A novel approach for health management online-monitoring of lithium-ion batteries based on model-data fusion

Abstract

In order to ensure the safe and stable operation of electric vehicles and energy storage systems, online monitoring of the state of health and the remaining useful life of lithium-ion batteries is the key to the health management of lithium-ion batteries. A novel approach for health management online monitoring of lithium-ion batteries based on mechanism modeling and data-driven fusion is proposed in this paper. An improved semi-empirical capacity degradation model of the lithium-ion batteries fully considering internal resistance and temperature is established. After the data sets of the lithium-ion batteries are de-noised by the wavelet packet, the parameters of the model are identified according to the genetic algorithm and a particle filter framework is designed to online update the parameters of the model. Through the fusion of the two, the remaining useful life and state of health of the lithium-ion batteries can be predicted accurately. The proposed method is verified by the battery cycle test data from the Advanced Life Cycle Engineering Center of University of Maryland and the NASA Ames Prognostics Center of Excellence, the mean absolute error and root mean square error of the remaining useful life for the lithium-ion batteries are respectively less than 20 and 25 cycles at constant temperature condition, and respectively less than 3.30 and 3.60 cycles at non-constant temperature condition. Compared with the existing methods, the proposed method has higher prediction accuracy and better fitting performance, which can provide a certain theoretical basis for the safe operation of lithium-ion batteries.