The Internet of Things (IoT) continues to expand rapidly, resulting in increasingly heterogeneous and complex wireless sensor networks (WSNs). Traditional anomaly detection approaches cannot cope with dynamic traffic patterns, high data volumes, and strict resource constraints. This study presents a hybrid XGBoost–CNN model that integrates XGBoost-based feature selection with a lightweight Convolutional Neural Network optimized for IoT environments. The proposed model was evaluated using real-world IoT traffic data and benchmarked against XGBoost, KNN, and SVM. Experimental results show that the hybrid approach improves detection accuracy by over 1%, increases throughput by 22–40%, and r... More >

Accurately determining the state of charge (SOC) is a critical factor in effective energy management for electric vehicles (EVs). Therefore, SOC variations in battery packs must be assessed with high precision. To simulate the complex processes within EVs that involve lithium-ion batteries (LIBs), an appropriate battery model is essential. Accurate parameter extraction through algorithmic methods is key to reliable SOC estimation. A dynamic, high-order equivalent circuit model, featuring two RC pairs in series with the battery's internal resistance, is employed to enhance parameter extraction. The values of the RC pairs are derived by solving equations that characterize the operational state... More >