基于改进粒子群算法的电站锅炉NOx排放预测控制及优化

摘要

A support vector machine(SVM) model of Noχ emission was developed and verified by the data on-spot. Then, operating parameters were optimized by an improved linear particles swarm optimization(ILPSO), and compared with the linear particles swarm optimization(LPSO). The results show that Noχ emissions are significantly lower, and ILPSO algorithm has better convergence. This is a better way to predictive control of Noχ emission in power plant boiler.%基于支持向量机算法建立锅炉NOx排放模型,并利用实炉热态数据对模型进行校验.应用一种改进的粒子群优化算法对锅炉运行参数进行优化,并与一般线性粒子群优化算法进行对比.研究结果表明,NOx排放量明显降低,且改进的优化算法收敛性更好,为锅炉NOx排放的预测控制提供了更好的方式.