Mechanical Characteristics Fault Diagnosis of Circuit Breaker Based on Quantum Genetic Neural Network and D-S Evidence Theory
Jianhua Liu, Mingping Zhou, Hou Chuanchuan · Gaoya dianqi · 2018
Focus on the present situation of the inaccurate diagnosis of the fault diagnosis system of circuit breaker,a new method of quantum genetic neural network and D-S evidence theory is proposed,using wavelet transform andspectrum analysis technology,getting the coil current wave and mechanical vibration wave of low frequency and highfrequency signal,extracting two types of band energy value as the input of the two independent quantum geneticradial basis function(RBF)neural network,and two preliminary diagnosis results are obtained,finally using D-Sevidence theory to integrate evaluation results of two neural network. The experimental results show that the RBFneural network with improved quantum genetic algorithm has a fast convergence speed and accurate result,and alsothe final result of D-S evidence fusing is more accurate,moreover,the diagnosis results are improved.