摘要
采用BP神经网络,可利用较少的输入参数建立地板辐射供暖系统热负荷预测模型,以大连市某超低能耗建筑为实测对象,根据实测的供暖期逐时热负荷数据建立了BP神经网络热负荷预测模型,并进行了改进。结果表明,采用基于多项式拟合改进的神经网络预测模型能够精确地预测一个单元未来24h的逐时热负荷,预测误差为5%左右。
BP neural network is used to predict hourly heating load and it requires fewer parameters for establishing the model. With the field survey data in a lower energy consuming building in Dalian city, establishes a prediction model of hourly heating load based on BP neural network and makes some improvement on it. The results show that the improved BP neural network prediction model based on polynomial fitting can accurately predict hourly heating load for one unit in the next 24 hours, and the prediction error is about 5%.
出处
《暖通空调》
北大核心
2011年第12期95-98,共4页
Heating Ventilating & Air Conditioning
基金
"十一五"国家科技支撑计划项目(编号:2006BAJ03B0101)
