A Neural Network Based State-of-Health Estimation of Lithium-ion Battery in Electric Vehicles

Abstract

Abstract As one of the main promising power sources in electric vehicles (EVs), lithium-ion battery plays an important role in EVs’ power system. Its state-of-health (SOH) estimation is a key technology in the battery management system (BMS). Many battery parameters can reflect the battery's health, like internal resistance, internal capacity, etc. In this paper, we use maximum available capacity to indicate the battery's SOH based on a back propagation (BP) neural network. The main contributions of this paper include: (1) a direct parameter extraction method by HPPC (Hybrid Pulse Power Characterization) test is employed to identify the battery parameters of the first-order equivalent circuit model. (2) The parameters can be used to train the three-layer BP neural network, therefore the structure and parameters including weights and thresholds of the network can be determined. This well-trained neural network is used to estimate the SOH. (3) The static and dynamic current profile tests are carried out to verify the accuracy of the training results of the proposed neural network. The experimental results indicate that the proposed neural network based estimation method can present accuracy and suitability for SOH estimation with low computation cost.