分层表冷器的水循环优化预测控制研究  

Research on Water Cycle Optimization and Predict Control of Stratified Surface Cooler

作　　者：王华秋 朱天虎 高鹏 WANG Huaqiu;ZHU Tianhu;GAO Peng(School of Liangjiang Artificial Intelligence,Chongqing University of Technology,Chongqing 401135,China;Shifang Cigarette Factory,China Tobacco Sichuan Industrial Co.Ltd,Shifang 618400,China)

机构地区：[1]重庆理工大学两江人工智能学院,重庆401135 [2]四川中烟工业有限责任公司什邡卷烟厂,四川什邡618400

出　　处：《控制工程》2024年第6期1004-1011,共8页Control Engineering of China

基　　金：国家科技部重点研发计划项目(2018YFB1700803)。

摘　　要：针对分层表冷器的协同控制问题,根据室外温度和室外相对湿度这2种环境变量,采用门控循环单元(gated recurrent unit,GRU)网络搭建表冷器温湿度系统预测模型,用于预测室内平均温度和室内平均湿度。首先,针对水循环优化算法收敛速度慢的问题,将生物地理学优化(biogeography-based optimization,BBO)算法中的迁入迁出行为嵌入到水循环算法(water cycle algorithm,WCA)中,以提高算法性能;然后,利用改进的水循环算法(improved water cycle algorithm,iWCA)滚动优化得到控制量,即主表冷阀、副表冷阀和电动三通阀的开度。仿真实验结果表明:与传统的预测控制算法相比,GRU-iWCA预测控制方法具有更高的跟踪精度和稳定性。For the problem of collaborative control of stratified surface coolers,the prediction model system of surface cooler temperature and humidity is established by gated recurrent unit(GRU)network based on two environmental variables,i.e.,prediction of average indoor humidity and average indoor temperature.Firstly,to address the problem of slow convergence of the water cycle optimization algorithm,the migration-in and migration-out behaviors of Biogeography-Based Optimization(BBO)algorithm are embedded into the water cycle optimization algorithm(WCA)to improve the performance of the optimization algorithm.Then,the control system obtains the control objects by receding optimization using the improved water cycle optimization algorithm(iWCA),i.e.,the main meter-cooling valve opening,the sub-meter-cooling valve opening and the electronic three-way valve opening.Finally,the simulation experiments show that the GRU-iWCA predictive control system has higher trace ability and stability compared with the other predictive control algorithm.

关 键 词：分层表冷器 门控循环单元 预测控制 水循环算法 滚动优化 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]