Multi-feature Analysis And Recognition Technology Of Aviation AC Fault Arc

In recent years,the aviation safety problem has received more and more attention,One of the main inducing factors is arc fault in AC wiring harness.The insulating material of the aircraft cable may be physically damaged by friction,stresses,foreign matter.At the same time,temperature,humidity,chemical erosion and other reasons will lead to aging decomposition of the insulation layer.The insulation performance become weak,and the possibility of electric breakdown increased.Then intermittent arc is formed on cable surface.In recent years,with the further development of the capacity of power distribution system,the wiring quantity is increasing,and the complexity of the cables are followed.With such electrical environment,aviation electrical safety issues caused by arc occurred frequently.And the hidden nature of the arc fault makes it difficult to be detected.Aviation safety issues become more serious.Therefore,it is of great significance to study the characteristics of the arc fault and determine an effective method to identify the arc fault of aviation AC wiring harness.In this thesis,the characteristic of current is extracted from different angles and the method of identifying AC fault arc is studied.First of all,according to the standard of aviation fault simulation test to experiment,obtain the analysis of the needs of different loads under the normal current data and fault arc current data,for feature extraction preparation;Secondly,arc current and normal current were comparative analyzed separately from three different aspects,Current waveform difference,signal modeling and spectrum characteristics.The results showed that,the arc current was not exactly the same as the normal current in these three aspects,but the difference between them was influenced by the load type,the line current level,the power frequency and so on.In order to verify the stability of the feature extraction method,analysis the changes of current frequency and sampling speed;Finally,in order to solve the problem of arc fault in different working conditions,a multi-feature fusion arc fault recognition technique was proposed.Constructing multifeature data matrix,And the extreme learning machine is introduced.The fault arc is classified and identified,and analyzed the performance of the classification model.The data analysis results show that the model has high accuracy,can provide a reliable reference for identification of aviation AC arc fault.