DouHe 200MW Power Plant Bolier Combustion Optimization Neural Network Modeling

As the energy and environmental situation of China is getting increasingly rigorous, Coal-Fired power plants are facing increasing pressure from energy saving and emission reduction. As its main energy consuming equipment, power plant boiler is getting more and more attention. First, the efficiency of power plant boiler in China is below the international advanced level. Secondly, in the process of peaking power, the boiler might be in low-load operation. In order to ensure the economic benefits, the research on boiler efficiency is inevitable for economic development Again, the boiler combustion produces massive amount of NOx and other pollutants, which creates enormous environmental burden. Under this background, high efficiency and low NOx emission boiler combustion optimization technology has become a hot topic for researchers.This thesis is based on an in-depth study of boiler combustion optimization technology and provides an analysis of neural network in terms of modeling for boiler combustion. It is also based on the actual work condition of the 200mw boiler in Douhe power plantand discusses our research on the application of using genetic algorithm to find parameters of optimization ofthermal powerboilercombustion. The maincontents are as follows1. A detailed analysis of boiler structure and combustion technology is provided. It is based on the original empirical methodology and an in-depth understanding of the process of combustion technology,critical factors of power plant boiler control, and its ways to impact the performance of boiler combustion process. It also determined the research topic and laid the foundation for subsequent theoretical analysis.2.Utilizing neural network modeling and genetic algorithms to find a solution and get the optimal result. Introducing the basic principles and working method for neural network model. Using the BP neural network and RBF neural network to build model for boiler combustion system and the simulation results are compared and analyzed. The output and input parameters are determined by cross study. Adopting genetic algorithms to find solution for the boiler combustion model. Using the genetic algorithms to optimize the controlled parameters of the combustion system, eventually obtained the optimal value of the parameter of improving the efficiency of combustion system and reducing, nitrogen oxide emission.3. Using MFC developed the software for optimizing the value of boiler combustion control parameter. Applying the calculation system from the software to the real-time control of the boiler. The actual results show that the algorithm optimized based on neural network model and genetic algorithm has great value in practical application. The thermo efficiency of the boiler system and NQx emission have all been optimized in this experiment, wnich brings enormous economic and environmental benefits. The last part of this thesis summarizes the deficiency of the research and presented the vision for the future development of boiler combustion technology.