Analysis Of Aircraft Secondary Power Failure Mode And Fault Feature Extraction

In multiple systems of the more electric aircraft,electrical energy has been used to replace hydraulic,pneumatic,and mechanical energy.As a result,the aircraft's power supply capacity has been increased significantly and the number of on-board electrical equipment is numerous.In order to ensure normal operation of aircraft systems,a highly reliable power system must be equipped.In addition to the main power supply,the secondary power supply is also an important part in the aircraft power supply system.Its operation not only affects normal operation of aircraft critical loads,but also has an important impact on reliability and stability of the aircraft power grid.The core device of the secondary power supply is the power semiconductor device.Due to the harsh working environment and complex working conditions on the aircraft,the reliability of the power semiconductor device which is core device of the secondary power supply has also been severely tested.Therefore,from the perspective of power semiconductor devices,the failure mode of the aircraft secondary power supply is analyzed,and the failure characteristics are extracted to achieve rapid fault diagnosis,which is of great significance for improving aircraft power grid reliability and ensuring aircraft airworthiness.In the aircraft power grid,the voltage size or form of the main power supply is transformed mainly by the secondary power supply.There are many types of secondary power sources in more electric aircraft.Among them,three types of secondary power sources have been studied,including transformer rectifier units,DC converters and inverters.First,three types of simulation models of secondary power supply are established,and then its failure modes are analyzed,including several failure modes such as open circuit,short circuit and the power device parameter drift.By studying impact of different fault conditions on power grid stability in different modes and the main factors of input and output characteristics changes,relevant data and conclusions are obtained.Finally,taking the aerospace transformer rectifier unit as the research object,a fault diagnosis method combining the stacked noise reduction autoencoder and particle swarm optimization support vector is proposed.Among them,the stacked noise reduction autoencoder method is used to extract the fault feature of the transformer rectifier unit fault signals,and the particle swarm optimization support vector machine method is used to process the feature vector to realize the location diagnosis of the fault diode,and the diagnosis result is traditional The fault diagnosis method was compared and analyzed.Theoretical analysis shows that the failure of power semiconductor devices is affected by many factors.During the gradual process of power devices from degradation to failure,device losses,device temperatures,and on-voltages all increase to a certain extent.The study also found that when the secondary power supply fails,the input and output characteristics of the secondary power supply are affected mainly by the number and location of faulty power devices.The stack noise reduction autoencoder method can better extract faults from the aircraft's secondary power supply,and the fault diagnosis combined with particle swarm optimization support vector machine has a higher fault diagnosis accuracy rate.