Adaptive Multitimescale Joint Estimation Method for SOC and Capacity of Series Battery Pack
Fang Liu, Dan Yu, Weixing Su, Shichao Ma, Fantao Bu
Abstract
Aiming at the estimation timescale selection problem in the multi-state joint estimation for state of charge (SOC) and capacity at cell level of series battery pack, an adaptive multi-timescale dynamic time-varying strategy (AMts-DtvS) is proposed. This strategy includes a triggered update strategy for the mean capacity of the battery pack, a time-varying polling update strategy for the differential SOC, and a triggered polling update strategy for the differential capacity. This strategy can adaptively adjust the timescale of multi-state joint estimation throughout the entire lifecycle of the battery pack based on the operating conditions, the degree of consistency deterioration and the change rate of capacity, achieving the goal of balancing complexity and estimation accuracy. Based on AMts-DtvS, an adaptive multi-timescale H infinity filter (AMts-HIF) algorithm is formed to achieve joint estimation for cell SOC and cell capacity of the battery pack. In the experimental section, based on 4 different datasets, the proposed AMts-HIF is compared with 3 different fixed timescale series battery pack state estimation algorithms. Through comparative verification, it can be concluded that the proposed AMts-HIF based on AMts-DtvS can obtain comparable estimation accuracy with less computational complexity in the discharge natural temperature risk scenario, Li (NiCoMn)O <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> battery natural aging scenario, and LiFePO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> battery natural aging scenario. In scenarios where capacity drop/consistency deterioration due to faults/low temperatures, etc, it is possible to obtain higher accuracy with comparable complexity.