Online state of health monitoring of lithium-ion battery based on model error spectrum for electric vehicle applications

Abstract

Lithium-ion battery, as the mainstream power source of electric vehicles, suffers the performance degradation in the long-term vehicular usage. Battery health monitoring is quite an important task of the battery management system. This paper presents a novel method of online estimating the battery state of health (SOH) based on the evolution of the model error during the battery aging process. The monitoring model is established by using the test data of fresh batteries. During the process of battery aging, the model precision gradually declines and there is an implicit correlation between the aging degree and the distribution of model error. The model error spectrum (MES) is established based on the kurtosis and skewness of the model error statistics. It is found that there is an exponential relation between battery SOH and the MES. Based on the above work, the online SOH estimation is proposed by obtaining the MES within the historical data in a pre-set time window. The experiments are used to evaluate the presented method. The results indicate that the precise estimation results of battery SOH can be obtained with the error less than 1.07%. It is expected that the presented method is useful in online health monitoring and prognosis applications of the lithium-ion batteries in electric vehicles.