锅炉燃烧系统神经网络建模及多目标优化研究  

Multi-objective Neural Network Optimization for Boiler Combustion Systems

作　　者：王志心[1] 包德梅[1] 曹黎明[1] 秦政[1] 

机构地区：[1]国电南京自动化股份有限公司,南京210003

出　　处：《发电设备》2012年第2期97-99,118,共4页Power Equipment

摘　　要：电站锅炉面临降低运行成本与降低污染物排放的双重要求。依据采集到的燃烧数据,采用神经网络训练得到了锅炉燃烧模型,用于预测锅炉在不同燃烧参数的NOx排放和燃烧效率。并采用多目标优化算法,通过调整锅炉运行参数,在锅炉高效燃烧与NOx低排放之间找到合理的平衡点。There are double requirements of both reducing operating cost and decreasing contaminant emission for power station boilers. A boiler combustion model has been obtained with the training of neural network based on combustion data collected, which can be used to forecast the NOx emission and combustion efficiency of the boiler with different combustion parameters. A reasonable balance point is found between high efficiency and low NOx emission of the boiler based on multi-objective optimization and operation data adjustment.

关 键 词：锅炉 燃烧 神经网络 多目标优化 NOX排放 热效率 

分 类 号：TK227.1[动力工程及工程热物理—动力机械及工程] TP183[自动化与计算机技术—控制理论与控制工程]