Cross-park computing-electricity bilateral matching driven by Bayesian reputation and diffusion-policy potential game

CHENG Lefeng; YU Feng; SUN Runbao; ZOU Tao

doi:10.3969/j.issn.2097-0706.2026.09.006

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Cross-park computing-electricity bilateral matching driven by Bayesian reputation and diffusion-policy potential game

Key Technology & Coordinated Control | 更新时间：2026-09-17

- Cross-park computing-electricity bilateral matching driven by Bayesian reputation and diffusion-policy potential game
- BLASTING Vol. 48, Issue 9, Pages: 68-83(2026)
- 作者机构：
  
  广州大学机械与电气工程学院，广州 510006
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China(52171331);Innovation Team Project for General Universities in Guangdong Province(2024KCXTD031);Guangzhou Municipal Education Bureau Higher Education Research Project(2024312278)
- DOI：10.3969/j.issn.2097-0706.2026.09.006
  CLC：
- Received：18 August 2026，
  
  Revised：2026-08-28，
  
  Online First：25 September 2026，
  
  Published：25 September 2026
- 稿件说明：
移动端阅览
程乐峰,余风,孙润宝等.贝叶斯声誉与扩散策略势博弈驱动的跨园区算力-电力双边匹配[J].综合智慧能源,2026,48(09):68-83.

CHENG Lefeng,YU Feng,SUN Runbao,et al.Cross-park computing-electricity bilateral matching driven by Bayesian reputation and diffusion-policy potential game[J].BLASTING,2026,48(09):68-83.
程乐峰,余风,孙润宝等.贝叶斯声誉与扩散策略势博弈驱动的跨园区算力-电力双边匹配[J].综合智慧能源,2026,48(09):68-83. DOI： 10.3969/j.issn.2097-0706.2026.09.006.

CHENG Lefeng,YU Feng,SUN Runbao,et al.Cross-park computing-electricity bilateral matching driven by Bayesian reputation and diffusion-policy potential game[J].BLASTING,2026,48(09):68-83. DOI： 10.3969/j.issn.2097-0706.2026.09.006.

摘要

“东数西算”格局下，跨区算力-电力交易在多个自治园区之间的组织面临多重相互交织的障碍，需构建一种可同时内生信任演化、保证匹配稳定并高效求解网络化博弈的跨园区交易组织机制。建立跨园区算力-电力双边匹配市场模型，将卖方履约可靠度刻画为不可观测的隐变量并与其绿色禀赋耦合，构造计及价格、时延、碳强度与声誉四维的成交综合效用。设计贝叶斯声誉演化机制：以Beta后验递推更新卖方声誉、以记忆保持因子实现历史证据的指数衰减，证明声誉得分几乎必然收敛至真实履约能力且后验方差以有效样本量的倒数速率收缩，并据此构造声誉加权的Gale-Shapley延迟接受双边匹配拍卖，证明其匹配结果不存在阻塞对。将卖方在互联网络上的报价激进度博弈建模为加权势博弈，以选择强度自适应的Fermi动力学求解，证明势函数沿演化轨迹单调上升并收敛至网络纳什均衡。提出图神经网络势博弈求解器与扩散策略生成式决策方法。结果显示，基于8买方-6卖方、24时段的跨区算例，声誉机制的平均稳态估计偏差为0.102，其中成交充分园区的偏差介于0.071~0.081，可靠还原卖方真实可信度。相较仅价格匹配基线，社会福利由0.410升至0.756（提升84.4%），平均碳强度由0.303降至0.272 t/（PFlops·h）（下降10.5%），而出清均价仅由517 元/（PFlops·h）变为522 元/（PFlops·h），表明增益源于对优质低碳卖方的精准甄别而非抬价；进一步计入网络化势博弈的完整机制将社会福利推升至0.880，其在社会福利、减排总量、抗操纵性、匹配公平与时延满足5个归一化维度上的性能包络完全包裹仅价格匹配、计及时延与仅计声誉3种配置，相对仅价格匹配基线实现帕累托改进。势函数沿Fermi动力学自341.9单调升至421.0；图神经网络求解器约10步即达95%收敛，较逐点Fermi迭代（约20步）实现约50%的收敛加速；园区数由6扩展至96时，其求解耗时的双对数斜率约1.15（近线性），而Fermi迭代约1.9（近平方），二者耗时差距扩大逾一个数量级。扩散策略的决策分布标准差达0.317，远高于确定性策略的0.020，呈显著双峰结构。鲁棒性方面：恶意虚报卖方占比由0升至50%时，上述机制的社会福利由0.822降至0.741，衰减仅9.8%，而朴素均值声誉由0.791降至0.358，衰减达54.8%；卖方新能源渗透率由30%升至90%时，机制带来的单位算力碳强度降幅由9.7%单调增至17.3%。故将信任的贝叶斯演化内生于匹配环节、将卖方策略耦合纳入势博弈框架、并以图学习与生成式决策加速均衡求解，可在不抬高出清价格的前提下同步改善跨园区算力-电力交易的社会福利、低碳性与抗操纵性；且该机制的减碳价值随电源侧清洁化程度的提高而放大、其求解优势随市场规模的扩大而放大，与“双碳”目标下的电源结构演进形成同向强化。所提机制与面向单园区的3层Stackelberg-演化-交替方向乘子法协同机制在研究对象、机制设计与求解方法上互为补充，共同构成算力-电力协同的完整方法体系。

