基于BP神经网络的燃煤锅炉大气污染物排放模型构建  

Establishment of air pollutants emission model of coal-fired boiler based on BP neural network

作　　者：奚增辉 王卫斌 洪祎祺 姚嵘 瞿海妮 Xi Zenghui;Wang Weibin;Hong Yiqi;Yao Rong;Qu Haini(State Grid Shanghai Electric Power Company,Shanghai 200122,China)

机构地区：[1]国网上海市电力公司,上海200122

出　　处：《煤化工》2023年第4期142-145,共4页Coal Chemical Industry

摘　　要：提出了一种基于BP神经网络的燃煤锅炉大气污染物排放量预测模型构建方法,即通过在BP神经网络中训练燃煤锅炉排放样本数据,纠正训练误差后优化权值和阈值,输出燃煤锅炉大气污染物排放数据,以输出结果为基础,得到脱硝反应器、脱硫塔出口的大气污染物氮、硫排放量。结果表明:构建的基于BP神经网络的燃煤锅炉大气污染物排放模型具有大气污染物排放量预测误差小、精度高、模型可行性高等优点,对评估和控制燃煤锅炉大气污染物排放具有重要意义。The method for constructing a prediction model for air pollutants emission from the coal-fired boiler was proposed based on the BP neural network.By training the emission sample data of the coal-fired boiler in the BP neural network,correcting the training errors,optimizing the weights and thresholds,and outputting the air pollutants emission data of the coal-fired boiler,and based on the output results,the emissions of air pollutants nitrogen and sulfur were obtained at the outlet of the denitrification reactor and desulfurization tower.The results showed that such a constructed model had small error,high accuracy,and high feasibility in predicting the air pollutants emission,which was of great significance for evaluating and controlling the air pollutants emission from the coal-fired boiler.

关 键 词：燃煤锅炉 大气污染物 BP神经网络 脱硫塔 脱硝反应器 

分 类 号：TK229[动力工程及工程热物理—动力机械及工程]