Research On Fault Diagnosis Of Low Voltage Circuit Breaker Based On VMD Parameter Optimization And Sample Entropy

As a device to protect AC and DC electrical appliances,low-voltage circuit breakers are widely used in the power distribution system of airports,factories,apartments,shopping malls and residential houses.They are the basis for ensuring the safety of users’ electricity use and it is one of the important protection and control components in low voltage electrical appliances.Therefore,the research on fault diagnosis of low-voltage circuit breakers has important research significance and engineering practical value.The main faults of low-voltage circuit breakers include electrical faults and mechanical faults.Studies have shown that 70% to 80% of faults are mechanical faults.Therefore,this article mainly studies the diagnosis of mechanical faults in low-voltage circuit breakers.1)Aiming at some of the shortcomings of traditional time-frequency signal analysis methods,a signal analysis method based on Variational Mode Decomposition is studied.By constructing simulation signals,the best mode number K and penalty factor α in VMD are analyzed and studied.The importance of parameters.And by constructing the simulation signal,the VMD method is compared with the empirical mode decomposition and the collective empirical mode decomposition method to reflect the superiority of VMD.2)Based on the selection of the optimal number of modes K and penalty factor α in VMD,the method of searching for the optimal values of parameters through particle swarm optimization is studied to find the optimal parameters of VMD,which avoids the complexity of selecting parameters by manual repeated experiments and reduces errors,Reduce the workload of the method and improve the efficiency.In addition,because the sample entropy has the characteristics of simple calculation,strong antiinterference performance,and insensitive to missing data,the concept of sample entropy is introduced,and the sample entropy is calculated for each component after VMD decomposition.By combining the sample entropy of each component into a feature vector,As the characteristic quantity of the fault signal.This method helps to improve the efficiency and accuracy of circuit breaker fault diagnosis.3)Collection and processing of experimental data.In the case of artificial setting of the universal circuit breaker in normal state,three-phase different periods,motor failure,loose rivets and gear oil shortage,the signal acquisition platform is used to obtain the vibration signal generated during the opening and closing of the circuit breaker.Research and perform data processing on the collected fault signals.First,perform the variational modal decomposition of the collected signals,and then calculate the sample entropy of all the signal components generated by the decomposition,and compose the feature vector to facilitate subsequent research.4)Introduce the basic principles of BP neural network and Support Vector Machine.Aiming at the normal state of universal circuit breaker,three-phase different period,motor failure,loose rivet and gear oil shortage,the previously processed data sample is used as the research object,It is trained and tested by SVM,and the data is also input to the BP neural network.Compared the two,the experiment proves that the SVM effect is better.It also reflects the feasibility and superiority of the fault diagnosis method combining VMD,sample entropy and SVM.