Research Of Prognostic And Health Management System Based On Flight Data

With the development of the aircraft systems, integration, intelligence, the moreand more people’s attention on problem of reliability, maintainability, testability,supportability, security and cost, offering opportunities for Prognostic and HealthManagement (PHM) technology. PHM represents a change in a diagnostic method,from the traditional turn-based diagnostic sensor system based on predictiveintelligence, communications reactive steering pilot’s3R (ie, at the exact time to takeprecise and accurate maintenance activities on the site). When introducing PHM is tounderstand and predict failure may occur, or in the event of unexpected failure triggersa simple maintenance activities in order to achieve autonomous security, reduce theuse and protection costs.The paper firstly combined with a variety of features and condition monitoringsystems, PHM system itself presents the characteristics of multi-layer structure of opensystem that can be embedded with the continuous development of science andtechnology in different modules to improve the system. PHM system includes thefollowing major functions:(1) failure detection;(2) failure prediction;(3) healthassessment;(4) component life tracking;(5) performance degradation trend tracking;(6) warranty tracking;(7) failure selective reporting;(8) decision support and resourcemanagement;(9) information management, accurate information on the exact timeaccurately communicated to staff.Based on the above functions, the main work of this paper is to design the systemarchitecture facilitates openness continuous improvement and development oflanguage development environment for a brief, focused on the flight data and flightdata decoding application subsystem. Presented on dynamic list of flight data decodingalgorithm can effectively decode them improve fly participation application rateaccording to different models and flight data. And the establishment of case-basedreasoning fault diagnosis expert system of aircraft. In this paper, a detailed study of the object-oriented technology and distributed case, said the case shows the structure ofthe organization. The system improves the case similarity algorithm uses hierarchicalretrieval algorithm search strategy combines with nearest neighbor and made a casestudy based on the similarity of the case. The system not only meets the aircraft PHMmanagement requirements, and simulate the fault diagnosis expert, to a certain extent,to overcome the existing method of fault diagnosis aircraft problem. Finally, thesystem was actually designed to run, showing that appear in most systems runningsystem interface.