Implementation method of avionics fault diagnosis system for airborne route maintenance

Abstract

The invention discloses an implementation method of an avionics fault diagnosis system for airborne route maintenance. According to the method, a Bayesian network is used as a framework to establish afault diagnosis model, and diagnosis analysis is carried out by combining a joint tree algorithm, so that the limitation of low efficiency of traditionally eliminating faults by inquiring a maintenance manual is solved, the dependence on actual experience of maintenance personnel is reduced, and the reasoning efficiency is improved; strong association rules of historical maintenance data are mined by using association rules, and parameter learning is carried out by combining expert experience, so that the limitation that traditional methods only depend on expert experience is solved, and thehistorical maintenance data are fully utilized; maintenance information of each member system is obtained in real time, the maintenance information and fault information input by maintenance personnelthrough a man-machine interaction window are used as observation evidences together, the observation evidences are input into the fault diagnosis model to update posterior probability distribution inreal time, and the dynamic diagnosis process of airborne route maintenance is achieved. The accuracy of fault diagnosis is improved, and a certain theoretical basis is provided for the design of a domestic airborne maintenance system.

Description

technical field [0001] The invention relates to the technical field of civil aviation airborne route maintenance, in particular to an implementation method of an avionics fault diagnosis system oriented to airborne route maintenance. Background technique [0002] The avionics system is referred to as avionics system for short, mainly including avionics systems such as navigation and communication. The structure of the avionics system is complex, there is a high degree of information interaction between each subsystem, and there is great uncertainty in the failure. After a failure, it is likely to cause failures in related structural parts. Sometimes the same failure symptom corresponds to several failure causes, and the same failure cause will cause multiple failure symptoms. There are complex forms such as related failures and multiple failures; The system is often used in an environment with complex interference factors. Even for the same type of aircraft, the types of fa...

AI Technical Summary

Problems solved by technology

[0004] 1. Technical issues: The structure of the avionics system is complex, the data is highly cross-linked, and there are complex forms such as multiple faults and related faults. One fault sometimes corresponds to multiple fault symptoms, and one fault symptom sometimes corresponds to multiple fault causes. Determining the one-to-one correspondence between failure phenomena and failure causes is a difficult problem; it is difficult for avionics systems to obtain complete maintenance data. How to effectively analyze historical maintenance data and expert experience on the basis of incomplete historical maintenance data? It is a difficult problem; historical maintenance data records are uneven, how to unify and enter them, how to quickly mine the correlation information between faults in historical maintenance data, and how to implement real-time fault diagnosis model with the increase of historical maintenance data Dynamic learning of parameters is a difficult problem

Images