A method for using BP neural network to monitor running state of a steam turbine gearbox

The relationship between the gearbox´s running state and the characteristic parameters is complex and nonlinear. In this paper, a diagnostic method for BP neural network gear box´s running state based on principal component analysis is proposed. The method is mainly extracted from 8 main characteristic parameters and 10 groups of training samples. On this basis, the BP neural network classifier is designed, and use the network to identify steam turbine gearbox´s running state identify the operational status, so as to facilitate timely maintenance, reduce production costs and create some economic benefits.