Research On Operating State Identification Of AC Contactor Contact System Based On Vibration Signal

In recent years,with the increasing growth of the national economy,in order to better promote the development of China’s smart grid and accelerate the goal of carbon peak and carbon neutralization,higher-level requirements are also put forward for switching appliances to ensure the safety and reliability of power transmission and distribution.How to improve the operation status information feedback ability of switching appliances is a hot issue in recent years.As an electrical equipment in low-voltage electrical appliances,AC contactor can be used for long-distance frequent connection and disconnection of lines and overload protection of motor starting circuit.It has a far-reaching impact on the development of its intelligence by deeply excavating the state information in its operation process,judging the degraded state of products and establishing a classification model for judging the operation state.In this thesis,the mechanical vibration signal of AC contactor contact system is taken as the main research object to extract the characteristic parameters.According to the characteristics of AC contactor with multivariate,multi-attribute interaction and dynamic change,the optimal characteristic parameter subset with high correlation with its operation state is obtained through the feature selection method based on maximum information coefficient,And input it into support vector machines to establish the state recognition model of AC contactor.Firstly,the mechanical structure of the AC contactor is studied,and the action process and failure mechanism of the contact are deeply analyzed.With the help of the AC contactor operation state research system,taking CJX2-5011 as the research object,the vibration signal,voltage and current signal of the contact system in the whole life process are obtained,Comprehensively calculate and mine the characteristic parameters in time domain and frequency domain of vibration signals used to characterize and identify the operating state of AC contactor contact system.Secondly,in order to retain the key information contained in the extracted feature parameters to the greatest extent,the development of feature selection strategy revolves around MIC method.Considering the importance and redundancy of the parameters,the dimension of the extracted original feature parameter set is reduced to obtain the optimal feature parameter subset.Finally,the optimal characteristic parameters are combined with the weighted relative deterioration method to characterize the operating state of AC contactor,and its operating state is divided.Then it is input into the GWO-SVM to construct the operation state recognition model of AC contactor.By comparing the accuracy of parameter subsets obtained by different selection methods for characterization and the accuracy of operation state recognition,it is concluded that the representation and recognition of the optimal feature subsets obtained by MIC method have better results than other methods.