Fault correlation analysis of CNC machine tools based on DEMATEL/ISM/Entropy weight method

Due to the advanced functionality and complex structure of CNC machine tools, faults are interrelated and can trigger a domino effect. Therefore, an effective approach to analyze the impact of fault propagation on the system and mitigate the domino effect is essential. Existing methods for identifying key subsystems, which rely solely on fault cause degree, centrality, or PageRank centrality, often overlook subsystem positions and indicator weights, thus limiting their effectiveness. To address this issue, an integrated DEMATEL/ISM/Entropy weight method is proposed. The approach begins by constructing a directed graph of fault propagation baesd on fault correlation analysis. Then, the integration of DEMATEL and ISM is used to develop a hierarchical model. The model incorporates influence degree, influenced degree, net influence degree, PageRank centrality, degree centrality, and betweenness centrality to form an importance matrix. The entropy weight method determines the indicator weights, and a weighted approach is used to identify key subsystems. A case study involving 9 correlated subsystems in a CNC machine tool validates the method, demonstrating that it outperforms the DEMATEL/ISM approach and achieves comparable accuracy with mainstream methods.