Joint Estimation of SOC and SOH of Lithium Battery Based on Dual Multi-Innovation Adaptive Unscented Kalman Filter

With the rapid development of UAVs, the estimation algorithm for State of Charge (SOC) with strong generalization ability and high accuracy can enhance the energy management performance of the entire UAV. However, as the number of cycles increases, the SOC of a lithium battery is easily affected by aging factors, and the noise from sensors in the vehicle Battery Management System (BMS) significantly influences data acquisition. This paper addresses the issue of insufficient accuracy in SOC and State of Health (SOH) estimation. To tackle this, a dual multi-innovation adaptive Unscented Kalman Filter (UKF) algorithm is proposed based on the multi-innovation theory and adaptive theory. The algorithm is compared with the Dual UKF (DUKF) and multi-innovation adaptive UKF-Extended Kalman Filter (MIAUKF-EKF) algorithms using the Urban Dynamometer Driving Schedule (UDDS). The results of the verification demonstrate that the DMIAUKF algorithm improves the accuracy.