Abstract:The single parameter can not accurately represent the engine performance degradation process, and traditional intelligent learning model is difficult to accurately fit the engine degradation model. In order to solve the above problems,an engine health index (HI) based on data fusion was proposed, and the residual life prediction method was presented combining multi model similarity matching with integrated model. Hierarchical clustering and contour coefficient were used to select parameters, which were integrated into engine health index. The genetic algorithm was used to optimize the random forest to fit the engine performance degradation process, and the multi model similarity matching was used to optimize the prediction results of the regression model. A turbofan engine simulation data set (C-MPASS) were selected to verify the effectiveness of the proposed method. The results show that the RMSE and MAE of the method are 6.128 and 4.901 respectively,and matching the health index and multi-model similarity greatly improves the prediction accuracy of engine remaining life.