Frequency Selective Filters for Estimating the Equivalent Circuit Parameters of Li-Ion Battery

The most difficult part of designing a battery management system (BMS) is accurately modelling the battery. So far, the most suitable model is the equivalent circuit model (ECM). The battery is a very complex system, and with the increase in model order, the accuracy should improve. However, the higher order model would face the tendency of over-parameterization having unfavorable estimation issues. In time domain it becomes difficult to fit the higher order difference equations, because of the problem of ill conditioning. Therefore, in this paper, the frequency domain parameter estimation is proposed to overcome the aforementioned limitations. Here, the ill-conditioning problem is resolved by filtering and estimation around the dominant frequencies. The estimated parameters are validated with the simulation data.