Overview of Machine Learning Methods for Lithium-Ion Battery Remaining Useful Lifetime Prediction

Abstract

Lithium-ion batteries play an indispensable role, from portable electronic devices to electric vehicles and home storage systems. Even though they are characterized by superior performance than most other storage technologies, their lifetime is not unlimited and has to be predicted to ensure the economic viability of the battery application. Furthermore, to ensure the optimal battery system operation, the remaining useful lifetime (RUL) prediction has become an essential feature of modern battery management systems (BMSs). Thus, the prediction of RUL of Lithium-ion batteries has become a hot topic for both industry and academia. The purpose of this work is to review, classify, and compare different machine learning (ML)-based methods for the prediction of the RUL of Lithium-ion batteries. First, this article summarizes and classifies various Lithium-ion battery RUL estimation methods that have been proposed in recent years. Secondly, an innovative method was selected for evaluation and compared in terms of accuracy and complexity. DNN is more suitable for RUL prediction due to its strong independent learning ability and generalization ability. In addition, the challenges and prospects of BMS and RUL prediction research are also put forward. Finally, the development of various methods is summarized.

Health prognostics for lithium-ion batteries: mechanisms, methods, and prospects

“…Lack of aging mechanisms introduction Ref. 46 Introduction of ML-based lifetime prediction methods Comparisons among different ML methods…”

Data-driven methodology for optimal Lithium-ion battery RUL prediction

“…The advantages of the Machine Learning (ML) and Deep Learning (DL) methods, such as the ability of ANN approaches to learning non-linear relationships between the features, have allowed researchers to develop many techniques that can help to predict the RUL of a Li-ion battery accurately [10]. The hybridization of diverse AI-based methods has also been developed in several studies [11]. In [12], the authors present a combination of particle swarm optimization with a Long Short-Term Memory (LSTM) network, a highly used recurrent Neural Network (RNN), an attention procedure for RUL prognosis and SOH observation of the Li-ion batteries.…”

“…The discussion centers on the trade-off between model interpretability and performance. Both have achieved good results in predicting the RUL of Li-ion batteries with a Root Mean Squared Error (RMSE) less than 26 cycles in the case of Gaussian Process Regression [11] and less than 6 cycles by using Recurrent Neural Networks (RNN) [15]. As far as traditional machine learning models are concerned, kernel regression approaches can capture the non-linearity between the extracted features and the RUL [16].…”

Hierarchical Diagnostics and Risk Assessment for Energy Supply in Military Vehicles

“…In the calculation methods, the battery parameters (SOC-state of charge, SOH-state of health) are estimated using different algorithms [25][26][27]. Some of these are highlighted: coulomb counting [28], modified coulomb counting [29], statistical approaches [30,31], hybrid methods [28], machine learning [32][33][34], and degradation pattern recognition with transfer learning [35].…”