Research On Fault Diagnosis Technology Of Scroll Compressor Based On Variational Mode Decomposition

Scroll compressor is a key mechanical machine which is widely used in industrial manufacture and daily life.Once the machines have abnormal or breakdown during running,it will directly affect the normal operation of the entire mechanical system,resulting in a huge economy loss,even endangering personal safety.Therefore,the in-depth study of scroll compressor condition monitoring and fault diagnosis technology will have important practical significance for making sure the machine smoothly running.This paper takes scroll compressor as the research object,basing on the study of variational modal decomposition,then combines with singular value decomposition,1.5-dimensional energy spectrum,multi-scale permutation entropy,and support vector machine to achieve the fault diagnosis,and establishes a fault diagnosis system through Lab VIEW platform.The main research contents of this article are as follows:1.In order to solve the problems of mode mixing and lack of theoretical basis in the traditional adaptive decomposition method,the variational mode decomposition(VMD)was introduced into the fault feature extraction and deeply study and analyze the factors affecting the performance of VMD decomposition.Aiming at the drawbacks of the two parameters of the decomposition layer and the penalty factor affecting the signal decomposition effect,the energy proportion method and envelope entropy optimization VMD are proposed to determine the best decomposition parameters,and the effectiveness of the improved method is verified by simulation signals.2.Aiming at the problem that the signal components of the scroll compressor are complex and it is difficult to effectively extract the fault features under the background of strong noise,a fault diagnosis method based on the combination of variational modal decomposition,singular value decomposition,and 1.5-dimensional energy spectrum is proposed.This method first performs an improved variational modal decomposition of the fault signal,uses kurtosis and correlation to screen the optimal component;then combines it with the singular value decomposition,and uses the singular value energy standard spectrum to determine the signal reconstruction order to restore the signal.Improve the signal-to-noise ratio;finally,perform 1.5-dimensional energy spectrum demodulation on the noise-reduction signal to extract the fault characteristics.The proposal is applied to the simulated and measure signals,and compared with other methods.The results show that this proposal has better performance in noise reduction and feature extraction.3.For the problems of state identification of scroll compressor under actual complex working conditions,a fault recognition method based on variational modal decomposition-based multiscale permutation entropy and parameter optimization support vector machine(SVM)is studied.This method first arranges the components of the entropy quantization variational modal decomposition in multi-scale to construct the feature vector;then applies the new sparrow search optimization theory(SSA)to the SVM parameter optimization process to establish the SSA-SVM model;finally,the feature vector is input Go to SSA-SVM to identify the state of the scroll compressor.The verification of examples shows that the fault diagnosis effect of this method is good,it can effectively distinguish the different fault types of scroll compressors,and has high recognition accuracy.4.Combined with the research of theoretical methods,a fault diagnosis system for scroll compressors based on Lab VIEW is developed to realize the functions of on-line monitoring and off-line diagnosis of scroll compressors,and the system is applied to actual projects.From the field test results,the functional modules of the system can effectively diagnose the operation status and fault types of the compressor,which achieve the expected goal.