摘要
本文对BP神经网络的构成、算法设置及BP模型的预测步骤进行介绍,并以深圳某一大型公共建筑为例进行了负荷模拟,将BP神经网络模型应用到冷负荷预测活动中,数值模拟结果表明BP神经网络模型对负荷与各输入变量有很好的映射能力。
This paper introduces the composition, algorithm setting and prediction steps of BP model. A large public building in Shenzhen serves as an example to undertake simulation process, and the BP neural network model is applied to the cooling load prediction. The numerical simulation results show that the BP neural network model has a good mapping ability to the load and input variables.
引文
[1]清华大学建筑节能研究中心.中国建筑节能年度发展研究报告[M].中国建筑工业出版社,2016.
[2]于新巧,陈征,汪汀,等.我国办公建筑用能行为现状调研与分析[J].建筑科学,2015,31(10):23-30.
[3]Yan D,Zhe T,Yong W,et al.Achievements and Suggestions of Heat Metering and Energy Efficiency Retrofit for Existing Residential Buildings in Northern Heating Regions of China[J].Energy Policy,2011,39(9):4675-4682.
[4]Yun K,Luck R,Mago P J,et al.Building hourly thermal load prediction using an indexed ARXmodel[J].Energy and Buildings,2012,54:225-233.
[5]Henninger R H,Witte M J.EnergyPlus Testing with Building Thermal Envelope and Fabric Load Tests from ANSI[J].ASHRAE standard,2013,140:2011.