基于SOA-RVFL预测模型的建筑节能控制方法研究  

Research on Building Energy Saving Control Method Based on the SOA-RVFL Prediction Model

作　　者：孙鸿昌 翟文文 Sun Hongchang;Zhai Wenwen(Shandong Dawei International Architecture Design Co.,Ltd.,Jinan 250101,China)

机构地区：[1]山东大卫国际建筑设计有限公司机电智能化设计院,济南250101 [2]山东大卫国际建筑设计有限公司,济南250101

出　　处：《智能建筑电气技术》2021年第6期76-80,共5页Electrical Technology of Intelligent Buildings

摘　　要：为了在满足建筑室内舒适性的同时更有效地节约能耗,本文提出了一种基于海鸥算法优化的随机向量功能连接网络(SOA-RVFL)策略对建筑能耗与温度进行预测,并通过预测结果动态调节建筑内的制热/制冷系统。策略在济南某公共建筑上应用,结果与传统的基线控制相比,降低了11.9%的建筑能耗。In order to satisfy the indoor comfort of buildings and save energy consumption more effectively at the same time,this paper proposes a seagull optimization algorithm based random vector functional link network(SOA-RVFL)strategy to predict the building energy consumption and temperature,and dynamically adjusts the heating/cooling system in the building through the prediction results.The strategy was applied to a public building in Jinan.Results indicated that the proposed method could reduce the building energy consumption by about 11.9% in comparison with the traditional base control method.

关 键 词：模型预测控制 随机向量函数连接网络 海鸥优化算法 建筑节能 

分 类 号：TU855[建筑科学]