应用深度强化学习的电池热管理系统控制策略

席椿富; 赵东鹏; 黄驰; 邹子豪; 黄琨杰; 谢翌

doi:10.7652/xjtuxb202602003

您当前的位置：

首页 >

文章列表页 >

应用深度强化学习的电池热管理系统控制策略

专题新能源电动车辆热管理 | 更新时间：2026-02-09

- 应用深度强化学习的电池热管理系统控制策略
- Control Strategy for Battery Thermal Management System Using Deep Reinforcement Learning
- 西安交通大学学报 2026年60卷第2期页码：24-37
- 作者机构：
  
  中国汽车工程研究院股份有限公司，401122，重庆
  天津大学机械工程学院，300350，天津
  重庆大学低品位能源利用技术及系统教育部重点实验室，400044，重庆
  重庆大学能源与动力工程学院，400044，重庆
  重庆大学机械与运载工程学院，400044，重庆
- 作者简介：
  
  作者简介：席椿富（1990—），男，工程师；
  谢翌（通信作者），男，教授，博士生导师。
- 基金信息：
  
  国家自然科学基金资助项目。(52472375)
- DOI：10.7652/xjtuxb202602003
  中图分类号： TM912
- 收稿：2025-06-30，
  
  纸质出版：2026-02-10
- 稿件说明：
移动端阅览
席椿富, 赵东鹏, 黄驰, 等. 应用深度强化学习的电池热管理系统控制策略[J]. 西安交通大学学报, 2026,60(2):24-37.

XI Chunfu, ZHAO Dongpeng, HUANG Chi, et al. Control Strategy for Battery Thermal Management System Using Deep Reinforcement Learning[J]. Journal of Xi'an Jiaotong University, 2026, 60(2): 24-37.
席椿富, 赵东鹏, 黄驰, 等. 应用深度强化学习的电池热管理系统控制策略[J]. 西安交通大学学报, 2026,60(2):24-37. DOI： 10.7652/xjtuxb202602003.

XI Chunfu, ZHAO Dongpeng, HUANG Chi, et al. Control Strategy for Battery Thermal Management System Using Deep Reinforcement Learning[J]. Journal of Xi'an Jiaotong University, 2026, 60(2): 24-37. DOI： 10.7652/xjtuxb202602003.

摘要

针对电动汽车电池热管理系统中传统控制方法温控精度不足及环境适应性差的难题，提出基于深度强化学习的智能控制体方法。基于电池热电耦合模型与制冷空调系统模型，应用强化学习中的双延迟深度确定性策略（TD3）算法进行控制策略训练，通过双重价值网络与延迟策略更新机制，克服传统强化学习中的过高估计问题。结果表明：在夏季充电的训练工况下，能够将电池包的平均温度控制在25℃左右；在冬季充电的训练工况下，能够将电池包的平均温度控制在20℃左右，电池模组之间的最大温差控制在1℃以内。同时，在控制动作上，智能体控制的压缩机转速的调整更为平缓，与比例-积分-微分控制、开关控制相比，智能体控制在夏季放电时最高节能了32.1%，充电时最高节能了15.8%，冬季放电时最高节能了17.0%，充电时最高节能了26.3%。此外，在环境条件变化时，智能体能够及时调整控制动作，将电池包的温度控制在目标温度附近。该研究利用TD3强化学习算法能够在多变的环境条件下平稳、精准地控制电池热管理系统，证明了强化学习在电池热管理中的可行性与有效性。

