The Research Of Health Assessment And Fault Diagnosis Method Of Civil Aircraft Air Conditioning System

Aircraft air-conditioning system is one of the most important subsystem of the aircraft.But currently,the study of the air-conditioning system health monitoring,fault diagnosis and health management is not as comprehensive and extensive as the study of the other important subsystem like the engine,landing gear etc.Due to the lack of sophisticated methods of the health monitoring,the air-conditioning system failures are generally sudden failures.Cockpit panel pack light and temperature control system failures happened frequently,especially during the hot summer months,which caused the frequent unscheduled maintenance and lead to great effect on economic benefit of airlines.Furthermore,air-conditioning system fault location,isolation and eliminate of domain airlines are mainly based on the maintenance manual provided by manufacturer.However,with the huge amount of information and content,it?s quite difficult to use the manual as a guidance to solve the problem like the system fault location,isolation and eliminate,which is a major problem in civil aviation maintenance field.Therefore,it has been a main point of airlines to reduce the unscheduled maintenance of the air conditioning system and isolate the fault rapidly and efficiently.This paper takes the air conditioning system as the object of research and studies the method of air-conditioning system health monitoring and fault diagnosis based on flight record data which provided the technological support for the realization of on-condition maintenance and health management of air-conditioning system.Firstly,this paper begins with the fundamental principle and common line failure mode and develops the air-conditioning system monitoring parameter system.In this paper,we select heat exchanger of the cold system as object of research based on the investigation result and propose the concept of air-conditioning system baseline model.We have developed the heat exchanger baseline model based on QAR data of six aircraft provided by a domestic airline and verified this model.The results show that the heat exchanger baseline model based on the actual field data can strongly reflect the degree of degeneration of heat exchanger performance.The anomaly can be detected before the air conditioning system failed caused by the performance degradation of heat exchanger,which avoid the unscheduled maintenance caused by these failure and realized the on-condition maintenance of heat exchanger.Secondly,with the concept of multivariate information fusion diagnosis,this paper has developed the air conditioning system Bayesian fault diagnosis network based on the expert experience and structural learning integration and verified the model with the actual case data.The result shows that Bayesian fault diagnosis network has good fault reasoning performance which can be used for fault diagnosis and isolation as an auxiliary troubleshooting tool.This finding reduce the time for troubleshooting and improve the utilization of aircraft.Finally,combined the current airline?s demand analysis of air conditioning health management and the context of previous research,we have designed the integrated PHM model of civil air conditioning system from the perspective of actual engineering application of airlines.This model can be used to assist the health monitoring,fault prediction and auxiliary diagnosis information supply which can realize the automated system health management.