结合天气因素的HTS-FOA-GRNN短期热负荷预测研究  

Research on HTS-FOA-GRNN Short-Term Heat Load Forecasting Combining Weather Factors

作　　者：顾伟 薛贵军 李水清[2] GU Wei;XUE Gui-jun;LI Shui-qing(North China University of Science and Technology,Hebei Tangshan 063210,China;Hebei Polytechnic University Intelligent Instrument Factory,Hebei Tangshan 063000,China)

机构地区：[1]华北理工大学,河北唐山063210 [2]河北理工大学智能仪器厂,河北唐山063000

出　　处：《机械设计与制造》2022年第1期81-83,87,共4页Machinery Design & Manufacture

基　　金：国家自然科学基金青年项目(61502143)。

摘　　要：随着社会的不断发展进步以及社会各界对环境保护问题的高度重视,如何做到让冬季北方地区的集中供热系统经济环保运行成为了目前我国供热行业的首要目标。短期热负荷预测的研究恰恰可以提高供热系统控制精度,优化供热系统控制方案,实现系统的经济环保运行。由于热用户的室内温度受外界环境温度的影响因素较大,因此在结合天气因素再进行神经网络建模会使预测更加精准。提出一种基于改进果蝇算法优化的广义回归神经网络(FOA-GRNN)以及结合天气因素的短期热负荷预测方法,选取供热负荷值、实时天气温度和二次网供水温度和二次网回水温度等四类数据,构建HTS-FOA-GRNN供热负荷短期预测模型。研究表明,所构建的HTS-FOA-GRNN模型具有很好的预测能力和优化效果。With the continuous development and progress of the society and the high attention paid by the community to environmental protection issues,how to make the central heating system in the northern part of the country economic and environmental protection has become the primary goal of China’s heating industry.The research on short-term heat load forecasting can improve the control precision of heating system,optimize the control scheme of heating system,and realize the economic and environmental protection operation of the system.Since the indoor temperature of the hot user is greatly affected by the external environment temperature,the neural network modeling combined with the weather factor will make the prediction more accurate.A Generalized Regression Neural Network(FOA-GRNN)based on improved drosophila algorithm optimization and a short-term heat load forecasting method combined with weather factors are proposed.Four types of data,including heat load value,realtime weather temperature and secondary network water supply temperature and secondary network return water temperature,are selected to construct HTS-FOA-GRNN short-term heat load forecasting model.The research shows that the constructed HTSFOA-GRNN model has good predictive ability and optimization effect.

关 键 词：HTS FOA GRNN 供热负荷预测 外界温度 二次网 

分 类 号：TH16[机械工程—机械制造及自动化] TP391.9[自动化与计算机技术—计算机应用技术]