Research On Vibration Signal Analysis And Fault Diagnosis Method Of SF₆ High Voltage Circuit Breaker

High voltage circuit breaker serves as a device that can play a wide range of safety protection function in all kinds of power systems,which is of great significance to its fault diagnosis research.In recent years,sulfur hexafluoride（SF₆）high-voltage circuit breakers have gradually become the mainstream choice in the high-voltage circuit breaker market,because the gas used in its arc extinguishing medium has excellent arc extinguishing performance and is widely favored.In the power system,the role of high-voltage circuit breakers is very critical.In the power system,once the high voltage circuit breaker fails,it may cause incalculable consequences to the power system.Therefore,the fault research of high voltage circuit breaker has always been an important topic in this field.Most of the faults of high voltage circuit breakers are mechanical faults.The research object of this paper is the operating mechanism of LW9-72.5k V high voltage circuit breaker.The vibration data under three different fault conditions are analyzed.Through the feature extraction of vibration signals and the construction of diagnostic models,different circuit breaker fault states are successfully diagnosed.The main contents of this paper are as follows:（1）This paper analyzes the research background and significance of the paper,analyzes the market prospect of the sulphur hexafluoride high-voltage circuit breaker,and summarizes the research status of the high voltage circuit breaker in recent years.（2）This paper analyzes the basic structure and principle of sulphur hexafluoride high-voltage circuit breaker,summarizes the common fault types and research objects,and finally three different types of mechanical fault data are described.（3）The original vibration signal data is preprocessed.The improved threshold wavelet denoising method is used to remove the interference signal.The signal is decomposed into multiple modal functions using the parameters in the variational modal decomposition optimized by the genetic algorithm.The modal function with high correlation is extracted to reconstruct the vibration signal,which is prepared for the next step of fault feature extraction.（4）The vibration signal is reconstructed using the selected intrinsic mode function,from which three different types of fault features form a composite feature set for circuit breaker state diagnosis.The three feature quantities are time domain features,power spectrum features,and nonlinear entropy features.Among them,17 time domain features are reduced to 4 by mutual information algorithm.The power spectrum line is drawn by two methods:segmented periodogram method and MTM method.The nonlinear entropy is refined composite multiscale sample entropy and refined composite multiscale dispersion entropy.（5）The fault diagnosis model of the least squares support vector machine based on the improved squirrel search algorithm is constructed.Three different strategies are used to improve the squirrel algorithm,and the effectiveness of the improved algorithm is verified by CEC benchmark function.The modified algorithm is then used to optimize the parameters of the least squares vector machine.Finally,the composite fault feature set is input into the model for circuit breaker state diagnosis.The experimental results show that the model can accurately judge the fault state of the circuit breaker.