Abstract

To address the challenges of insufficient temperature control accuracy and poor environmental adaptability in traditional control methods for electric vehicle battery thermal management systems，an intelligent agent control method based on deep reinforcement learning is proposed. Based on a battery electro-thermal coupling model and an air conditioning refrigeration system model，the twin delayed deep deterministic policy gradient（TD3）algorithm in reinforcement learning is applied to train the control strategy.By utilizing dual critic networks and a delayed policy update mechanism，the issue of overestimation common in traditional reinforcement learning is overcome.Results show that under summer charging conditions，the average battery pack temperature can be controlled around 25℃，while under winter charging conditions，it can be maintained around 20℃，with the maximum temperature difference between battery modules controlled within 1℃.Moreover，the compressor speed adj ustments made by the intelligent agent are smoother.Compared with proportional-integral-derivative control and on-off control，the intelligent agent control achieves energy savings of up to 32.1% during summer discharging，15.8% during summer charging，17.0% during winter discharging，and 26.3% during winter charging. Additionally，when environmental conditions change，the agent can promptly adj ust control actions to maintain the battery pack temperature near the target.This study demonstrates that the TD3 reinforcement learning algorithm can achieve stable and precise control of the battery thermal management system under varying environmental conditions，proving the feasibility and effectiveness of reinforcement learning in battery thermal management.

关键词

Keywords

references

潘家华,董秀成,崔洪建,等.欧洲能源危机及其影响分析[J].国际经济评论, 2023(1):9-37.

PAN Jiahua, DONG Xiucheng, CUI Hongjian, et al. Analysis of European energy crisis and its impact[J]. International Economic Review, 2023(1):9-37.

樊大磊,王宗礼,王彧嫣,等.2024年国内外油气资源形势分析及展望[J].中国矿业, 2025 , 34(1):46-54.

FAN Dalei, WANG Zongli, WANG Yuyan, et al. Analysis of oil and gas resources situation at home and abroad in 2024 and outlook[J].China Mining Magazine, 2025 , 34(1):46-54.

HUANG Aishi, ZHANG Lei, CHENG Wenxuan, et al.CO 2 emissions associated with China's real estate development:2000—2020[J ] .Journal of Environmental Sciences, 2025, 156:495-505.

李晓易,谭晓雨,吴睿,等.交通运输领域碳达峰、碳中和路径研究[J].中国工程科学, 2021, 23(6):15-21.

LI Xiaoyi, TAN Xiaoyu, WU Rui, et al.Paths for carbon peak and carbon neutrality in transport sector in China[J].Strategic Study of CAE, 2021 , 23(6):15-21 .

乔英俊,赵世佳,伍晨波,等.“双碳”目标下我国汽车产业低碳发展战略研究[J].中国软科学, 2022(6):31-40.

QIAO Yingjun, ZHAO Shijia, WU Chenbo, et al. Research on low-carbon development strategy of China's automotive industry with the “carbon-peak and carbon-neutrality”goal[J].China Soft Science, 2022 (6):31-40.

欧阳明高.能源革命与新能源智能汽车[J].中国工业和信息化, 2019(11):21-24.

OUYANG Minggao.Energy revolution and new energy intelligent vehicles[J].China Industry & Information Technology, 2019(11):21-24.

孙叶,刘锴.里程焦虑对纯电动汽车使用意愿的影响[J].武汉理工大学学报(交通科学与工程版), 2017, 41(1):87-91.

SUN Ye, LIU Kai.Impact of mileage anxiety on intention to use pure electric vehicle[J].Journal of Wu-han University of Technology(Transportation Science& Engineering), 2017, 41(1):87-91.

于旭东,徐爽,李科迪.低温环境对纯电动汽车续驶里程的影响因素研究[J].汽车电器, 2023(2):1-3.

YU Xudong, XU Shuang, LI Kedi.Study on factors of low-temperature driving range of pure electric vehicle[J].Auto Electric Parts, 2023(2):1-3.

ZHANG Ziqi, LIU Cichong, CHEN Xiaoning, et al. Annual energy consumption of electric vehicle air conditioning in China[J].Applied Thermal Engineering, 2017, 125:567-574.

李哲,韩雪冰,卢兰光,等.动力型磷酸铁锂电池的温度特性[J].机械工程学报, 2011, 47(18):115-120.

LI Zhe, HAN Xuebing, LU Languang, et al.Temperature characteristics of power LiFePO 4 batteries[J ] . Journal of Mechanical Engineering, 2011, 47 (18 ):115-120.

LU Languang, HAN Xuebing, LI Jianqiu, et al.A review on the key issues for lithium-ion battery management in electric vehicles[J].Journal of Power Sources, 2013, 226:272-288.

