SOC estimation optimization method based on parameter modified particle Kalman Filter algorithm
- 146 Downloads
Traditional Kalman Filter algorithm requires the system noise to be Gaussian distribution, but the power battery operating condition generally can not meet the requirement due to complexity and disturbance by the environment. However, the Particle Filter algorithm can adapt to various forms of system noise. In this work, the calculation process of the standard Particle Filter algorithm is improved based on the engineering characteristics of SOC estimation. In the calculation process, the key parameters including the total number of particles and the effective particle threshold are optimized and verified under FTP75 and NEDC conditions. The systematic error under different conditions is evaluated, based on the vehicle platform computing capacity, the proposed total number of particles is 1000, the effective particle threshold is 0.01. In this case, the SOC estimation accuracy can reach 1–2%, meeting the practical requirements.
KeywordsSOC estimation Particle Filter algorithm Optimization modified parameters Recommended parameters
This research was supported by the National Natural Science Foundation of China (51477125), the Hubei Science Fund for Distinguished Young Scholars (2017CFA049), the Wuhan youth morning project (2016070204010155), and the Fundamental Research Funds for the Central Universities (WUT: 2017II40GX).
- 25.Pathuri, B.V., Unterrieder, C., Huemer, M.: Battery internal state estimation: a comparative study of non-linear state estimation algorithms. In: Proceedings of the 2013 IEEE Vehicle Power and Propulsion Conference (VPPC), Beijing, China, 15–18, pp. 1–6 (October 2013)Google Scholar