Alagappa Chettiar College of Engineering and Technology

State of Charge (SOC) estimation plays a crucial role in managing the performance and longevity of electric vehicle (EV) batteries. Various estimation techniques, including various Kalman Filters (KF), have been employed for this purpose. The mathematical process and limits of different KF family algorithms are analysed. In this study, an impact of nonlinear KF parameters on SOC estimation accuracy for EV batteries is analyzed. Specifically, the performance of Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), and Particle Filter (PF) is examined. EKF is most preferred for SOC estimation whereas introducing sigma points in UKF is the basic difference between EKF and UKF. In comparison to EKF and UKF, PF is more flexible since it applies to nonlinear systems with non-Gaussian distributions. Utilising a sequence of noisy measurements, KF is a recursive method that approximates the state of a dynamic system. It requires accurate parameters such as process noise covariance (Q) and measurement noise covariance (R) to work effectively. In this work, the effect of these parameters on state of charge estimation using nonlinear KFs has been analysed under different settings of noise covariance matrices. Tuning these parameters is essential for optimizing the performance of the KF and ensuring accurate and reliable state estimation in real-world applications. To find parameter values that provide the best balance between accuracy, robustness, and computational efficiency, the tuning process involves experimentation, simulation, and validation using real-world data. According to the results observed, the filter parameters such as three diagonal elements of Q are selected as 2e-8, 0.005, 0.003 and R as 2e-6. By evaluating the state estimation accuracy under different driving scenarios, robustness and sensitivity of filtering technique in estimating SOC for EV batteries has been studied. This analysis contributes to the optimization of SOC estimation algorithms, enhancing the reliability and efficiency of EV battery management systems.

This study presents a time-coupled, multi-objective distributionally robust chance-constrained (MODRCC) framework for resilient grid restoration using Mobile Emergency Generators (MEGs). The model unifies (i) time-expanded logistics for MEG routing, crew scheduling, and refuelling, (ii) islanding-feasible DC-OPF under line outages, and (iii) Wasserstein-ball ambiguity to hedge uncertainty in attack severity and travel-time delays. Disjunctive linearization and second-order-cone (SOC) embeddings yield a tractable MISOCP that is evaluated inside an NSGA-II evolutionary search to generate the Pareto frontier between total cost and resilience. Experiments on IEEE-24 and IEEE-118 (12-hour horizon, 24 periods) show that, at comparable budgets, the proposed method reduces expected unserved energy (EUE) by 14–20

Ti-6Al-4 V (grade 5 titanium alloy) is widely applied in critical structures owing to its high strength-to-weight ratio and corrosion resistance; however, its strong reactivity at elevated temperatures complicates fusion welding. Laser beam welding (LBW) was employed in this study using a Nd:YAG laser to produce butt joints. Experimental weld-bead geometry was measured and compared with three-dimensional transient thermal simulations in COMSOL Multiphysics using a conical Gaussian heat source with temperature-dependent material properties. The predicted weld profiles closely matched experimental results, confirming the reliability of the finite element model. Microstructural evaluation revealed acicular α′ martensite in the fusion zone (FZ), resulting in the highest hardness of 383 HV, compared with 332 HV in the heat-affected zone (HAZ) and 297 HV in the base metal (BM). Energy-dispersive X-ray spectroscopy detected only 0.02

Ensuring the stability and reliability of power transmission networks requires accurate and timely fault detection. Conventional fault detection methods often struggle with adaptability, computational efficiency, and robustness to noisy conditions. This work proposes an intelligent transmission line fault detection and classification through an advanced deep learning based model. The proposed method converts time series current and voltage signals into structured feature representations by using fast Markov transition field transformation, and processes these feature representations with a ResNet34 based deep feature extractor with squeeze and excitation module. The proposed model enhances classification accuracy while decreasing the computational complexity, enabling real time fault detection. A 735 kV transmission line system is modeled in the Simulink environment to generate a diverse dataset under various fault scenarios. The proposed model can detect such faults with an accuracy of 99.98

The diagnosis and treatment of brain tumors can be challenging. They are a main cause of central nervous system disorder and uncontrolled proliferation. Early detection is also very important to ensure that the intervention is successful and delayed diagnosis is a significant factor contributing to lower survival rates for specific types. This is because the doctors lack the necessary experience and expertise to carry out this procedure. Classification systems are required for the detection of brain tumor and Histopathology is a vital part of brain tumor diagnosis. Despite the numerous automated tools that have been used in this field, surgeons still need to manually generate annotations for the areas of interest in the images. The report presents a vision transformer that can analyze brain tumors utilizing the Convolution Neural Network framework. The study’s goal is to create an image that can distinguish malignant tumors in the brain. The experiments are performed on a dataset of 4,855 image featuring various tumor classes. This model is able to achieve a 99.64% accuracy. It has a 95% confidence interval and a 99.42% accuracy rate. The proposed method is more accurate than current computer vision techniques which only aim to achieve an accuracy range between 95% and 98%. The results of our study indicate that the use of the ViT model could lead to better treatment and diagnosis of brain tumors. The models performance is evaluated according to various criteria, such as sensitivity, precision, recall, and specificity. The suggested technique demonstrated superior results over current methods. The research results reinforced the utilization of the ViT model for identifying brain tumors. The information it offers will serve as a basis for further research on this area.