基于GRA-SA-BP神经网络算法的商场空调冷负荷预测模型  

Cold Load Prediction Model of Shopping Mall Air Conditioning Based on GRA-SA-BP Neural Network Algorithm

作　　者：黄馨乐 于军琪[1] 赵安军[1] Huang Xinle;Yu Junqi;Zhao Anjun(School of Building Services Science and Engineering,Xi’an University of Architecture and Technology,Xi’an,710055)

机构地区：[1]西安建筑科技大学建筑设备科学与工程学院,西安710055

出　　处：《建设科技》2020年第24期113-116,共4页Construction Science and Technology

摘　　要：本文提出了一种基于GRA-SA-BP神经网络算法预测商场的空调冷负荷,利用关联度分析法,剔除输入变量中和输出变量间关联度低的因素,采用模拟退火算法对BP神经网络的权值进行优化,并对优化后的神经网络进行训练,得到冷负荷预测结果。结果表明,GRA-SA-BP模型具有较高的预测精度和可靠性,满足实际的应用需求。In this paper,a method based on GRA-SA-BP neural network algorithm is proposed to predict the air-conditioning cooling load of shopping malls.By using the correlation analysis method,the factors with low correlation between input variables and output variables are eliminated.The simulated annealing algorithm is used to optimize the weights of BP neural network,and the optimized neural network is trained to obtain the cooling load prediction results.The results show that the GRA-SA-BP model has high prediction accuracy and reliability,which can meet the actual application requirements.

关 键 词：灰色关联度分析 模拟退火算法 空调冷负荷 预测模型 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TU831.2[自动化与计算机技术—控制科学与工程]