Research On Fault Diagnosis Of S700K Switch Machine Based On EEMD Multi-scale Fuzzy Entropy

As one of three outdoor railway signal equipments,the switch machine is used to convert turnout.It locks turnout and gives right position when the turnout is in place.It has an important impact on the safety and efficiency of railway transportation.At present,China railway signal system guarantees switch machine to operate safely and reliably by means of periodic maintenance and the maintenance of sunroof points.This maintenance method is inefficient,the workload of workers is heavy.Also it maybe cause misjudgments and missed judgments that only depends on the experience of workers to identify fault type.With the development of China railway towards high-speed and heavy-load operation,the S700 K AC electric switch machine is widely used in current railway lines.In order to improve the safety and reliability of its work,it is inevitable that the intelligent data analysis and processing technology is applied to realize the intelligent perception and safety early warning of the working states of S700 K AC electric switch machine.In view of the problems raised above,the main research contents of this article are as follows:Firstly,the change of output power of S700 K AC electric switch machine can reflect the change of push-pull force of switch rail of turnout,and the push-pull force of switch rail can reflect the internal operating states of the switch machine.This article takes S700 K AC electric switch machine as object,proposes a fault feature extraction method based on EEMD(Ensemble Empirical Mode Decomposition)multi-scale fuzzy entropy.The fault diagnosis method applies the theory of ensemble empirical mode decomposition,the fuzzy entropy and the parameter normalization knowledge.The power curve of switch machine is adaptively decomposed into a series of intrinsic mode function from high frequency to low frequency.Then the fuzzy entropy on each IMF is extracted and regarded as the characteristic parameters to establish the feature set of fault diagnosis of S700 K electric switch machine.Secondly,this article uses the grey correlation degree algorithm and fuzzy clustering method to identify fault type of S700 K electric switch machine according to the fault characteristic.The grey correlation degree algorithm firstly calculates correlation degree values between the curves to be tested and the each fault curve.Then choosing reasonable resolution coefficient tests the function of fault diagnosis system based on the resolution value principle and the accuracy of fault diagnosis exists difference under different resolution coefficient.The fault diagnosis based on the fuzzy clustering method,firstly the original feature pattern matrix is normalized by the translation standard deviation transformation and the translation range transformation.Then the index of describing the similarity between samples is introduced to establish the fuzzy similarity matrix of the original feature pattern matrix.Lastly the fuzzy equivalent matrix is constructed by the transfer closure method and the dynamic clustering graph is formed for fault identification.In summary,using the fault power curve of S700 K electric switch machine in a certain electricity section as test samples verify the function of fault diagnosis system.The experimental results verify the feasibility of two fault diagnosis methods applied to fault diagnosis of S700 K electric switch machine.