Dynamic equipment residual life prediction method based on data driving

The invention relates to a dynamic equipment residual life prediction method based on data driving, and relates to the technical field of residual life prediction, and the method comprises the steps: collecting the data of dynamic equipment; processing the data to obtain first data; degeneration features are extracted, and the degeneration features are integrated into a degeneration feature vector; defining a health index, and obtaining a composite health index; and establishing a nonlinear Wiener degradation model, and based on the first data, estimating characteristic parameters of the nonlinear Wiener degradation model by adopting a least square method to realize prediction of the residual life of the dynamic equipment. According to the method, the observation values of the dynamic equipment are collected, and the observation values are processed and subjected to feature extraction, so that the degradation mode of the dynamic equipment can be captured, and the performance change of the dynamic equipment ca