Research On Thermal Prediction Based On Uniform Design And BP Neural Network

Our country is in a fast development of modernization, the demand for electricity grow day by day, how to coordinate economic development and environmental protection is a major task. Coal is our country’s main energy of power generation,and power plant based on coal is the main generation method in china, thus, how to deal with coal-fired power plant economic operation and the problem of environmental pollution is a mission which matters a lot.Carbon content in fly ash and nitrogen oxide emissions is two main parameters which reflects the running effect of power plant, carbon content in fly ash is an important parameter which shows the efficiency of unit, and coal-fired power generation of nitrogen oxides produce a lot environmental damage, therefore, these two would be two important parameters as the starting point, analysis and discuss how to deal with the economic efficiency and environmental protection of power plant through controlling the two parameters. By studying the protection of this two parameters, achieve economic and effective control, and efficient and environmentally operation of power plant. For the large coal-fired power plant, boiler combustion optimization and operation optimization is the important method which could improve the economics and the security of unit and to achieve energy-saving and reducing emissions of unit in order to improve the efficiency and reduce the emissions of NOx of the boiler combustion. In this paper, the network model combining uniform design and BP neural network was established, taking the actual boiler load, second wind average valve position, the coal output, first wind pressure and other12factors as input, the carbon content of fly ash and NOx emissions as the output of a power unit, to forecast the fly ash carbon content and NOx. Analysing of the prediction results proved that the present can meet the demand of power plant monitoring well in accuracy and stability.