A Novel Approach for Accurate SOC Estimation of Lithium-Ion Electric Vehicle Batteries Using a (Q, S, R)-$γ$-Based Dissipativity Observer
提出一种基于耗散性理论的观测器，用于在模型不确定和噪声环境下精确估计电动汽车锂离子电池的荷电状态，实验验证其精度优于自适应无迹卡尔曼滤波，最大精度达99.23%。
For the first time, this article presents a dissipativity-based observer design for accurate state-of-charge (SOC) estimation, essential for improving the safety, performance, and lifespan of lithium-ion batteries (LIBs) in electric vehicle (EV) battery management systems (BMSs). However, model uncertainties and measurement noise significantly affect estimation accuracy. To address this, a novel observer design based on ( $\mathcal {Q}, \mathcal {S}, \mathcal {R}$ )- $\gamma $ -dissipativity theory is developed, formulated within a linear matrix inequality (LMI) framework, and integrated with the Lyapunov-Krasovskii functional (LKF) approach. The proposed observer ensures robustness and stability in SOC estimation under uncertain and noisy conditions. A one-resistor capacitor (1-RC) equivalent circuit model (ECM) is adopted for battery modeling, with experimental validation performed on a Panasonic 18650PF cell. The proposed method is compared against the adaptive unscented Kalman filter (AUKF) under four drive cycles: the urban dynamometer driving schedule (UDDS), the aggressive US06 supplemental federal test procedure, the Los Angeles 92 (LA92), and the highway fuel economy test (HWFET). Results show that the proposed observer achieves root-mean-square errors (RMSEs) of 0.77%, 0.50%, 0.65%, and 0.48% and mean absolute errors (MAEs) of 0.59%, 0.42%, 0.50%, and 0.40% under UDDS, US06, LA92, and HWFET, respectively. This corresponds to RMSE reductions of 28.38%, 88.93%, 67.25%, and 38.35% compared with AUKF. Notably, the proposed method achieves a maximum accuracy of 99.23%, surpassing the latest reported accuracy of 98.50%.