摘要
本文以南京市某地源热泵空调系统为实测对象,根据实测的逐时负荷数据建立了人工神经网络负荷预测模型,并进行了预测。结果表明,该模型能够精确地预测一个单元未来24小时的逐时负荷,预测误差为5.20%左右。
Taking the heat pump system at a ground source in Nanjing as the measured object, this paper constructs a prediction model for artificial neural network load based on the measured hourly load data, and makes a prediction. The results show that this model can accurately predict the hourly loads of a unit in the future 24 hours, with the prediction error at about 5.20%.
出处
《建筑科学》
北大核心
2014年第2期72-75,共4页
Building Science
基金
建筑安全与环境国家重点实验室开放课题"基于数据挖掘技术的空调优化管理系统的研究"(BSBE2011-05)
关键词
人工神经网络
地源热泵
负荷预测
artificial neural network, ground source heat pump, load prediction
