Health Assessment Of Complex Product Component Combined With Class Interval Estimation

In order to meet the various functional requirements of complex products,the complexity and intelligence of its components are constantly increasing.As a result,the fault loss is getting larger and larger,and the maintenance cost is increasing.Therefore,it is necessary to establish a health management system for the operation status of complex products and evaluate the status of products in real time.This helps to identify potential anomalies,facilitate timely development of appropriate maintenanee work,improve product reliability and safety,and ensure that it can operate in a healthy and stable state.This thesis proposes a health assessment method for complex product components based on health index combin ed with group distance.estimation.First,using Mahalanobis-Taguchi System(MTS)for attribute reduction,deleting the characteristic parameters that are not useful for health assessment,and filtering out the eombination of effective feature p arameters.Next,for the class interval problem in histogram method,the closeness between parameter probability dertsity function curve and the upper contour line of histogram is used as the decision criterion.After the optimal class interval is obtained,a health index(HI)is established to determine the degree of degradation of the product's health status.The test proves that the method can effectively remove the abnormal fluctuation caused by eavironmental disturbances,and can also diagnose the degradation of eomplex product components when the Mahalanobis distance is lower than the threshold.Compared with the Mahalanbis time series,it can detect product component degradation earlier and has better stabilitr and real-time performance.