Voltage and current signals de-noising with wavelet transform matrix for improved SOC estimation of lithium-ion battery

The electromagnetic environment of the lithium-ion battery in the Electric Vehicles (EVs) is severe. Moreover, the load current of the battery in the EVs changes drastically and randomly depending on the EV driving condition. As a result, the voltage and current signals measured by the Battery Management System (BMS) normally contain the noise such as the white noise. This results in the estimation error of the State of Charge (SOC). A new voltage and current de-noising approach based on Wavelet Transform Matrix (WTM) is proposed in this paper to improve the accuracy of the SOC estimation using Extended Kalman Filter (EKF) algorithm. This approach reduces the computation complexity and the measuring noise is de-noised effectively. It was validated by the experimental results on a 1665132 model laminated Li(NiCoMn)O2 battery with the rated capacity of 200 Ah and rated voltage of 3.6 V. The voltage of the battery ranges from 3.2 V to 4.2 V. The accuracy of the SOC estimation is improved significantly and the error is limited within 1.5% less than the error of 6% by EKF without de-noising.