Research Of Thermal Power Plant Coal Proportioning Optimizationbased On SVM

Thermoelectric enterprise needs to improve the efficiency of coal-fired boilerurgently due to the rising coal costs and the serious environmental pollution ofcoal-fired boiler. It aims to reduce pollutant emissions, enhance competitiveness andimprove the efficiency of the enterprise. Under the condition of the coal quality goesworse and worse, carrying on the blending ratio of coal ensures the safe running of theboiler, which is an important way to realize energy conservation and emissionsreduction.Based on the analysis of a certain power plant in Xinjiang and the ratiooptimization point of view from the actual coal-fired boiler, this paper analyzed theindex which affected the safety of boiler production, and then it established a coalproportioning optimization index system. After using support vector institutions, it builtcoal ash melting point prediction model. With the model, this paper took the predictedresults into coal ratio model to optimize. During the modeling process, firstly, it appliedsupport vector machine(SVM) to forecast the coal ash melting point, and took thepredict results into the optimization model, then used tons of steam boiler coalconsumption cost minimum as the objective function and coal quality index and boilercombustion efficiency index as the restriction condition of boiler coal proportioningoptimization. Finally it did an actual blending research according the actual productionsituation of a power plant in Xinjiang, and it proved to be a good practical application.The nonlinear prediction advantages of small samples of SVM improved the ratio ofcoal optimization system operation speed and accuracy, which made convenience forthermal power plant coal ratio. Under the condition of satisfying the requirements of theboiler production load, the least amount of boiler coking, the smallest carbon emissions,it achieved the optimal management of the coal into the furnace, coal saving andemission reduction.