Applications of artificial neural network based battery management systems: A literature review

Lithium-ion batteries have gained significant prominence in various industries due to their high energy density compared to other battery technologies. This has led to their widespread use in energy storage systems, electric vehicles, and portable electronic devices. However, lithium-ion batteries still face limitations, particularly concerning safety issues such as overheating and aging. BMS play a crucial role in ensuring safe and effective operation by providing control and monitoring functions. Among the key challenges in BMS is the accurate prediction of SOH, SOC, and RUL of the battery. Additionally, fault detection of lithium-ion batteries is an essential function of BMS. Given the complex electrochemical characteristics of lithium-ion batteries, there is a growing interest in developing advanced BMS that can accurately estimate the battery state. This review article focuses on the increasing popularity of ANN based methods for predicting the state of lithium-ion batteries. The literature review encompasses a wide range of studies on ANN-based battery management systems. The BMS applications and prediction methods for SOH, SOC, and RUL are thoroughly classified. The review covers state-of-the-art ANN methods, including feedforward neural network, deep neural network, convolutional neural network, and recurrent neural network, and provides a comparative analysis. The article also highlights current trends and identifies gaps in BMS applications. Furthermore, it offers insights and directions for future research and development, aiming to upgrade existing BMS or create advanced BMS systems. The comprehensive analysis presented in this review article serves as a valuable resource for research and studies seeking to enhance BMS capabilities and improve battery management strategies in the context of lithium-ion batteries.