摘要
针对大型电站锅炉空气预热器受热面积灰状况进行了分析研究。应用3层神经网络构建了300MW电站锅炉空气预热器受热面积灰监测数学模型,选择锅炉负荷、烟气差压、排烟温度等参数作为输入向量,以反映空气预热器积灰状况的污染系数作为输出向量,利用电厂DCS系统采集的机组实时数据,经规格化处理后作为样本集对网络进行训练。训练过程中,通过添加动量项和变步长改进了BP算法。将建立的模型应用于华电国际青岛发电公司#2炉的空气预热器在线积灰监测,取得了较好的结果。
The ash deposit monitoring model of air preheater for power generation boiler based on the BP neural network is presented in this paper.The neural network is a three-layer BP network and is improved by additional momentum item and variable step size.The network uses the parameters such as boiler load,flue gas pressure-drop of air preheater,flue gas temperature before and after the air preheater etc.,as the inputs,ash deposit factor as the output.Training data are from on-line DCS after pre-selected and normalized.The model is used to monitor ash deposit of No.2 300MW-boiler air preheaters in Qingdao Power Station.The result shows that the model can identify the sootblowing process of air preheaters and can be used to monitor the ash deposit status of air preheater and optimize sootblowing intervals.
出处
《能源研究与利用》
2006年第3期14-16,20,共4页
Energy Research & Utilization
关键词
神经网络
空气预热器
积灰
neural network
air preheater
ash deposit