Research On Fault Diagnosis Methods For Aircraft Electromechanical Systems

With the continuous development of the aviation manufacturing industry,the structure of aircraft electromechanical systems is becoming increasingly complex and precise.Relying on ground maintenance personnel for maintenance and traditional expert system diagnosis methods,it can no longer meet the needs of electromechanical system fault diagnosis.Manual troubleshooting has low efficiency,high cost,and is easily influenced by subjective human factors.Expert systems rely on expert experience and knowledge,lack self-learning ability,and are difficult to cope with new faults.In recent years,machine learning technology has developed rapidly and achieved good results in the field of fault diagnosis.Based on the research and analysis of common fault signals of aircraft electromechanical systems,this paper explores the application of machine learning algorithms in the fault diagnosis of aircraft electromechanical systems,and deeply studies the integrated learning diagnosis algorithm.The main work is as follows:1.Researched fault diagnosis methods based on machine learning,in-depth research on convolutional neural network CNN,recurrent neural network RNN,research on ensemble learning algorithm(Ensemble Learning)and its base classifier combination strategy.Design and build a CNN single classifier diagnostic model and a GRU single classifier diagnostic model,and summarize and analyze the advantages and disadvantages of the above two neural network diagnostic models.2.Aiming at the shortcomings of single classifier model diagnosis,a fault diagnosis method for aircraft electromechanical system based on ensemble learning is proposed.CEEMDAN-PCA is used to preprocess the fault data,and then input into XGBoost model to complete fault diagnosis,which improves the accuracy of fault diagnosis of electromechanical system.And reduce the false alarm rate and missed judgment rate of diagnosis.3.The machine learning diagnostic model is used in the design and implementation of fault diagnosis software.The software provides CNN fault diagnosis model,GRU fault diagnosis model,CPXbt fault diagnosis model.To verify whether the software performance test meets the actual diagnosis requirements,it proves that the aircraft electromechanical system fault diagnosis software can have a good diagnosis effect.In summary,this paper studies the application of machine learning algorithm on aircraft machinery system fault diagnosis,and is configured to realize auxiliary decisionmaking for the maintenance of ground maintenance personnel to improve accessibility efficiency.