FENG Xuning, OUYANG Minggao, LIU Xiang, et al.Thermal runaway mechanism of lithium ion battery for electric vehicles:a review[J].Energy Storage Materials, 2018, 10:246-267 .

王从飞,曹锋,李明佳,等.碳中和背景下新能源汽车热管理系统研究现状及发展趋势[J].科学通报, 2021, 66(32):4112-4128.

WANG Congfei, CAO Feng, LI Mingjia, et al.Research status and future development of thermal management system for new energy vehicles under the background of carbon neutrality[J].Chinese Science Bulletin, 2021, 66(32):4112-4128.

CHENG Hangyu, JUNG S, KIM Y B.Battery thermal management system optimization using deep reinforced learning algorithm[J].Applied Thermal Engineering, 2024, 236(Part C):121759.

张蕾,杨洋,马菁,等.液冷动力电池系统热管理控制策略优化探究[J].电源学报, 2024, 22 (S1 ):127-133.

ZHANG Lei, YANG Yang, MA Jing, et al.Optimization of thermal management control strategy for liquid-cooled power battery system[J].Journal of Power Supply, 2024, 22(S1):127-133.

CEN Jiwen, JIANG Fangming.Li-ion power battery temperature control by a battery thermal management and vehicle cabin air conditioning integrated system[J]. Energy for Sustainable Development, 2020, 57:141-148.

WEI C, HOFMAN T, CAARLS EI, et al.Zone model predictive control for battery thermal management including battery aging and brake energy recovery in electrified powertrains[J]. IFAC-PapersOnLine, 2019, 52(5):303-308.

HASKARA I, HEGDE B, CHANG Chenfang.Reinforcement learning based EV energy management for integrated traction and cabin thermal management considering battery aging[J].IFAC-PapersOnLine, 2022 , 55(24):348-353.

JOO S, LEE D, KIM M, et al.Multi-agent reinforcement learning based actuator control for EV HVAC systems[J].IEEE Access, 2023, 11:7574-7587.

XIE Yi, LIU Zhaoming, LI Kuining, et al.An improved intelligent model predictive controller for cooling system of electric vehicle[J].Applied Thermal Engineering, 2021, 182:116084.

HASAN A.Going below the wet-bulb temperature by indirect evaporative cooling:analysis using a modified ε -NTU method[J ] .Applied Energy, 2012, 89(1):237-245.

SHARMA K V, SUNDAR L S, SARMA P K.Estimation of heat transfer coefficient and friction factor in the transition flow with low volume concentration of Al 2 O 3 nanofluid flowing in a circular tube and with twisted tape insert[J ] .International Communications in Heat and Mass Transfer, 2009, 36(5):503-507.

VARJU E, POÓS T.New dimensionless equation for mass transfer at evaporation of open liquid surface in mixed or forced convection[J].International Journal of Heat and Mass Transfer, 2024, 227:125522.

JINDAL P, KATIYAR R, BHATTACHARYA J. Evaluation of accuracy for Bernardi equation in estimating heat generation rate for continuous and pulsedischarge protocols in LFP and NMC based Li-ion batteries[J].Applied Thermal Engineering, 2022, 201(Part A):117794.

封居强,司玉文,伍龙,等.基于动态综合型等效电路模型的动力电池特性分析[J].储能科学与技术, 2020, 9(3):986-992.

FENG Juqiang, SI Yuwen, WU Long, et al.Analysis of dynamic battery characteristics based on dynamic synthesis equivalent circuit model[J].Energy Storage Science and Technology, 2020, 9(3):986-992.

JOHNSON V H.Battery performance models in ADVISOR[J].Journal of Power Sources, 2002, 110(2):321-329.

杨鹏.基于模型预测控制的电动汽车整车热管理控制策略研究[D].重庆:重庆大学, 2022.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

应用深度强化学习的电池热管理系统控制策略

面向飞行器智能协同控制的分层双时延策略梯度强化学习方法

不同跨临界CO2直冷系统架构下动力电池冷却系统性能分析

联合遗传算法和强化学习的虚拟网络功能映射与调度方法

锂电池冷板新波形微流道传热特性研究及结构优化