Research on mathematical model of electrode boiler based on neural network

Abstract

Aiming at the difficulty of controlling the peaking control of the electrode boilers in the “Three North” area (northeast, northwest and north) in China, the power supply of the electrode boiler with BP neural network and Bayesian regularized neural network is put forward Switch active power and main steam temperature modeling prediction method. Based on a 20MW electrode boiler in a power plant in northeast China firstly, the input function of the network was determined by the correlation function method. Then, the BP neural network and the Bayesian neural network were used to establish the electrod boiler model. Finally, the model prediction results were obtained. The results show that the error of Bayesian neural network is smaller than that of BP neural network, and it is more suitable for the peak-load control of electrode type boiler.