Abstract
Lithium ion batteries (LIBs) have revolutionized the era of electrical energy storage by offering high energy density and longer life cycles in various applications such as electric vehicles, electronic gadgets, satellites and power grids. To achieve optimum and reliable performance throughout their life cycle, accurate monitoring of their state parameters such as state of charge (SOC), state of health (SOH), and remaining useful life (RUL) needs to be estimated precisely. Filter-based and data driven techniques estimate these parameters accurately even under dynamic battery operation. In this paper, first, we have given details about experimental techniques through which LIB state parameters are estimated, but due to poor nonlinearity handling capacity of these models, we showcase the potential of various filter-based and data driven techniques with a variety of features extracted from LIBs. Subsequently, we discuss the working and performance of various filter based and data driven algorithms utilised in predicting the state parameters of batteries such as SOC, SOH & RUL in detail. Additionally, a comparative table comprising features, predictive techniques and performance is made to highlight the effectiveness of each method. Finally, we propose a strategy to improve the estimation accuracy of LIBs. Overall, the paper provides a comprehensive review of various estimating lgorithms and their potential in predicting the state parameters of LIBs with an aim to develop an intelligent framework for required applications and highlights the challenges which are yet to be overcome.