论文部分内容阅读
对一台300 MW机组W火焰锅炉进行研究,采用最小二乘支持向量机,建立了炉内燃烧温度、污染物生成以及飞灰含碳预测模型;并进一步结合遗传算法,建立了运行优化模型,实现了燃烧过程炉内火焰形貌的重建。研究结果表明所建燃烧预测模型能较准确地预测煤粉的燃尽程度、NOx排放,炉内火焰的形态、火焰中心位置以及火焰的温度水平。优化目标不同,得到的运行条件有显著差别。随负荷的不同,针对不同优化目标得到的运行参数的差别也不相同。以燃尽为优化目标得到的优化运行工况炉内火焰温度高于以控制NOx为优化目标的炉内火焰温度。
A 300 MW unit W flame boiler was studied. The least square support vector machine was used to establish the combustion temperature and pollutant generation in the furnace as well as the predictive model of fly ash carbon content. Combining with genetic algorithm, The reconstruction of the flame morphology in the combustion process is realized. The results show that the proposed combustion prediction model can accurately predict the burnout degree, NOx emission, the shape of the furnace flame, the center of the flame and the temperature of the flame. The optimization objectives are different, the operating conditions are significantly different. With the different load, the different operating parameters obtained for different optimization goals are not the same. Optimized operating conditions with burnout as the optimized target The flame temperature in the furnace is higher than the flame temperature in the furnace optimized for NOx control.