State of charge estimation for space lithium-ion battery with super twisting sliding mode observer

Accurate determination of State-of-Charge (SOC) in lithium-ion batteries is critical for the reliable operation of space power systems. While the SOC estimation accuracy of space batteries is challenging due to the space environment variations, frequent charge-discharge cycles and limited onboard computational resources, etc. To address this, this paper proposes a novel super-twisting sliding mode observer (STSMO)-based SOC estimation algorithm for space lithium-ion batteries. The battery model employs a Dual Polarization (DP) equivalent circuit, enhanced through temperature-dependent parameter identification to establish a high-fidelity temperature-compensated model, and validated across various on-orbit temperatures. A STSMO-based SOC estimation algorithm is designed to overcome the limitations of conventional sliding mode observers(CSMO). Then, a hardware test system is developed, based on low-power embedded ARM Cortex-M3 processor, and experiments are conducted using on-orbit discharge data of a Cubesat. Simulation and experimental results demonstrate that the designed method achieves higher estimation accuracy compared to the CSMO and extended Kalman filter (EKF), and exhibits robustness against temperature variations. The proposed algorithm can be easily implemented in embedded processors, enabling accurate and temperature-robust on-board real-time battery SOC estimation of various space applications.