to turn off the TOP and BOTTOM switches of each arm at the same time. However, this dead-time technique causes an unexpected pole voltage to be applied to the PMSM on the VSI output voltage, causing distortion and resulting in control nonlinearity. The disturbance voltage that causes nonlinearity is difficult to measure directly with the sensor. Therefore, this paper analyzes the nonlinearity of the controller due to the distorted voltage caused by the dead time during PMSM operation using the DTC algorithm and predicts the distorted output voltage using the extended Kalman Filter (EKF) to improve control stability. As a result, The algorithm proposed in this paper has verified the improvement of torque ripple and stator flux ripple through experiments and simulations. Full article

The doubly fed induction machine (DFIM)-based DC voltage generator is equipped with a stator-connected diode rectifier. The six-pulse diode rectifier as a nonlinear circuit introduces harmonics in the stator and rotor current and distorts the machine stator voltage, as well as the stator flux. This causes electromagnetic torque oscillations and instantaneous power components oscillations. The torque oscillations adversely impact the mechanical parts of the drive-train and oscillations of the p component of instantaneous power influence DC-bus voltage oscillations. The oscillations can be somewhat cancelled by control methods. However, cancellation of electromagnetic torque is not strictly coupled with cancellation of oscillations of the p component of instantaneous power. The paper presents an analysis of influence of the control methods aimed at a reduction of torque oscillations on the output voltage oscillations level in the stand-alone DFIM-based DC voltage generator. Field-oriented control FOC with current controllers and space vector modulation-based direct torque control DTC-SVM with flux module regulation have been compared with control in which electromagnetic torque is one of the commanded variables, whereas the second variable is the dot product of stator flux and rotor current space vectors. The contributions of this paper are the introduction of a new variable in the second control path in the DTC-SVM method instead of flux vector length and the proof that it can reduce torque and DC-bus voltage oscillations in the DFIG-DC system. Additionally, this paper reveals that for proper stator-to-rotor-turns ratio of a doubly fed machine necessary for reduction of the rotor converter power, lower DC-bus voltage can be obtained than is required for full realization rotor side voltage requested by rotor current controllers. This is the reason why, regardless of the control method, torque oscillations cannot be always fully cancelled, and a comparative study of the methods at these conditions has been conducted in simulation and in laboratory tests. Full article

A number of cross-sectional epidemiological studies suggest that poor oral health is associated with respiratory diseases. However, the number of cases within the studies was limited, and the studies had different measurement conditions. By analyzing data from the National Health and Nutrition Examination Survey III (NHANES III), this study aimed to investigate possible associations between chronic obstructive pulmonary disease (COPD) and periodontitis in the general population. COPD was diagnosed in cases where FEV (1)/FVC ratio was below 70% (non-COPD versus COPD; binary classification task). We used unsupervised learning utilizing k-means clustering to identify clusters in the data. COPD classes were predicted with logistic regression, a random forest classifier, a stochastic gradient descent (SGD) classifier, k-nearest neighbors, a decision tree classifier, Gaussian naive Bayes (GaussianNB), support vector machines (SVM), a custom-made convolutional neural network (CNN), a multilayer perceptron artificial neural network (MLP), and a radial basis function neural network (RBNN) in Python. We calculated the accuracy of the prediction and the area under the curve (AUC). The most important predictors were determined using feature importance analysis. Results: Overall, 15,868 participants and 19 feature variables were included. Based on k-means clustering, the data were separated into two clusters that identified two risk characteristic groups of patients. The algorithms reached AUCs between 0.608 (DTC) and 0.953% (CNN) for the classification of COPD classes. Feature importance analysis of deep learning algorithms indicated that age and mean attachment loss were the most important features in predicting COPD. Conclusions: Data analysis of a large population showed that machine learning and deep learning algorithms could predict COPD cases based on demographics and oral health feature variables. This study indicates that periodontitis might be an important predictor of COPD. Further prospective studies examining the association between periodontitis and COPD are warranted to validate the present results. Full article

In this paper, a full analysis of voltage vectors (VVs) in the DTC algorithm is presented. The analytical analysis shows that the application of specific VVs results in false switching states called uncontrollable angles (UCAs). A robust scheme that ensures the elimination of UCAs is proposed for medium and high speeds with (18) subsectors (SSs). Simulation results are obtained and validated using MATLAB/Simulink. Full article

Modern precision agriculture applications increasingly rely on stable computer vision outputs. An important computer vision task is to discriminate between soil and plant pixels, which is called plant segmentation. For this task, supervised learning techniques, such as decision tree classifiers (DTC), support vector machines (SVM), or artificial neural networks (ANN) are increasing in popularity. The selection of training data is of utmost importance in these approaches as it influences the quality of the resulting models. We investigated the influence of three modeling parameters, namely proportion of plant pixels (plant cover), criteria on what pixel to choose (pixel selection), and number/type of features (input features) on the segmentation quality using DTCs. Our findings show that plant cover and, to a minor degree, input features have a significant impact on segmentation quality. We can state that the overperformance of multi-feature input decision tree classifiers over threshold-based color index methods can be explained to a high degree by the more balanced training data. Single-feature input decision tree classifiers can compete with state-of-the-art models when the same training data are provided. This study is the first step in a systematic analysis of influence parameters of such plant segmentation models. Full article