基于熵权法多分布融合的数控机床故障分布建模OA

To address the limitations of fault distribution models for complex systems such as computer nu-merical control(CNC)machine tools,where differences of fault distribution models are not considered for va-rious subsystems or component types,a method based on the entropy weight method is proposed to integrate different distribution models.First,multiple candidate distribution types that pass hypothesis tests are select-ed.Maximum likelihood estimation is then applied to fault data with right censored time to obtain distribution model for each subsystems or component types.Finally,taking the empirical function model as the bench-mark,the entropy values of the individual distribution model and their corresponding weights in the fused model are calculated.These weighted models are subsequently fused into the final fault distribution model.Fault data from ten CNC machine tools are analyzed,and the results show that the CNC machine tool fault dis-tribution model obtained by mult-i distribution fusion exhibits strong robustness,and the results obtained from the analysis are highly reliable.