基于历史数据分析锅炉燃烧优化方法.pdfVIP

基于历史数据分析的锅炉燃烧优化方法 于爽、刘浩 天津华能杨柳青热电有限责任公司,天津市300380 Boiler combustion optimization based on historical data analysis YU Shuang,LIU Hao Yangliuqing electric power plant of Huaneng,Tianjin,300380 ABSTRACT :Large coal-fired power plant boiler 火电厂锅炉燃烧系统为研究对象,探讨研究了基于 combustion efficiency and low emission of polluting 历史数据分析的锅炉燃烧优化理论与应用。火电厂 gases for conserving energy and protecting the 锅炉设备庞大,过程多变量、大延迟、输入输出特 environment are important. The combustion 性复杂、耦合性强、燃用煤种和操作条件变化较大， optimization can increase the efficiency of boiler and 无法用理论方法建立燃烧模型。本文采用 BP 神经 reducing nitrogen oxides emissions, thus it was widely 网络技术，根据锅炉燃烧过程中保存的历史数据， applied,This article studies the boiler combustion 利用 BP 神经网络建立锅炉燃烧系统模型，并采用 systems, based on historical data analysis of the boiler 遗传算法非线性寻优技术,从模型中找出不同运行 combustion optimization theory and application. The 参数下最佳的运行方式，以此指导锅炉燃烧调整， power plant has huge boiler equipment, multiple 实现锅炉燃烧系统的优化运行，对锅炉的节能降耗 variables, delaying the process, input and output 和降低环境污染都有重要意义。 characteristics of a complex, strong coupling, fueled with coal and the larger operating conditions change. 关键词：低NOx 排放；燃烧优化；BP 神经网络； It is unable to create combustion model by theoretical methods. This article used BP neural network 遗传算法 technology, according to boiler burning process in the save of history data, using BP neural network 1 概述 established boiler burning system model. It uses 由于火电厂的锅炉燃烧会排放很多有 genetic algorithm nonlinear technology and finds the 毒物质例如硫氧化物 SOx、氮氧化物 NOx best running way from model in different run parameter, to guide boiler burning adjustment, and 和粉尘等，这些物质都是危害人体健康和破 achieve boiler burning system of optimization run. 坏大气环境的剧毒污染物。所以，如何降低