Study Of Stastion Boiler Combustion Optimization Based On Improved K-Meansclurstering Algorithm

With economic growing speedily in these years, the energy demand has increased dramatically and pollution problem has emerged. The worsening environment urges us to use resources more rationally and protect the environment. As the major source of pollution, power station boiler has attracted many attentions and the optimization of boiler combustion system has become one of the key study objects in the world.An improved clustering method which was used to optimize the boiler combustion was proposed in this paper, based on the research of a thermal station’s 300 MW Circulating fluidized bed boiler, aimed to improve the thermal efficiency of boiler combustion and reduce the emission of NOX. In the first place, the method of boiler combustion efficiency computation and the forming mechanism of NOX were analyzed. Moreover, an advanced K-means clustering algorithm called K-medoids algorithm was formed and a K-BBO-CLUSTER algorithm based on the combination of BBO and K-medoids clustering algorithm was proposed in this paper. Then, exchange the adjustable parameter to optimize the boiler combustion system. In order to verify the feasibility of this method, a Bayesian based least squares support vector machine was used as a test model in this paper, and 720 sets of optimized operation data collected from a CFBB were used for experiment. The results of boiler combustion efficiency and the emission of NOX was satisfied, therefore proved the accuracy and reliability of this method.The experiment results indicated that the advanced clustering algorithm that proposed in this paper could effectively improve the efficiency of boiler combustion and reduce the emission level of NOX. These results well fitted the energy emission reduction policy and has great significance on environmental protection as well as resources and energy conservation.