Abstract: Boiler as the pivotal equipment of burning, its safe operation is essential, because boiler has complex structure, damage, wear, acid gas corrosion and improper operation will cause malfunctions. In order to avoid failure, this paper combines wavelet transform and neural network to constitute wavelet neural network and applies it to boiler fault diagnosis. Experiment results show that the wavelet neural network fully inherits the advantages of wavelet transform and neural network, this method has better fault diagnostic capabilities, the fault diagnosis accuracy is obvious better than BP neural network.


Key words: boiler, wavelet neural network, fault diagnosis, BP neural network