Online Prediction Method of RUL of Lithium Battery Based on CEEMDAN-ILSTM

Abstract

Aiming at the problems of non-linearity of remaining life data caused by the uncertainty of lithium battery discharge behavior and local over optimization in the process of optimizing parameters by swarm intelligence algorithm, a short-term and long-term memory neural network (CEEMDAN-ILSTM) prediction method based on data decomposition and improved whale optimization algorithm is proposed. In this method, the CEEMDAN algorithm is used to decompose the original data to preliminarily simplify the hidden change law in the data, and the improved whale optimization algorithm is used to optimize the parameters of the neural network to find out the network parameters that can make the current prediction model have better prediction performance. Through the evaluation on the lithium battery data set provided by NASA, compared with the current mainstream methods, the method proposed in this paper has better prediction performance than other methods.