Parameter Optimization of Current Estimator for Lithium Polymer Battery (LiFePO4)

Abstract

Battery current estimation rises as there is a need to eliminate current sensor components in battery management system (BMS). Elimination of current sensors is intended to reduce the cost of BMS production. In addition, the elimination of current sensors can also reduce the total power consumption in the BMS. Instead of utilizing current sensors to read the current values, a current-estimation scheme is installed in the BMS. In this paper a current-estimation algorithm is proposed based on a simple battery model by updating the internal capacitance that changes over time based on a polynomial function of State of Charge - Open Circuit Voltage (SOC-OCV) relationship. The optimal order of the polynomial function is then sought in the hope of minimizing current estimation errors. To demonstrate the effectiveness of the proposed method, current estimation was performed for the pulsed-load test. The current estimation results are then compared to the current sensor readings. The results show that the current estimate is able to follow the trend of the current sensor readings.