Research On Fault Diagnosis Of Hydraulic Pipe Based On Optimized VMD And BP Neural Network

The hydraulic pipeline system of aeroengine is one of the most important accessory systems of aeroengine.Because the hydraulic pipeline system of aeroengine is in the extremely harsh environment of high temperature,high pressure and high power,the vibration of hydraulic pipeline is an essential problem Obstacle failure has always been one of the main problems affecting the safety of the hydraulic pipeline system,because the failure of the hydraulic pipeline in the engine is easy to cause the damage of the hydraulic pipeline system and even the whole engine can not work,which is easy to produce a series of safety problems.At the same time,due to the complex structure of hydraulic pipeline and the influence of pump source excitation,the hydraulic pipeline fault diagnosis process is easy to be disturbed by strong noise,and the accuracy of fault diagnosis results is not ideal.Therefore,this paper proposes a fault diagnosis method of aeroengine hydraulic pipeline based on optimized VMD and BP neural network.Firstly,the genetic algorithm is used to optimize the parameters of VMD modal component K value and penalty factor alpha value,so that it can determine the optimal parameters adaptively,avoid over decomposition or under decomposition of decomposition results,and reduce the subjective error of human factors The best component of hydraulic pipeline crack fault and pit fault is extracted by comparing and analyzing the fault state and health state of hydraulic pipeline,which provides a certain basis for pattern recognition of BP neural network model.Secondly,using VMD method,the best component obtained from the hydraulic pipeline processing is used as the input of BP neural network,and input into the BP neural network model for training;the results show that the hydraulic pipeline fault diagnosis method based on GA optimization VMD and BP neural network realizes the accurate identification of hydraulic pipeline crack and pit fault,and verifies the VMD-BP method in aviation engine The recognition rate can reach 99.4% in the actual fault signal of hydraulic pipeline.Finally,based on the same aviation hydraulic pipeline vibration data set,the proposed VMD-BP neural network fault diagnosis method is compared with EMD-BP neural network,BP neural network and support vector machine fault diagnosis methods,the diagnosis performance of VMD-BP neural network is better than other intelligent diagnosis methods,and it also shows that the fault diagnosis method of optimized VMD-BP neural network proposed in this paper realizes the fault diagnosis of hydraulic pipeline The intelligent classification of pressure pipeline crack fault signal and pipeline pit fault signal provides a certain reference for the intelligent fault diagnosis of aeroengine hydraulic pipeline.