The Research On Health Monitoring And Fault Diagnosis Method For Civil Aircraft Engine Starting System

The Engine Start System of Aircraft(ESS)is one of the important accessory systems for aircraft engines.Its normal operation is directly related to whether the aircraft can take off on time.At the same time,when the engine encounters an emergency in the air,the starting system also plays an indispensable role.For a long time,the starting system has not attracted much attention like the landing gear and the engine bleed air system.There are two main reasons for this: First,the starting system itself has a complicated structure,and the internal system includes other subsystems such as the ignition system and the fuel system.The second is that the starting system is greatly affected by the environment,especially in winter.The starting system often has a phenomenon that the starting valve cannot be opened.These two points cause the difficulty of monitoring the starting system.At the same time,the starting system is also one of the subsystems with high failure rate,which greatly affects the airline’s on-time flight rate and economic benefits.Since the starting system does not implement automatic monitoring,sudden failures often occur on the route.At present,the maintenance of the starting system by the airline is based on the manufacturer’s maintenance manual.If the fault occurs,the manual will not only cause the target to be unclear,the difficulty of review,the long time of fault isolation and so on.Therefore,airlines urgently need to carry out scientific research on how to reduce the number of unscheduled maintenance of engine starting systems and the diagnosis of fast fault isolation of routes.Based on this,this paper selects the starting system as the research object,carries out the research on the health monitoring and automatic fault diagnosis method of the engine starting system based on flight data,and provides certain technical support for the early realization of the visual maintenance of the engine starting system.First of all,this paper starts from the principle of the starting system and the actual maintenance record of the airline.The performance data of the starting system health monitoring,the environmental parameter system,and the set of characteristic values reflecting the health status of the running system were established,and the baseline of the health monitoring performance of the starting system was established.Subsequently,based on the Mahalanobis distance technique,the model is verified using the actual QAR data of the airline.The results show that the health baseline model of the starting system based on flight data can reflect the health status and performance degradation trend of the starting system well,thus achieving early detection of performance degradation,preventive maintenance in advance,and Reduced the number of unscheduled repairs to the starting system.Secondly,using the concept of multi-information fusion diagnosis,using the expert experience and system principle,the baseline model of random forest parameters of the starting system is established,and the fault tree of the starting system is constructed.The verification based on the actual flight data shows that the random forest fault diagnosis model can reflect the degradation trend of a certain parameter well,and the fault tree can reflect the fault reasoning performance well.The above research results provide technical assistance for airline personnel to achieve rapid troubleshooting and reduce aircraft downtime.Finally,from the perspective of airlines,this paper listens to their actual needs and designs an integrated civil aircraft starting system PHM model based on the previous method model.In this way,it can help the airline achieve the automation of the start system data reading,information storage,health monitoring and fault diagnosis.