Data-driven model for state of health estimation of lithium-ion battery

The trend of electrification within the automotive industry has resulted in the development of electric vehicles (EVs) due to their sustainable and environment-friendly feature. One of the major components of these cars is their energy storage device, as its performance is very crucial for the success of the EV. But the major concern is the properties of the energy storage or battery change with time and usage. Therefore, the degradation of battery properties is interesting, especially the capacity decline. To understand and counter this degradation, it must be measured with high precision. This paper develops a data-driven approach to estimate the health of the batteries by using the support vector machine model. The proposed method identifies the keen points of the operation of an EV battery under aging conditions by performing accelerated aging and trains them with the adapted data-driven approach. To access the aging conditions, an equivalent electrical model of a cell of EV battery is simulated and operated for various current pulses. Further, when tested with unkown operation of a battery, the developed estimator predicted its health with an accuracy of 96.25%. The results depicted that the proposed concept gives an improved measure of the battery state and a pattern on how the capacity degradation occurs with time.