Application of statistical parameter identification in battery management system

As a vital part of the battery employed systems, battery management system (BMS) must correctly estimate values descriptive of the battery's present operating condition. Without a doubt, state of charge (SOC) is a key battery state for BMS to estimate. This paper carefully analyzes the operating performance of VRLA battery through a large number of experimental tests; a statistical parameter identification strategy is employed in electromotive force (EMF) SOC estimation method, then this EMF method is combined with ampere-hour (Ah) counting through a parallel feedback weighting structure, finally an improved SOC estimator is built up. A BMS platform based on TI DSP2407 is established; besides, a BMS monitoring system is built in VC++ as an auxiliary tool. The results tend to be satisfactory.