Research On Mechanical Fault Diagnosis Of High Voltage Circuit Breaker Based On Improved SVDD Algorithm

The power system is the core of the national energy system,and its safe operation is related to the national strategic security.As one of the important electrical equipment in the power system,the high-voltage circuit breaker has an important impact on the safe and reliable operation of the power system.Therefore,how to diagnose the fault of high-voltage circuit breaker efficiently and accurately is one of the important problems in power system security research.Most of the faults of high-voltage circuit breakers are mechanical faults.The vibration signal can reflect the operation state of high-voltage circuit breakers comprehensively The research on mechanical fault diagnosis of high-voltage circuit breakers based on the vibration signal has become the main direction.In this paper,through the research of feature extraction method and classification recognition algorithm for the vibration signal of high-voltage circuit breaker,the mechanical fault of high-voltage circuit breaker can be identified and diagnosed quickly and accurately.First of all,based on the research of relevant diagnosis technology at home and abroad,this paper introduces the basic structure and common fault mechanism of LW9-72.5 type high-voltage circuit breaker.Based on this,the vibration signal of high-voltage circuit breaker is collected by using appropriate acceleration sensor and signal acquisition device,and the wavelet packet decomposition transformation method is used The entropy value and relative energy value of wavelet packet nodes are extracted as the feature vector of fault diagnosis.Then,the kernel entropy component analysis method is used to reduce the dimension of the high-dimensional feature data set,aiming at the problem that the high-dimensional feature data set brings a lot of redundant information and affects the classification effect.This method can maintain the overall popular structure of the original data and make it have certain distinguishing ability,Finally,an improved support vector data description algorithm is proposed to build a high-voltage circuit breaker mechanical fault diagnosis classifier.Based on the traditional support vector data description algorithm,a penalty factor is added to the algorithm,so that it can consider the influence of two kinds of samples on the classification boundary at the same time,and make the classification interface formed by the classification algorithm close to the middle of the two kinds of samples,which greatly enhances the generalization performance.The simulation results of the mechanical fault diagnosis method of high voltage circuit breaker based on the improved support vector data description algorithm show that the method can overcome the problems of the traditional method,such as too much dependence on training samples,insufficient detection range,and the inability to effectively identify the new type of fault,so as to realize the efficient classification and identification of the operation condition and fault type of high voltage circuit breaker,and effectively improve the high performance The accuracy of mechanical fault diagnosis ensures the normal operation of the equipment and power system.This method is suitable for the production practice of high voltage circuit breaker fault diagnosis.