Remaining useful life of Lithium-ion batteries based on EMD-GSA-ELM

The widespread popularity of new energy vehicles has a higher demand for accurately determining battery health and ensuring battery safety performance. Considering that the capacity regeneration phenomenon during using lithium batteries is difficult to track, the model is difficult to build, and the prediction accuracy is not high. In order to reduce the computational complexity and improve the efficiency of establishing a lithium-ion battery remaining life prediction model, a method of fusion of empirical mode decomposition and extreme learning machine based on gravity search algorithm is proposed to establish a lithium battery remaining life prediction model. First, introduce EMD to analyze the battery's early health factor and capacity data to obtain fluctuation and trend components. Second, obtain a battery capacity decay prediction model by learning and training on different components. Finally, use NASA's lithium-ion battery test data to verify the performance of the EMD-GSA-ELM model. The experimental results show that the proposed model can better track the capacity decline and has better accuracy.