Abstract:To solve the problem that a single characterization parameter could not express the paths of degradation completely and the correlation between multiple parameters could cause a bad effect,a life prediction method based on master metadata for large-scale complex electromechanical equipment was presented. Firstly, mutual information from multiple features was fused by using the method of principal component analysis, and the master metadata that could characterize the equipment degradation and had no correlation with each other were obtained. Then, a Wiener degradation process model based on the master metadata was constructed. Finally, a specific analysis and verification was made by using the data adopted from NC machine tool.The examination identifies the effectiveness of the proposed method in large-scale complex electromechanical equipment by eliminating the correlation between the characterization parameters.