A review of machine learning state-of-charge and state-of-health estimation algorithms for lithium-ion batteries

Vehicle electrification has been proven to be an efficient way to reduce carbon dioxide emissions and solve the energy crisis. Lithium-ion batteries (LiBs) are considered the dominant energy storage medium for electric vehicles (EVs) owing to their high energy density and long lifespan. To maintain a safe, efficient, and stable operating condition for the battery system, we must monitor the state of the battery, especially the state-of-charge (SOC) and state-of-health (SOH). With the development of big data, cloud computing, and other emerging techniques, data-driven machine learning (ML) techniques have attracted attention for their enormous potential in state estimation for LiBs. Therefore, this paper reviews the four most studied types of ML algorithms for SOC and SOH estimation, including shallow neural network (NN), deep learning (DL), support vector machine (SVM), and Gaussian process regression (GPR) methods. The basic principles and uniform flowcharts of different ML algorithms are introduced. Then, the applications of each ML algorithm for state estimation within recent years are comprehensively reviewed and compared in terms of used datasets, input features, hyperparameter selection, performance metrics, advantages, and disadvantages. Based on the investigation, this review discusses the current challenges and prospects from four aspects, aiming to provide some inspiration for developing advanced ML state estimation algorithms.@2023 The Authors Publised by Elsevier ltd This is an open access article under the CC BY license