Vehicle lithium battery health state on-line prediction method adapting to fast charging strategy

The invention relates to the technical field of lithium batteries, in particular to a vehicle lithium battery health state online prediction method adapting to a fast charging strategy. The method is composed of two parts of off-line model development and multi-task model online application. The off-line model development part comprises hyper-parameter adjustment and model training, the multi-task model online application part comprises model transfer learning, online measurement is performed on a target application monomer during model transfer learning, voltage characteristics of private attributes of a battery and cycle times of common attributes of the battery are obtained, and the voltage characteristics of the private attributes of the battery and the cycle times of the common attributes of the battery are obtained. And carrying out feature fusion on the voltage features and the cycle index, inputting the fused voltage features and cycle index into a trained offline model for prediction, and carrying ou