State of charge and state of health estimation strategies for lithium-ion batteries

Abstract

Due to the widespread use of renewable energy sources, lithium-ion batteries have developed rapidly because renewable energy sources such as photovoltaics and wind, which are very much affected by the environment and their power output can be better leveled if lithium-ion batteries are used. Battery state of charge (SOC) characterizes the remaining battery power, while battery state of health (SOH) characterizes the battery life state, and they are key parameters to characterize the state of lithium-ion batteries. In terms of battery SOC estimation, this paper optimizes the extended Kalman filtering (EKF) algorithm weights to adjust the weights during high current bursts to obtain better SOC tracking performance, and optimizes the back propagation (BP) neural network for SOH estimation to obtain better weights to further obtain more accurate battery SOH. The feasibility of the optimized algorithm is validated by the experimental platform.