基于BP神经网络下空调负荷预测模型的研究  

Study on the Air-Conditioning Load Prediction Model Based On Bp Neural Network

作　　者：林育贤[1] 冯圣红[1] LIN Yu - xian FENG Sheng - hong(Beijing University of Civil Engineering and Architecture, College of Environment and Energy Engineering, Beijing key lab of heating, gas supply, ventilating and Air Conditioning Engineering, Beijing 100044, China)

机构地区：[1]北京建筑大学环境与能源工程学院,供热供燃气通风及空调工程北京重点实验室,北京100044

出　　处：《应用能源技术》2016年第9期46-51,共6页Applied Energy Technology

摘　　要：在目前暖通空调领域已有的研究中,人工神经网络大多数都用在空调系统的负荷预测上,尤其以BP神经网络的研究最为深入。但是,当前并没有形成系统的、简便的神经网络负荷预测模型。为建立一套完整的神经网络负荷预测模型,需要考虑诸多因素。文中将对BP神经网络的结构和参数值进行详细地分析,并分析讨论输入层和隐含层神经元数目的确定以及样本集长度的寻优方法。最后,通过实际工程数据对神经网络负荷预测模型中各环节进行确定,从而最终确定负荷预测模型的最佳结构,证明本模型是行之有效的。In the recent research in the field of HVAC,most of the artificial neural network is used in the load prediction of air-conditioningsystem,especially in the research of BP Neural Network.However,there is no systematic and convenient neural network load prediction model. In order to establish a complete neural network load forecasting model,many factors need to be considered. The structure and parameters of BP Neural Network are analyzed in detail; the method of determining the number of neurons in the input and hidden layers and the length of the sample set is analyzed and discussed. Finally,the each link in the neural network load predictionmodelwill be confirmed via an actual engineering subject,the best structure of the load prediction model will be settled eventually,which prove the model is effective.

关 键 词：BP神经网络 空调负荷预测模型 输入层 输出层 样本集长度 

分 类 号：TB6[一般工业技术—制冷工程] TU831[建筑科学—供热、供燃气、通风及空调工程]