摘要
结合制冷机组运行原理,建立了基于门控循环单元网络(GRU)的某卷烟厂空调制冷机组能效预测模型,采用交叉熵鲸鱼优化算法(CWOA)对制冷运行工况进行寻优,该优化策略能跟随冷负荷变化动态调节设备运行参数。基于这2种智能算法的决策系统既解决了卷烟厂空调制冷机组机理建模困难的问题,又解决了制冷机组运行能耗高的问题。以某卷烟厂空调制冷机组实测运行数据为基础进行了预测和优化仿真实验,预测结果表明制冷机组GRU模型在训练集和测试集上的均方根误差(RMSE)分别为1.047%和1.186%,预测精度高于LSTM网络模型,优化结果显示:CWOA优化后的能效值相比其他优化算法更高,较实际运行平均节能9%。因此,基于GRU-CWOA算法的智能模型可用于卷烟厂空调制冷机组的能效预测优化。
The energy efficiency prediction model of air-conditioning chillers in a cigarette factory is established based on the gated recurrent unit network(GRU) according to the chillers operation principle in this paper. Then the cross-entropywhale optimization algorithm(CWOA) is used to optimize the chillers operation conditions to form the optimization strategy which can dynamically adjust the equipment operation parameters with the change of cooling load. The decision-making system based on these two intelligent algorithms not only solves the problem of difficult mechanism modeling of air-conditioning chillers in cigarette factory, but also solves the problem of high energy consumption of chillers. The prediction and optimization simulation experiment are performed based on the actual operation data of the air-conditioning chillers in a cigarette factory in this paper. The prediction results show that the root mean square error(RMSE) of GRU model in the training set and the test set are 1.047% and 1.186% respectively. The prediction accuracy is higher than that of LSTM network model. The optimization results show that the energy efficiency optimized by CWOA is higher than that by other optimization algorithms, it saves 9% energy on average compared with the actual operation. Therefore, the intelligent model based on GRU-CWOA algorithm can be used in the prediction and optimization of energy efficiency of air-conditioning chillers in cigarette factories.
作者
徐伟民
邬剑升
余数
桂腾跃
王华秋
向力
XU Weimin;WU Jiansheng;YU Shu;GUI Tengyue;WANG Huaqiu;XIANG Li(China Tobacco Zhejiang Industrial Co.,Ltd.,Ningbo 315504,China;Liangjiang Artificial Intelligence College of Chongqing University of Technology,Chongqing 401135,China;Chongqing Taihe Air Conditioning Automatic Control Co.,Ltd.,Chongqing 400038,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2022年第9期137-145,共9页
Journal of Chongqing University of Technology：Natural Science
基金
国家科技部重点研发计划项目(2018YFB1700803)
重庆市科委一般自然基金项目(cstc2019jcyj-msxmX0500)。
关键词
制冷机组
能效预测
能效优化
门控循环单元网络
交叉熵鲸鱼算法
chillers
energy efficiency prediction
energy efficiency optimization
gating recurrent unit network
cross-entropywhale optimization algorithm
