Research On Characteristic Of NOx Emission And Combustion Monitoring Based On Image Digital Technology

For large coal-fired boilers, establishing and maintaining steady combustion is essential to ensure the safety and efficiency. Diagnosis techniques as a main and basic method for safety of plant are needed. An optical method using CCD camera and image processing system for combustion diagnosis is the new technique that is being developed along with the advance of computer science, optics technology and mathematical techniques. In this paper, the main content including the monitoring temperature field visualization and NO_x emission is as follows:(1) Flame temperature is measured in running boilers on varied operating conditions by a combustion monitoring system. The combustion monitoring system includes an optics delivering system, a CCD camera, and an image collecting system. Research on the characteristics of NO_x emission was also carried out. The purpose of this research is to found the influence of varying operating conditions and coal types on the NO_x emission.(2) Numerical simulations were conducted for two different operating cases of combustion and NO_x emission, for the number 3 coal- fired furnace at XIBAIPO power plant. The comparison between the dates from computer simulation and the real-time measurement at the power plant showed good coincidence and correlation. It is the preliminary foundation to combing diagnosis system and control system.(3) The characteristics of NO_x emission were analyzed by SVM. The results obtained by SVM have been compared with that of ANN. The effective means that??? the characteristics of NO_x emission was forecasting by SVM could reduce NO_x emission and increase boiler efficiency by adjusting combustion conditions. What's more, the influencial factors on NO_x emission were studied by means of grey relational analysis, the results can be a reference for the adjustment of combustion conditions for reducing NO_X emission (4) The digital images of flame can be classified better by the method based on fassy immune network algorithm. The eigenvectors of flame images obtained from plant trials are extracted by digital image technology and classified. Recognizing the image classification to judge whether the combustion status of flame changing or not. The results of the experiment show that the said method has very high recognition accuracy to judge changes of combustion status.