Leveraging reinforcement learning for optimal control of chiller plant with complex hydraulic and thermodynamic characteristics

IF 6.6 2区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Energy and Buildings Pub Date : 2025-06-16 DOI:10.1016/j.enbuild.2025.116037

Xiao Wang , Huiming Xu , Cheng Fan , Shengyu Tan , Xuyuan Kang , Yang Shi , Da Yan

{"title":"Leveraging reinforcement learning for optimal control of chiller plant with complex hydraulic and thermodynamic characteristics","authors":"Xiao Wang , Huiming Xu , Cheng Fan , Shengyu Tan , Xuyuan Kang , Yang Shi , Da Yan","doi":"10.1016/j.enbuild.2025.116037","DOIUrl":null,"url":null,"abstract":"<div><div>The optimal control of chiller plants is vital for enhancing the energy efficiency of buildings. The operation combination of cooling towers and pumps has a significant impact on the heat exchange of the cooling water system and the chiller performance. However, optimizing the control strategy of a chiller plant with complex hydraulic and thermodynamic characteristics is challenging. This study proposes a framework for reinforcement learning (RL) control on the cooling sides of chiller plants. A detailed physical model of a chiller plant was constructed to provide a reliable training environment for the RL model. The proposed RL controller was deployed in a chiller plant in a commercial complex in Guangzhou, China. Compared with a rule-based control, the proposed RL control can save 741 MWh (8.1%) of electricity consumption over eight months. The optimal control strategies provide insights into the chiller plant system characteristics, which offer promise for broader implementation on a broader scale in building energy systems.</div></div>","PeriodicalId":11641,"journal":{"name":"Energy and Buildings","volume":"344 ","pages":"Article 116037"},"PeriodicalIF":6.6000,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Energy and Buildings","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0378778825007674","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"CONSTRUCTION & BUILDING TECHNOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

The optimal control of chiller plants is vital for enhancing the energy efficiency of buildings. The operation combination of cooling towers and pumps has a significant impact on the heat exchange of the cooling water system and the chiller performance. However, optimizing the control strategy of a chiller plant with complex hydraulic and thermodynamic characteristics is challenging. This study proposes a framework for reinforcement learning (RL) control on the cooling sides of chiller plants. A detailed physical model of a chiller plant was constructed to provide a reliable training environment for the RL model. The proposed RL controller was deployed in a chiller plant in a commercial complex in Guangzhou, China. Compared with a rule-based control, the proposed RL control can save 741 MWh (8.1%) of electricity consumption over eight months. The optimal control strategies provide insights into the chiller plant system characteristics, which offer promise for broader implementation on a broader scale in building energy systems.

查看原文本刊更多论文

利用强化学习进行复杂水力和热力特性的冷水机组最优控制

制冷机组的优化控制对提高建筑物的能源效率至关重要。冷却塔和水泵的运行组合对冷却水系统的热交换和冷水机组的性能有重要的影响。然而，具有复杂水力和热力特性的冷水机组的控制策略优化是一个挑战。本研究提出了一种强化学习（RL）控制框架，用于冷水机组的冷却侧。建立了冷水机组的详细物理模型，为RL模型提供了可靠的训练环境。所建议的RL控制器已在中国广州的一个商业综合体的冷水机组中部署。与基于规则的控制相比，RL控制在8个月内可节省741兆瓦时（8.1%）的电力消耗。最优控制策略提供了对冷水机组系统特性的深入了解，为建筑能源系统在更大范围内的更广泛实施提供了希望。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Energy and Buildings 工程技术-工程：土木

CiteScore

12.70

自引率

11.90%

发文量

863

审稿时长

38 days

期刊介绍： An international journal devoted to investigations of energy use and efficiency in buildings Energy and Buildings is an international journal publishing articles with explicit links to energy use in buildings. The aim is to present new research results, and new proven practice aimed at reducing the energy needs of a building and improving indoor environment quality.