基于人工神经网络的电站锅炉积灰实时监测系统  

Real-time Monitoring System for Ash Deposit in Utility Boiler Based on Artificial Neural Network

作　　者：杨祥良[1] 安连锁[1] 孙鑫强[1] 孙保民[1] 沈国清[1] 

机构地区：[1]华北电力大学电站设备状态监测与控制教育部重点实验室,北京102206

出　　处：《动力工程学报》2010年第3期206-209,共4页Journal of Chinese Society of Power Engineering

摘　　要：针对燃煤电站锅炉对流受热面积灰,提出了一种基于人工神经网络的积灰实时监测方法:利用受热面清洁吸热量和实际吸热量定义灰污特征参数;通过电厂现有的DAS系统得到的温度、压力和流量等参数可获得大量样本点;建立神经网络模型并进行训练.在燃用神华煤的某300 MW锅炉上进行了试验.结果表明:实测吸热量与预测吸热量的最大误差不超过10%,平均误差为3%左右.该方法可准确预测锅炉对流受热面的积灰情况.Aiming at ash deposit on convective heating surface in coal-fired utility boiler,a real-time monitoring method for ash deposit based on artificial neural network was proposed.First,the fouling characteristic parameter was defined by use of the heat absorption of the clean heating surface and the actual heat absorption.Then a lot of samples of temperature,pressure and flow rate were obtained from the existing DAS system in the power plant.Finally,the artificial neural network model was established and trained.The experiments were carried out in a 300MW boiler with Shenhua coal.Results show that the maximum error between the actual and the predicted heat absorption is no more than 10%,and the mean error is about 3%.The method can accurately predict the ash deposit of boiler convection heating surface.

关 键 词：燃煤锅炉 积灰 吹灰 人工神经网络 在线监测 

分 类 号：TK31[动力工程及工程热物理—热能工程]