Research On Fault Diagnosis Of High Voltage Circuit Breaker Based On Compressed Sensing Theory

High voltage circuit breaker is one of the key equipment in the power supply system.It plays the dual functions of control and protection in the whole power supply network.In the normal working time,the high-voltage circuit breaker can cut off the working voltage,and can also cut off the short-circuit current and overload current within the specified time limit,so as to ensure the safe operation of the power grid.Therefore,ensuring the smooth operation of high-voltage circuit breaker is very important to improve the reliability of power grid.The opening and closing coil in the high-voltage circuit breaker is one of the key components of the high-voltage circuit breaker.59.9% of the faults of the circuit breaker are the faults of its mechanism and auxiliary circuit.The traditional Shannon Nyquist theory fault detection method has the problems of low fault detection rate and high signal sampling rate,while the compressed sensing method uses the sparse characteristics of the signal to collect and compress the data at a sampling speed far less than Nyquist sampling theorem,which reduces the hardware pressure on data acquisition,transmission and storage.Therefore,based on the compressed sensing theory,this thesis analyzes the opening and closing coil of high-voltage circuit breaker,and takes the collected current signal as fault data for fault diagnosis.Firstly,the research status of fault diagnosis of high voltage circuit breaker is briefly introduced,and the relevant theories and fault types of opening and closing coil current are introduced in detail.According to the working principle of the opening and closing coil of high-voltage circuit breaker,the simulation model of the opening and closing coil is established in MATLAB/Simulink,the current signal is obtained through the simulation model,and the fault data is obtained by modifying the relevant parameters of the opening and closing coil current model.Secondly,the compressed sensing theory is studied,the sparsity of the original signal is analyzed,the discrete Fourier transform is trained as an initialization dictionary,and the trained learning dictionary is used to sparse represent the original signal.Since SVD can increase the minimum singular value of the measurement matrix,SVD is applied to the optimization algorithm of the measurement matrix to improve the accuracy of signal reconstruction.Finally,two kinds of reconstruction algorithms are studied.Based on the orthogonal matching pursuit algorithm,an improved cosine generalized regularized orthogonal matching pursuit algorithm is proposed.The algorithm further determines the preselected atoms of regularized secondary selection through cosine similarity coefficient to obtain the optimal atomic support set.Through experiments,it is concluded that the reconstruction algorithm proposed in this paper has better reconstruction performance.Finally,support vector machine(SVM)classification model is introduced into the fault diagnosis of high-voltage circuit breaker.In order to improve the diagnosis accuracy,PSO-SVM fault diagnosis model is obtained by optimizing particle swarm optimization(PSO).Then,the compressed sensing theory is used to process the fault current signal of the opening and closing coil current of the high-voltage circuit breaker.In the process of reconstructing the measured value,the fault characteristic parameters are obtained,and the extracted fault characteristic parameters are input into the support vector machine optimized by particle swarm optimization algorithm to realize the fault diagnosis of the high-voltage circuit breaker.Experiments show that compared with the traditional methods,the fault diagnosis accuracy of this method is improved and the fault diagnosis speed is accelerated.