The Technology And Application Of Furnace Combustion Intelligent Diagnosis

Combustion diagnosis of boiler is most important in the whole power station's daily running. During the course of boiler combustion, there are a lot of complex chemical and physical reactions. The traditional flame diagnostic system could not resolve the problems about the diagnosis's accuracy and the system's self-adaptation. To resolve this question, people make use of computer image processing to deal with flame images and to judge the burners' burning states. By analyzing the coal flame digital images, a new set of Boiler Burners' Intelligent Diagnostic System is constructed, which is based on the Neural Networks. Firstly, it analyses the coal flame image's appearance and characteristics and selects several eigenvalues to make up of inputs of diagnostic system. Secondly, after particular description of the principle and the construction and the training ideas of the diagnostic system, the imitational results of the system are presented by testing data. Then, the neural network is adjusted for self-adaptation. In the 4th chapter the questions of NOX emissions are detailed investigated and the creation mechanism is analyzed particularly considering the influence of temperatures. And the demarcate experiment of temperatures field is introduced in this chapter. The system's software modules and their functions are described in the 5th chapter. Finally, the author puts forward some problems occurred in the system's realization and propositional solutions.