Abstract
Based on the characteristic of turbine-generator set vibration fault diagnosis, a fault diagnosis model with artificial neural network (ANN) is proposed. Firstly, the shortcoming of conventional back-propagation (BP) network is analyzed, and a new fast BP algorithm is presented. Through setting up neuron error threshold function, only when neuron's error is bigger than the error threshold, the neuron's parameters can be adjusted, the algorithm can avoid the excessive learning of neural network and learning error oscillation, and increase BP network learning speed. Secondly, the model of turbine-generator set vibration fault diagnosis by the fast BP algorithm is set up. The results of verification show that the model has faster speed and higher diagnosis precision.
