Adopting combined strategies to make state of charge (SOC) estimation for practical use

Abstract

The estimation of state of charge (SOC) requires the tradeoff between high accuracy and robustness in the design of the battery management system. There are varieties of studies being carried out around this issue, aiming to balance the model complication, algorithm complexity, estimation accuracy, as well as robustness. In this work, in order to solve the SOC estimation problem under real complex working conditions, we introduce a strategy that combines battery modeling tactics and algorithm developing techniques to make it. In detail, we employ a combined model and build discrete state-space equations based on it. For improving the estimation accuracy, we use the recursive least squares method with forgetting factor to identify the parameters of the model. The particle filter embedded genetic algorithm is employed for SOC estimation, which overcomes the particle degradation and diversity loss for further enhancing the accuracy and robustness of estimation. Finally, real road test data is applied to investigate the estimation performance of the developed SOC estimation strategy.