Modelling of A18560 Lithium-ion Battery Based on System Identification Method

Abstract

In the automobile industry, the battery management system (BMS) plays a big role in supporting the industry by giving protection to the battery from upcoming impairment. Its thorough study in forecasting battery lifelong and the preservation of the battery's condition upon battery's operation are the core business in BMS. The BMS executes various methods in order to accomplish its goal through single or multiple function integration, state of charge (SOC) and state of health (SOH) determination, battery's life continuation, battery's cell balancing, the communication and data acquisition onboard and off-board modules and the analyzation of battery's historical data via monitoring and storing battery's system. While in this paper, we would like to propose SOC based on the Discharger Test Technique that plays an important role in BMS as it could estimate the critical characteristic of the battery. The changes of battery voltage value are collected until the battery reach the minimum voltage limit for safety purpose. The collected data during the discharging process is used to model battery via different types of System Identification approach or BlackBox model including Transfer function, Nonlinear, and State Space model. The models are manually tuned to get the best result and are compared among them. The result shows the Nonlinear model represented the greatest percentage among others, hence giving the best fit of the battery model and offering accurate Lithium-ion of SOC monitoring status.