APPLICATION OF FUZZY SPIKING NEURAL NETWORK ON FAULT DIAGNOSIS OF GAS BLOWER SETS

This paper is designed to handle thegas blower failure with regard to the fuzzy classification boundaries and the traditional neural network algorithms that are difficult to solve application problems on the contradictions between an instance of scale and the network scale. Fuzzy spiking neural network was used in the Fault Diagnosis of Gas blower sets, leading to diagnostic algorithms proposed for fuzzy spiking neural network. The first step is to use the receptive fields of neurons, the algorithm converts the input mode into the output pulse train of the neuron, and then pulse sequence with a population coding and unsupervised learning for clustering analysis, thus being able to better overcome thegas blower failure with regard to the classification bout:duly, and the related clustering analysis of invalidity. The applications showed that the algorithm effectively solves the fuzzy boundaries of thegas blower failure, greatly improving the accuracy of fault diagnosis.