摘要
本文对大数据时代背景下的空调系统数据的介绍,通过分析BP人工神经网络的运算原理、拓扑结构与数学表达,提出了SPSS时序性分析与BP神经网络结合的预测模型,应用于空调系统逐时负荷预测。以南京某办公建筑为例,提出了样本以工作日与非工作日分类,选择在预测日前一月作为样本时域,通过Matlab软件实现预测模型进行数据分析,相对误差保持在10%以下。
This paper introduces the data of air conditioning system in the era of big data.By analyzing the operation principle,topological structure and mathematical expression of BP artificial neural network,a prediction model combining SPSS time series analysis and BP neural network is proposed,which is applied to the hourly load forecasting of air conditioning system.Taking an office building in Nanjing as an example,the sample is classified into working days and non-working days.The sample is selected in the time domain of one month before the forecast date.The prediction model is analyzed by using MATLAB software,and the relative error is kept below 10%.
作者
张峰
李苏泷
ZHANG Feng;LI Su-long(School of Energy and Power Engineering,Nanjing University of Science and Technology)
出处
《智能建筑与智慧城市》
2019年第7期34-35,41,共3页
Intelligent Building & Smart City
