A performance degradation assessment method of equipment based on multi-attributes

Abstract

The invention discloses a device performance degradation evaluation method based on multiple attributes, which uses a three-layer wavelet packet decomposition algorithm to analyze a device performancedegradation data sample, and takes the energy feature of a third-layer node as a feature vector of the device performance degradation, thereby realizing multi-scale feature extraction of data change.Then, K-Means baed clustering method is used to cluster the eigenset of the equipment in normal state, and the cluster center, the maximum cluster radius and the maximum angle of the cluster are obtained, which are fused as the benchmark of the evaluation of equipment performance degradation state. The distance and angle which can describe the change of equipment running state are constructed, and the measurement index of equipment degradation state is established based on the distance and angle. Finally, the effective analysis of equipment performance degradation is realized. The invention can accurately and timely analyze the performance degradation degree of the equipment, and formulate the maintenance plan or the replacement plan of the equipment pertinently.

Description

technical field [0001] The invention belongs to the technical field of equipment safety, and in particular relates to a multi-attribute-based equipment performance degradation evaluation method. Background technique [0002] The equipment will experience a series of different degrees of performance degradation from normal to complete failure during service operation. In order to ensure the safe and reliable operation of the equipment, it is necessary to accurately and timely analyze the degree of performance degradation through the monitoring of the degradation process. Make a maintenance plan or replacement plan for the equipment in a timely manner. Therefore, using performance monitoring data during equipment service to analyze and evaluate equipment performance degradation is of great significance for preventing accident losses caused by equipment failure, and it is also a hot topic in current research. [0003] In practice, the whole life cycle data of equipment is ofte...

AI Technical Summary

Problems solved by technology

However, the existing methods mostly describe the degree of deviation between the degraded data and the normal state data from the perspective of a single attribute (such as distance), and lack of similarity measurement analysis from the perspective of multiple attributes, and its effectiveness is difficult to guarantee; at the same time, the normal state of equipment The effective characterization of the following data features will be the benchmark for analyzing performance degradation. The existing methods lack consideration of the overall characteristics of data changes, making it difficult to guarantee the validity of normal state characterization

Images