State of charge estimation of ternary lithium-ion batteries at variable ambient temperatures

Abstract

Among the factors that affect lithium-ion batteries, ambient temperature has a great influence on the charge and discharge rate, general battery performance, and storage capacity of the battery, hence accurate state of charge (SOC) estimation over a wide range of temperatures is essential and necessary. In this paper, a Second-order Thevenin equivalent circuit model is established with ambient temperature compensation over varying ambient temperature conditions. An improved Fixed Range Forgetting Factor-Adaptive Extended Kalman Filtering (FRFF-AEKF) algorithm with the Saga-Husa Adaptive filter (SHAF) is proposed for the accurate estimation of SOC at variable ambient temperature. The temperatures −10°C, 10 °C, 25 °C, and 40 °C are set to depict low, normal, and high ambient temperatures in which lithium-ion batteries in electric vehicles (EVs) and most electronic devices usually operate. Tests conducted include the hybrid pulse power characterization (HPPC) for obtaining data required for parameterization, and for SOC estimation and verification, the Beijing Bus Dynamic stress (BBDST). The object of the tests was a ternary lithium-ion battery representing those used in electric vehicles (EVs). To improve adaptability, reduced noise, and achieve faster convergence with increased accuracy, the range of forgetting factor is fixed between a variable range of 0.997–1. To verify the performance of the proposed algorithm under varying conditions and different ambient temperatures, the SOC estimation result was compared with that of the adaptive extended Kalman filtering (AEKF) algorithm. The results show that the proposed improved algorithm implemented with the established Second-order Thevenin equivalent model has a 1.1 % better estimation accuracy compared to the results of the AEKF algorithm. This confirms that the proposed method can achieve accurate SOC estimation and is adaptive to variable ambient temperatures.