电工技术学报  2024, Vol. 39 Issue (23): 7481-7497    DOI: 10.19595/j.cnki.1000-6753.tces.231896
电力系统与综合能源 |
基于计算热点转移的5G车联网能量实时协同管理策略
陈恺1, 付宇2, 孙毅1, 李跃2, 杨泓玥1
1.华北电力大学电气与电子工程学院 北京 102206;
2.贵州电网有限责任公司电力科学研究院 贵阳 550005
Online Energy Management Strategy for 5G-Vehicle Network Based on Computing Hotspot Transfer
Chen Kai1, Fu Yu2, Sun Yi1, Li Yue2, Yang Hongyue1
1. School of Electrical and Electronic Engineering North China Electric Power University Beijing 102206 China;
2. Electric Power Research Institute of Guizhou Power Grid Co. Ltd Guiyang 550005 China
全文: PDF (2831 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 由于5G基站与电动汽车规模持续增长,亟待开展能量管理,有利于减碳降本,缓解电网压力。但车联网的信息与能量高度耦合,使得基站独立管理能量易恶化车主的通信质量,且基站空调热负荷内部转移能力有限,电动汽车的车载算力调度潜力亟待发掘。为此,该文首先提出联合电动汽车充放电、5G网络与空调负荷和储能调度的能量管理模型,构建车联网时间平均用能成本最小化目标函数;其次,考虑网络状态不确定性,利用改进Lyapunov优化将随机优化问题转换为确定性问题;再次,利用车载计算零热点特性,提出基于主从博弈的计算热点转移策略,此外,考虑电网通信安全需求,提出融合启发式与连续凸近似的求解方法;最后,仿真表明,所提方法与现有实时能量管理方法相比,可有效削减车联网整体用能成本。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
陈恺
付宇
孙毅
李跃
杨泓玥
关键词 5G边缘计算电动汽车能量管理Lyapunov优化热点转移主从博弈    
Abstract:With the development of smart vehicle network promoted by China, the number of electric vehicle (EV) and 5G base station (BS) is increasing rapidly, while also bringing huge operation pressure and energy consumption to power grid and network operator, respectively. Therefore, how to coordinate the flexible resource of EV and BS to reduce electric load peak for power grid are important issues. At the same time, as EV equipping with powerful chip, BS can transfer some tasks to idle parked EV for cut down energy consumption of BS. However, due to the communication of EV is coupling with 5G BS, additional communication delay caused by BS energy management will affect the ordered charging scheduling signal between power grid and EV, increasing charging cost of EV. In addition, if 5G BS provided low delay communication for EV, it will bring more energy consumption. Moreover, the optimization of air condition load of BS is still lack of attention, which energy consumption is related with computing hotspot. To this end, this paper establishes a Stackelberg game based online energy management strategy faced to 5G vehicle network, aiming at minimizing the time average battery charging cost of EV and electricity purchasing cost of BS.
First of all, considering vehicle computing characteristics, and the coupling between hot load of air condition and computing tasks, a task migration model between 5G BS and EV is established to transfer the hot load and electric load of BS. Secondly, according to the analysis of communication delay to EV ordered charging, this paper proposes a calculation method of EV real charging capacity considering 5G wireless communication delay, and a communication delay penalty function for 5G network operator. Thirdly, based on above model, considering the stochastic events including task arrival and renewable energy, this paper investigates a long-term average electricity purchasing cost of BS and EV charging cost minimizing problem. Finally, an improved Lyapunov optimization method is proposed to transform the problem to a real time optimization problem. The real time problem can be divided into one upper layer subproblem and two lower layer subproblems, including computing electricity and battery (dis)charging problem, EV charging and communication resource allocation problem, and Stackelberg game-based computing hotspot transferring and BS air condition load control problem, respectively. By introducing an incentive function for EV accepting tasks of BS, the existence of the Nash equilibrium is proved, and this paper gives a close form solution on computing hotspot transferring. Meanwhile, other subproblems can be solved by successive convex approximation method.
In this paper, four scenarios are designed for verifying the performance of our proposed strategy, and analyze the impact of network scale, random task traffic, control parameter, and EV computing capacity on EV charging and BS electricity cost optimization. The simulation results show that proposed method can reduce electricity purchasing cost for both EV and BS without prior knowledge, and the hot load of BS is cut down without adding extra charging cost of parked EV. The following conclusions can be drawn from the simulation analysis: (1) The cooperation between EV and 5G BS can optimize the distribution of energy and information flow, reduce the energy consumption caused by cooling computing hotspot of BS. (2) By introducing improved Lyapunov optimization method, the network resource and energy scheduling of 5G BS and EV can be made online, and the resource scheduling result will not break the backup energy constraint and communication quality.
Key words5G edge computing    electric vehicle    energy management    Lyapunov optimization    hotspot transfer    Stackelberg game   
收稿日期: 2023-11-04     
PACS: TM93  
基金资助:贵州电网有限责任公司科技资助项目(GZKJXM20222149)
通讯作者: 陈 恺 男,1997年生,博士研究生,研究方向为能源互联网,电力信息与通信技术等。E-mail:kulewubi@163.com   
作者简介: 付 宇 男,1983年生,高级工程师,硕士,研究方向为配网运行分析。E-mail:1641356033@qq.com
引用本文:   
陈恺, 付宇, 孙毅, 李跃, 杨泓玥. 基于计算热点转移的5G车联网能量实时协同管理策略[J]. 电工技术学报, 2024, 39(23): 7481-7497. Chen Kai, Fu Yu, Sun Yi, Li Yue, Yang Hongyue. Online Energy Management Strategy for 5G-Vehicle Network Based on Computing Hotspot Transfer. Transactions of China Electrotechnical Society, 2024, 39(23): 7481-7497.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.231896          https://dgjsxb.ces-transaction.com/CN/Y2024/V39/I23/7481