Fast Fault Diagnosis System Based on Data Mining AR Algorithm

Abstract

Aero-mechanical parts are an important part of the aircraft, and the maintenance of their failures also consumes a lot of manpower and financial resources. Therefore, the fault diagnosis research of aero-mechanical parts is of great significance for ensuring the safety of human life and reducing economic losses. With the development of fault diagnosis technology, the monitoring data is becoming more and more abundant and complex. The traditional methods of processing and analyzing the monitoring data have become more difficult, and it is difficult to establish accurate mathematical models. Therefore, the rapid diagnosis method of aviation machinery parts Become the research focus of fault diagnosis. This paper constructs a rapid fault diagnosis system for the construction of aviation machinery parts. Based on the input of past cases, new cases, literature cases, and book knowledge, the case library is refined and the graph library and rule term library are added. AR algorithm is used to mine and obtain Useful association rules between the decision attributes (failure mode, failure mechanism, failure reason, etc.) of the failure information in the database and the basic attributes (basic information other than the decision attributes), to achieve the purpose of assisting failure analysts in rapid fault diagnosis.