The Target Value Optimization Of Operating Parameters Of600MW Supercritical Units

Compared with developed country, the standard coal consumption rate of powersupply in China is much higher, which means a huge potential for energy saving. A betterapproach to optimize operation of thermal power units can reduce coal consumption andimprove the environment. In this paper, main operating parameters of a600MW super-critical unit is optimized to improve thermal economy based on association rule-based datamining techniques and neural network algorithm.Different analysis and optimization methods of thermodynamic system are discussed.Combined with the characteristics of the research object, association rules and neuralnetwork algorithm are selected to improve the performance. According to experience andtheory, operating parameters of supercritical boiler combustion system and steamcirculation system are analyzed, and the main operating parameters which impact thermaleconomic of the units are selected. Then the association rules algorithm is applied todetermining the input and output parameters of the neural network model.Optimization models of the boiler side and turbine side in supercritical unit aredeveloped based on BP neural network algorithm. Then input and output parameters of theboiler side and turbine side are optimized and coupled. The results show that the relativeprediction error of unit power and main steam in boiler side model is less than1.5%and2%, respectively. The relative prediction error of standard coal consumption rate of powersupply in turbine side model is less than1%.The prediction of optimized target values of main operating parameters in power plantmeet the engineering requirements, and GUI interface developed in this paper can guideoperating personnel scheduling optimization.