Sezgin Kaçar

This study investigates and compares Sliding Mode Observer (SMO)-based rotor position and speed estimation methods used in sensorless drive applications of permanent magnet synchronous motors (PMSMs). The estimation performance of these methods in different speed domains is evaluated. Comparative analyses were performed in MATLAB/Simulink, and simulation results were obtained under sensorless vector control of PMSMs. The results show that improved and adaptive SMO methods provide better low-speed performance compared to traditional approaches. Thus, a comprehensive performance evaluation is presented, highlighting the advantages and limitations of different SMO-based methods for sensorless control applications in PMSM drives.

This study presents a MATLAB/Simulink-based analysis to investigate the effect of differential gear ratios on the energy consumption of a 13-meter low-entry electric city bus. Four differential ratios (5.12, 5.73, 6.19, and 7.36) were simulated under SORT-1, SORT-2, and SORT-3 driving cycles, representing different urban operation patterns. The developed simulation model incorporates vehicle longitudinal dynamics, motor efficiency maps, and battery energy flow. Results indicate that increasing the differential ratio reduces the specific energy consumption in all driving cycles. The lowest value of 1.747 kWh/km was obtained at a 7.36 ratio under the SORT-3 cycle. The recovery ratio remained stable around 27 ± 1 %, confirming consistent regenerative braking performance. The findings highlight the importance of differential ratio selection for optimizing the drivetrain efficiency of electric buses and can serve as a guideline for future powertrain configurations in urban public transport applications.

This study presents a sliding mode controller design for DC motor speed control using optimization algorithms. The design of sliding mode controllers typically requires expert input during the parameter determination phase. Traditionally, these parameters are set through trial-and-error methods based on the experience of specialists. However, this approach can be both time-consuming and costly. The application of optimization methods automates the parameter-tuning process, reducing human intervention and, in turn, minimizing both design time and costs. The goal of this study is to enhance the performance of optimization methods by hybridizing them with chaotic systems. The random structures of chaotic systems allow optimization algorithms to explore a broader solution space, thereby improving their performance. The analyses conducted in this study reveal that hybrid chaotic algorithms outperform their original ones. The data indicate that the use of hybrid algorithms generally leads to a decrease in Steady-State Error. Additionally, it is observed that when all hybrid algorithms are employed, the sliding mode controller does not exhibit any overshoot. The results demonstrate that the sliding mode controller performs effectively, achieving low settling time, rise time, and steady-state error, while also preventing chattering. Among the methods examined, the sliding mode controller optimized with the Chaotic Henry Gas Solubility Optimization algorithm delivers the best performance, ensuring optimal system stability.

Objective: The aim of this study is to evaluate the ability to detect emotions from human facial expressions via facial recognition technologies and analyze the effectiveness of deep learning models in this process. Method: This research was conducted between 01.04 and 01.07.2024. The data of the study were taken from the open access site https://www.kaggle.com/datasets/msambare/fer2013 (Kaggle, 2024). Python 3.8 is used in this study. The FER-2013 (Facial Expression Recognition 2013) dataset is a comprehensive collection of facial images labeled with various emotions. The dataset contains 35,887 grayscale facial images. Each image has a size of 48 × 48 pixels. The dataset consists of images belonging to 7 emotion categories: anger, disgust, fear, happy, sad, confused, and neutral. Results: In our experiments on the FER2013 dataset, we evaluated the performance of three different models: MobileNetV3-L, EfficientNetV2-L, and our proposed EfficientMobileNet. The evaluation criteria were based on sensitivity, specificity, accuracy, and F1 scores to assess the effectiveness of each model comprehensively. The EfficientMobileNet model outperformed MobileNetV3-L and EfficientNetV2-L in all measured performance metrics. EfficientMobileNet was the most successful model for predicting emotions, with an accuracy of 77.6%. Conclusion: The impressive results obtained by EfficientMobileNet on the Fer2013 dataset show potential for wider application, especially in image classification tasks involving low-quality or small-scale images. This performance supports the idea of the potential for further improvements in neural network architecture and model efficiency and accuracy. Future work should focus on optimizing the model for more challenging datasets, studying the impact of different architectural adjustments, and investigating the scalability of EfficientMobileNet across various domains and applications.

This study presents a novel meta-heuristic optimization method that combines the Henry Gas Solubility Optimization (HGSO) technique with symmetric chaotic systems. By leveraging the randomness of chaotic systems, the parameters of the HGSO algorithm that require random generation are produced through chaotic processes, allowing the algorithm to exhibit chaotic behavior in its pursuit of optimal values. This innovative approach is termed Chaotic Henry Gas Solubility Optimization (CHGSO), with the primary objective of enhancing the performance of the HGSO method. The randomness of the data obtained from chaotic systems was validated using NIST-800-22 tests. The CHGSO method was applied to both 47 benchmark functions and the optimization of parameters for a PID controller utilized in the speed control of a DC motor. To evaluate the effectiveness of the proposed method, it was compared with several widely recognized algorithms in the literature, including PSO, WOA, GWO, EA, SA, and the original HGSO algorithm. The results demonstrate that the proposed method achieved the best performance in 43 of the benchmark functions, outperforming the other algorithms. In the context of controller design, the PID parameters were optimized using the error-based ITSE objective function. According to the controller responses, the proposed method has achieved the best results in the simulation studies, with a settling time of 0.035 and a rise time of 0.014 without overshooting, and in the experimental studies, with a settling time of 0.15 and a settling time of 1.4%. When the results are examined, it is observed that it has achieved successful results in the controller design problem.

Object detection in high resolution enables the identification and localization of objects for monitoring critical areas with precision. Although there have been improvements in object detection at high resolution, the variety of object scales, as well as the diversity of backgrounds and textures in high-resolution images, make it challenging for detectors to generalize successfully. This study introduces a new method for object detection in high-resolution images. The pre-processing stage of the method includes ISA and SAM to slice the input image and segment the objects in bounding boxes, respectively. In order to improve the resolution in the slices, the first layer of YOLO is designed as SRGAN. Thus, before applying YOLO detection, the resolution of the sliced images is increased to improve features. The proposed system is evaluated on xView and VisDrone datasets for object detection algorithms in satellite and aerial imagery contexts. The success of the algorithm is presented in four different YOLO architectures integrated with SRGAN. According to comparative evaluations, the proposed system with Yolov5 and Yolov8 produces the best results on xView and VisDrone datasets, respectively. Based on the comparisons with the literature, our proposed system produces better results.