Fault Diagnosis Of High Voltage Circuit Breaker Based On D-S Evidence Theory

High-voltage circuit breaker as an important equipment in the power system,its operating state directly affects the stability of the entire power system and the reliability of power supply.Therefore,the research on fault diagnosis technology of high-voltage circuit breaker not only beneficial to ensure the stability and credibility of the power system,but also provide guarantee for the construction of smart grid in China.In practical work,due to access,transmission,external interference and loss and other reasons result in high-voltage circuit breaker fault diagnosis results are not accurate,inconsistent or even contradictory.In this paper,using fault diagnosis technology of high voltage circuit breaker based on D-S evidence theory.The final diagnostic results are obtained by making decision-level fusion of uncertain and incomplete preliminary diagnosis results.First of all,this thesis deeply studies the working principle of high-voltage circuit breaker,and analyzes the types of common faults and the causes of the faults.Then,this thesis introduces the basic theory of D-S evidence theory and its advantages in fault diagnosis,the study focus on its deficiencies,and put forward an improved method of correction of conflict evidence to its insufficiency of conflict evidence fusion.The improved method is simulated by Matlab,and compared with other improved methods to analyze its strengths and weaknesses.Finally,the method is applied to the fault diagnosis of high-voltage circuit breaker,and the final diagnosis result is obtained by Matlab algorithm simulation experiment.High-voltage circuit breaker fault signal collected in this thesis,including opening and closing coil current signal,acoustic signal and vibration signal.In the first place,feature signals are extracted from the collected fault signals,and empirical mode decomposition is performed by using Hilbert-Huang Transform.The intrinsic modal function obtained by the decomposition is taken as the characteristic information of the original signal.Then the support vector machine is used to perform feature level fusion,and the fault feature information is trained and tested to get the accuracy of each fault type identification.The feature information of the fault signal to be tested is identified by support vector machine,and the recognition results are weighted by SVM test accuracy to get the basic reliability of each fault type.Finally,the basic reliability of each fault is fused with the improved D-S evidence theory to get the final fault diagnosis result of the high-voltage circuit breaker.