Thermal Power Plant Boiler Combustion Optimization Of Key Technologies

Combustion process in boiler is physically and chemically complicated reactions, and no method available at present can be employed to analyze the whole combustion process fully in terms of the mechanism of boiler combustion. In this thesis, model prediction technique is found to be an effective way to solve the problem through analyses of some major combustion optimization techniques at home and abroad. Neural network is the most commonly used model prediction technique which views boiler combustion process as a black-box operation process. By training neural network with mass historical data and adjusting model structure, the model will finally acquire the function of mimicing boiler combustion process. On the basis of neural network, boiler can operate in the optimization interval by adjusting boiler operational parameters through multi-goal optimization approach, realizing that boiler can operate economically and safely without polluting the environment. Based on the approach, Several mature closed loop control systems of boiler combustion performance optimum have been developed in some European and American countries, which were successfully carried out.This thesis studies two key techniques for boiler combustion optimization system—model prediction technique and Best-First search Algorithm. The thesis falls into four chapters. In chapter l,the necessity and feasibility of implementation of closed loop control system of boiler combustion performance optimum for thermal power plant is confirmed through a review of combustion optimization, and the goal for adopting closed loop control system of boiler combustion performance optimum based on the two techniques is set up by comparing some major combustion optimization techniques. In chapter 2, taking boiler #1 of a certain thermal power plant in Zhejiang province as an example, the application of neural network model prediction technique is studied. On the basis of chapter 3 and chapter 4, a boiler combustion optimization model is set up, and solutions to Best-First search employing genetic algorithms are studied. In chapter 4, Software Architecture of boiler combustion optimization system are discussed by referring to the overseas advanced technology, and the effect of the technology applied is gained by on-the-spot investigation.