摘要
为了解决火电厂供热和供电的矛盾、增强机组的调峰范围,需要建立精确的热负荷预测模型。针对热电联供系统热负荷影响因素多、耦合性非线性强等特点,利用T-S模糊神经网络算法建立热负荷预测模型,用于预测未来时间热负荷的变化。通过基于减法聚类的模糊C均值聚类算法进行结构辨识,再利用模糊神经网络的混合学习算法进行参数辨识。为了建立热负荷的精确模型,选取热负荷的主要影响因素作为变量。其中,将室外温度、供水温度、回水温度、供水流量作为输入变量,热负荷作为输出变量,并从热电厂DCS上采集连续两天24 h的历史数据,将前一天的数据作为训练集和后一天的数据作为检测集。在MATLAB上进行试验。仿真结果显示:98%的训练样本的相对误差在±0.4%之间,且98%的检测样本的相对误差在±0.6%之间。这表明预测模型的拟合度精确度较好,可以为供热机组调峰研究提供依据。
In order to solve the contradiction between heating and power supply of thermal power plants and enhance the peaking range of the unit,it is necessary to establish an accurate thermal load prediction model.In view of there are many factors affecting the thermal load of the cogeneration system and the strong nonlinear coupling,the T-S fuzzy neural network algorithm is used to establish thermal load forecasting model to predict changes in heat load over time.The fuzzy C-means clustering algorithm based on subtractive clustering is used to identify the structure,and then the hybrid learning algorithm of fuzzy neural network is used to identify the parameters.For establishing an accurate model of the heat load,the main influencing factors of the heat load are selected as the variables,in which the outdoor temperature,the water supply temperature,the return water temperature,and the water supply flow rate are taken as input variables and the heat load is taken as the output variable.The historical data of 24 hours in two consecutive days are collected from the DCS of the thermal power plant;the data of the previous day are used as the training set and the data of the next day are used as the detection set.Experiments are carried out on MATLAB,the simulation results show that the relative error of 98% of the training samples is between ±0.4%,and the relative error of 98% of the detected samples is between ±0.6%;these mean that the accuracy of the prediction model is good.It can provide basis for peaking research on cogeneration units.
作者
姜平
赵保国
张海伟
李丽锋
王鹏程
王欣峰
苑文鑫
JIANG Ping;ZHAO Baoguo;ZHANG Haiwei;LI Lifeng;WANG Pengcheng;WANG Xinfeng;YUAN Wenxin(Shanxi Hepo Power Generation Co.,Ltd.,Yangquan 045001,China;Department of Automation,Shanxi University,Taiyuan 030006,China;School of Mathematical Sciences,Shanxi University,Taiyuan 030006,China)
出处
《自动化仪表》
CAS
2019年第11期20-23,共4页
Process Automation Instrumentation
