Pressure Difference Prediction of Air Preheater in Coal-Fired Power Plant Based on BP Neural Network and Support Vector Regression

Abstract

The high pressure difference between the inlet and outlet of the air preheater in coal-fired power plants has always been a problem that plagues the operation of the power plant. If the pressure difference of the air preheater can be predicted in advance and adjusted in time, it will be beneficial to the safe operation of the power plant. Based on the distributed control system (DCS) big data of a 660 MW coal-fired boiler, the BP neural network and support vector regression (SVR) methods were used to model the pressure difference between the inlet and outlet of air preheater in the power plant.Comparing the prediction results of the two models, it is found that the BP neural network model is more suitable for the prediction of the pressure difference between the inlet and outlet of the air preheater under the background of big data. Aiming at the inherent shortcomings of BP neural network model such as local optimization and slow convergence speed, the particle swarm optimization (PSO) was used to improve it, and a PSO-BP neural network model was proposed. The results show that the PSO-BP neural network model has the best comprehensive performance, the highest accuracy in predicting the change in the pressure difference between the inlet and outlet of the air preheater, and the strongest generalization ability.

Keywords

• thermal power generation
• air preheater
• bp neural network
• support vector regression (svr)
• particle swarm optimization (pso)