基于模型预测控制的冰蓄冷空调系统优化控制策略研究  

Research on Optimal Control Strategy for Ice Storage Air Conditioning System Based on Model Predictive Control

作　　者：鲍佳馨 谈卓越 王慧龙 Bao Jiaxin;Tan Zhuoyue;Wang Huilong(School of Civil and Transportation Engineering,Shenzhen University,Shenzhen Guangdong 518052)

机构地区：[1]深圳大学土木与交通工程学院,广东深圳518052

出　　处：《中外能源》2024年第7期96-104,共9页Sino-Global Energy

基　　金：国家自然科学基金青年项目“室温设定值调节下空调能耗回调特性及柔性用能控制方法”(编号:52308104);深圳可持续发展科技专项(双碳专项)“建筑高效节能及电力柔性技术研究与应用示范”(编号:KCXST20221021111203007)。

摘　　要：在峰谷电价背景下,冰蓄冷空调系统可以利用谷值电价进行蓄冷,在电价较高时段释放蓄冷量,从而减少峰值用电量。为此,提出一种基于模型预测控制(MPC)的冰蓄冷空调系统优化控制策略,目标是系统能耗成本最低,同时解决传统运行策略不当而导致的冰槽使用效率低下问题。基于参照建筑搭建TRNSYS-Python联合模拟仿真平台,采用极端梯度提升(XGBoost)算法构建建筑未来24h逐时冷负荷预测模型,运用模型预测控制理论,考虑建筑负荷预测偏差,逐时滚动求解冰蓄冷空调系统蓄放冷最优控制策略,在保证全天冷量合理分配的同时,实现峰谷电价背景下系统运行能耗成本最低。与传统规则策略相比,优化控制策略可以针对不同冷负荷需求采取灵活的控制措施,在冷负荷需求较高时,将有限的冷量优先用在电价高峰期,有效规避高额电费支出,降低能耗成本;在冷负荷需求较低时,通过精准预测,仅储存必要冷量,确保冰量在一天中被完全利用,提升冰蓄冷系统效率。在基于设计日负荷100%、75%、50%三种工况下,优化控制策略比传统规则策略分别可节约能耗成本7.95%、12.64%和10.18%。Under the background of peak-valley electricity price,ice storage air conditioning system can store cooling capacity during valley hours and release the cooling capacityduring peak hours,thus reducing the peak electricityconsumption.Therefore,an optimal control strategy for ice storage air conditioning system based on model predictive control(MPC)is proposed,with an aim to minimize the energy costs of the system and solve the inefficient use of ice tank caused by improper operation strategy.Based on the TRNSYS-Python joint simulation platform built with reference to the building,the extreme gradient lift(XGBoost)algorithm is adopted to construct an hourly cooling load prediction model for the building in the future 24h.The model predictive control theory is applied to solve the optimal control strategy for cold storage and release of the ice storage air conditioning system on an hourly rolling basis,taking into consideration the deviation of the building load prediction,so as to minimize the operating energy costs of the system under the background of peak-valley electricity price while ensuring the reasonable allocation of cooling capacity throughout the day.Compared with the traditional rule-based strategy,the optimal control strategy can take flexible control measures for different cooling load demands.When the demand for cooling load is high,the limited cooling capacity is preferentially used during peak hours,effectively avoiding high electricity costs and reducing energy costs.When the demand for cooling load is low,only the necessary amount of cooling capacity is stored based on accurate prediction,ensuring that the ice is used up in a day and improving the efficiency of the ice storage system.Compared with the traditional rule-basedstrategy,the optimalcontrol strategy can save the energy costs by 7.95%,12.64% and 10.18% based on a design daily load of 100%,75% and 50%,respectively.

关 键 词：冰蓄冷空调系统 模型预测控制 优化控制策略 冷负荷预测 峰谷电价 能耗成本 

分 类 号：TU831.3[建筑科学—供热、供燃气、通风及空调工程]