State of charge estimation of lithium ion battery for electric vehicle using cutting edge machine learning algorithms: A review

The accurate estimation of State of Charge (SoC) in Lithium-Ion Batteries (LIBs) is paramount for enhancing the efficiency and reliability of Electric Vehicle (EV) applications. In this study, the significance of using cutting-edge Machine Learning (ML) approaches for the estimation of SoC in LIBs was highlighted for EV applications. Various advanced algorithms, including Hybrid Long Short-Term Memory (HLSTM) networks, Hybrid Gate Recurrent Unit (HGRU), Multistage Temporal Convolutional Networks (MTCN), Recurrent Deep Belief Forward (RDBF) networks, Hybrid Slap Swarm Algorithm (HSSA), Robust Adam Optimizer (RAO), and Wavelet Neural Network (WNN), are comprehensively investigated for effective SoC estimation. Simulation results reveal the superior performance of these algorithms in estimating SoC for LIB systems, particularly in EV applications. The Effectiveness of above-specified algorithms is thoroughly evaluated through rigorous training and validation processes. The algorithms HSSA, RAO, and WNN demonstrate effective performance by achieving reduced Mean Square Error (MSE) and Mean Absolute Error (MAE) within the given operating range, corresponding to the levels of SoC estimation. © 2024