| In the recent years, with the lack of petroleum resources and the degradation of air environment, the energy saving and environmental protection of electric vehicle is the development direction of automobile industry. Battery is the main power source of electric vehicles and its price occupies a large proportion in the whole vehicle cost. Thus, it is great significance that battery management system (BMS) for on-board battery under effective control and management, to improve the battery performance and prolong the battery life. The state of charge estimation of battery is the core function of BMS, not only can it reflects the current remaining power of battery, but also provide a reasonable control strategy for vehicle.In this paper, in order to improve the SOC estimation based on the extended kalman filter, the battery model with hysteresis effect adjusted factor and the suboptimal fading factor of extended kalman filtering algorithm are proposed. Firstly, a series of performance testing experiment for VRLA are designed to acquire the static relationship between SOC and its influence factors, and under the analysis of the hysteresis effect of battery, the battery model with hysteresis effect adjusted factor is put forward. Secondly, battery model parameters are identified by the least square recurrence method. Finally, the compare of simulation voltage of model and true voltage of battery validate the model is reasonable.SOC estimation based on extended kalman filter is low precision when the battery model is uncertainty and the statistical noise is inappropriate. Thus, a suboptimal fading factor extended kalman filter (SFEKF) algorithm is proposed, it can remain strong track with the true SOC in the condition of poor compatibility of model and inaccurate noise statistics. Then in the same UDDS conditions, the SFEKF and EKF are adopted to estimate the SOC of VRLA, the simulated result show that the SFEKF algorithm has a higher accuracy and better robustness in SOC estimation. |
PDF Full Text Request