Investigations on Using Intelligent Learning Techniques for Anomaly Detection and Diagnosis in Sensors Signals in Li-Ion Battery—Case Study

This research paper aims to design and implement an intelligent least short time memory (LSTM) deep learning classification technique detect possible anomalies in measurements dataset within a particular Li-ion battery type. For the state of charge (SOC) faults estimation, Joint State Parameter Extended Kalman Filter (JEKF) estimator is developed. The SOC accuracy performance excellent, with less than 0.5% error during steady-state, compared 2% reported literature. implementation JEKF parameter estimation chosen preset Simulink Simscape generic model. It also helpful generate healthy faulty measurement proposed LSTM classifier technique. model wisely selected for “proof concept” purpose, validation, algorithms’ robustness, accuracy, effectiveness. Compared traditional EKF fault diagnosis isolation (FDI), model-based strategy, data-driven-based algorithm high (around 80%) loss close zero. Therefore, this feature makes data collection directly from sensors possible, which beneficial generating online scenarios. Additionally, can remarkably classify all detected independent uncertainties, unmodeled dynamics. Also, high-performance root mean square (RMSE) 0.0588 (voltage fault), approximately 5.5×10−7 (healthy) 8.87 × 10−6 (current fault) shallow neural network (DLSNN) reveals obvious superiority both FDI strategies.