Research On Fault Diagnosis Of Hydraulic System Based On CEEMDAN And Support Vector Machine

Due to the hydraulic system of the tamping car has a high working intensity and a high load,if the fault cannot be solved in time when the rail is repaired,it will affect the smooth operation of other trains.In order to solve the problems in the practical application of the hydraulic system of the tamping truck,this paper studies the noise reduction method of the fault vibration signal and the method of the fault feature extraction.The classifier is also studied.Support vector machine is used as the classifier to diagnose and identify the fault signal.At the same time,the improved particle swarm optimization algorithm is used to optimize the parameters of the support vector machine to improve the classification accuracy.This article introduces the EMD noise reduction method,EEMD noise reduction method,CEEMD noise reduction method,wavelet packet noise reduction method and CEEMDAN noise reduction method,and proposes an improved wavelet packet combined with CEEMDAN threshold noise reduction method.And on this basis,the fault feature extraction method of the hydraulic system of the tamping truck based on the combination of the total empirical mode decomposition CEEMDAN and fuzzy entropy is studied.The collected signal has a lot of noise.The signal is first denoised by the wavelet packet method,and then the signal is reconstructed to reduce the noise content.The reconstructed signal is decomposed by the CEEMDAN method to obtain the intrinsic modal component IMF.Secondly,according to the continuous mean square error standard,the signal-dominated and noise-dominated IMFs are respectively used to denoise with different improved threshold methods.Then,the respective processed IMF components are combined to reconstruct a useful signal with high signal-to-noise ratio.Finally,the IMF component of the final denoising fault signal is obtained by using adaptive total empirical mode decomposition,and the fuzzy entropy algorithm is used to calculate the fuzzy entropy of the first few main IMF components to form the fault feature vector.The experimental results show that the improved wavelet packet combined with CEEMDAN's threshold noise reduction method can achieve better noise reduction effects.This paper studies the fault classification model and method of the hydraulic system of the tamping vehicle based on the support vector machine SVM.At the same time,the improved particle swarm optimization algorithm IPSO is used to optimize the model parameters of the support vector machine to improve the diagnosis and recognition ability of the classifier.Then,the improved particle swarm optimization algorithm is selected to select the optimal parameters for the SVM to construct the SVM optimal fault diagnosis model,and the classification model is applied to the fault diagnosis of the hydraulic system.Experiments have verified the effective improvement of its classification performance.