A mechanical equipment residual service life prediction method based on KPCA-LSTM

Abstract

The application provides a kind of mechanical equipment remaining useful life prediction method based on KPCA-LSTM, comprising the following steps: collecting the degradation data of multiple sensors of mechanical equipment, the obvious sensor information of degradation trend is analyzed by kernel principal component, the contribution rate of each principal component is obtained;Determine the data corresponding to the first principal component and the second principal component, and obtain the comprehensive health index by fusion;The comprehensive health index is first-order derivative and second-order derivative, the degradation point of the equipment is determined, and the remaining useful life of the mechanical equipment is determined according to the degradation point of the mechanical equipment and the time point of failure;A Bayesian optimization LSTM remaining useful life prediction model for mechanical equipment is constructed;The degradation process of the remaining useful life of the mechanical equipment is predicted by the LSTM model. The method uses a multi-sensor data level fusion method, so that the fusion result can represent the degradation process of the equipment, and the Bayesian optimization LSTM prediction model is used to predict the degradation trend of the remaining useful life of the mechanical equipment.

Images