New energy automobile power battery remaining service life adaptive prediction method based on WSA-LSTM algorithm

The invention relates to a new energy automobile power battery remaining service life self-adaptive prediction method based on a WSA-LSTM algorithm, and the method achieves the comprehensive analysis of the degradation characteristics of a power battery in the charging and discharging process through the construction of a health characteristic complete set. According to the method, the Pearson and Spearman correlation coefficient calculation and an entropy weight method are combined to extract main features with relatively large influence, and irrelevant influence factors are eliminated, so that the input monitoring data are simplified; according to the method, input characteristic data are trained by combining a whale Swarm Algorithm (WSA) and a long short term neural network (LSTM), so that the residual service life of the power battery is quickly and accurately predicted. Based on real charging and discharging data of the new energy automobile, the method is verified to have good robustness, dynamic accura