Cascading Fault Analysis and Control Strategy for Computer Numerical Control Machine Tools Based on Meta Action

Abstract

Cascading faults of computer numerical control machine tools will lead to frequent machine faults, but an effective fault analysis and control strategy can significantly reduce the frequency of such cascading faults. The numerous studies have focused on faulty parts using established fault diagnosis models centered on faulty parts and developed measures for controlling machine tool reliability from the parts' perspective. However, the internal parts of machine tools are strongly coupled, and when one part or assembly begins operating abnormally, it can easily initiate the abnormal operation of other parts or the entire assembly, which can lead to failures. Thus, a fault diagnosis method that considers only individual faulty parts ignores the problem of fault propagation between parts, which may lead to failure to locate the root cause. To address this problem, this paper proposes a cascading fault analysis and control strategy based on meta action. First, the functional components are decomposed into meta actions using the “function-motion-action” principle. Second, the directed edges and hierarchical digraph of meta action fault propagation are obtained. Third, the PageRank algorithm is used to calculate the fault propagation impact degree of each directed edge, and the key paths of meta action fault propagation are extracted based on the hierarchical digraph and the impact degree. Finally, the fault tree analysis method is applied to analyze the meta action on key paths, and feasible machine control measures are developed. The feasibility and effectiveness of the method are demonstrated using an example machining center.