摘要
由于锅炉设备庞大,运行条件复杂,煤种多变等因素,很难建立锅炉NOx排放与效率的函数模型。利用最小二乘支持向量机(LS-SVM)建立了以锅炉NOx排放与热效率为输出的混合模型,并对此模型进行了校验。结果表明,该模型具有调节参数少、运算速度快、结果稳定、预测精度高等优点,可以根据燃煤特性以及各操作参数准确预报锅炉在不同工况下的NOx排放和效率。针对模型的多目标优化问题,采用多目标粒子群优化算法MOPSO(multiple objective particle swarm optimization)对某工况进行优化仿真,在提高效率的同时降低了NOx排放。
It is very difficult to develop the NOx emission and efficiency model because of the huge boiler architecture, the complex operating conditions, the large variation of the coal used and so on. A mixed model for boiler NOx emission and its efficiency was established by the Least Square Support Vector Machines （LS-SVM）. The model can predict the NOx emission and boiler efficiency simultaneously with fast calculating speed and high accuracy under various operating conditions. The multiple objective particle swarm optimization （MOPSO） was introduced to solve the multiple objective problems. The simulation shows the proposed algorithm can increase boiler efficiency and reduce its NOx emission in one single run.
出处
《热科学与技术》
CAS
CSCD
2007年第1期26-31,共6页
Journal of Thermal Science and Technology
作者简介
李素芬（1955-），女，副教授，主要从事地热能利用及热能系统优化研究．
