State of energy estimation for lithium-ion batteries using adaptive fuzzy control and forgetting factor recursive least squares combined with AEKF considering temperature

Abstract

As the main energy storage component of electric vehicles (EV), lithium-ion battery state estimation is an essential part of the battery management system (BMS). State of Energy (SOE) is one of the important state parameters, and its accurate estimation effectively reduces the potential safety hazards in the use of lithium-ion batteries, improves the efficiency of energy utilization, and alleviates the mileage anxiety of drivers. To solve the problem that the prediction of SOE of lithium-ion batteries is greatly influenced by temperature, a novel method called adaptive fuzzy control forgetting factor recursive least squares-Adaptive extended Kalman filtering (AFCFFRLS-AEKF) is formed. A fuzzy logic controller is designed for adaptive adjustment of the online parameter recognition forgetting factor with the change of working conditions. To solve the problem that the open-circuit voltage (OCV) changes with the influence of temperature in the variable temperature range, the regression analysis method is used in modeling to realize the regression analysis of OCV in a wide temperature range. Estimation accuracy is verified under two working conditions. The error of the estimation considering the temperature effect converges within 1 %, which achieves higher estimation accuracy and stronger robustness.