Joint state of charge and state of energy estimation of special aircraft lithium-ion batteries by optimized genetic marginalization-extended particle filtering

Abstract

With the continuous development and widespread application of special aircraft, accurately estimating the performance and status of battery systems has become crucial. This paper focuses on the joint estimation of State of Charge (SOC) and State of Energy (SOE) under complex operating conditions using the proposed Genetic Marginalization-Extended Particle Filtering (GM-EPF) algorithm with the Dynamic Forgetting Factor Recursive Least Square (DFFRLS) algorithm. To enhance estimation accuracy, the paper first introduces DFFRLS algorithm for real-time model parameter recognition. Then, the GM-EPF algorithm is applied to combine the dynamically updated parameters from DFFRLS with particle filtering techniques, further improving the precision and robustness of the SOC and SOE estimations. The joint estimation algorithm of SOC and SOE based on DFFRLS ensures stable recognition with error control within 5.6 %. The joint estimation algorithm of SOC and SOE based on GM-EPF reduced the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) of battery SOC estimation by 82.91 % and 87.56 %, respectively, and the MAE and RMSE of SOE estimation by 84.61 % and 85.53 %, respectively. The joint estimation method of SOC and SOE for lithium-ion batteries in special aircraft based on composite model optimization has improved the controllability and safety of lithium-ion batteries as power sources in the field of special aircraft.