Abstract

Cross-region computing-electricity trading among autonomous parks under China's "Eastern Data， Western Computing" initiative is obstructed by many coupled difficulties. A cross-park trading organization mechanism is demanded that endogenizes trust evolution， ensures matching stability， and efficiently solves the induced networked game. A cross-park computing-electricity bilateral matching market was established in which the sellers' fulfillment reliability was characterized as an unobservable latent variable coupled with their green endowment， and a composite transaction utility spanning price， latency， carbon intensity， and reputation was defined. A Bayesian reputation evolution mechanism was designed， where sellers' reputations were updated by Beta-posterior recursion， and a memory retention factor was introduced to achieve the exponential decay of historical evidence. It was proved that the reputation score converged almost surely to the actual fulfillment capability， and the posterior variance contracted at a rate inversely proportional to the effective sample size. Based on this， a reputation-weighted Gale-Shapley deferred-acceptance bilateral matching auction was constructed， proving that the matching outcome had no blocking pairs. The sellers' pricing-aggressiveness game over the interconnection network was modeled as a weighted potential game， which was solved by Fermi dynamics with adaptive selection intensity. The potential function was proved to increase monotonically along the evolutionary trajectory and converge to a network Nash equilibrium. A graph neural network（GNN）-based potential game solver and a diffusion-policy generative decision-making method were proposed. The results showed that， in an 8-buyer/6-seller， 24-period cross-region case， the mean steady-state estimation error of the reputation mechanism was 0.102， and the error ranged from 0.071 to 0.081 for the parks with sufficient trading frequency. This indicated that the mechanism reliably recovered sellers' true trustworthiness. Compared with the price-only matching baseline， social welfare rose from 0.410 to 0.756 （+84.4%） and the average carbon intensity fell from 0.303 to 0.272 t/（PFlops·h）（-10.5%）， whereas the clearing price shifted only from 517 to 522 yuan/（PFlops·h）. It indicated that the gain was mainly attributed to accurate identification of high-quality， low-carbon sellers rather than from price increases. Adding the networked potential game lifted welfare further to 0.880， with its performance envelope over the five normalized dimensions—social welfare， total emission reduction， manipulation resistance， matching fairness， and latency satisfaction—fully enclosing those of the price-only matching， latency-aware， and reputation-only configurations， thereby achieving a Pareto improvement over the price-only matching baseline. The potential function rose monotonically from 341.9 to 421.0 under Fermi dynamics. The GNN solver reached 95% convergence within about 10 iterations， realizing about 50% convergence acceleration compared with the pointwise Fermi iteration requiring approximately 20 iterations. As the number of parks increased from 6 to 96， its solution time scaled with a log-log slope of about 1.15 （nearly linear） versus about 1.9 （nearly quadratic） for Fermi， and the difference in solution time widened to more than one order of magnitude. The diffusion policy produced a decision distribution standard deviation of 0.317， substantially higher than the 0.020 obtained by the deterministic policy， exhibiting a pronounced bimodal structure. In terms of robustness， as the share of maliciously misreporting sellers increased from 0 to 50%， the social welfare achieved by the proposed mechanism decreased from 0.822 to 0.741， representing only a 9.8% reduction. In contrast， the naive mean reputation method experienced a decrease from 0.791 to 0.358， corresponding to a 54.8% reduction. As the renewable energy penetration rate of sellers increased from 30% to 90%， the reduction in carbon intensity per unit of computing power achieved by the proposed mechanism increased monotonically from 9.7% to 17.3%. Endogenizing Bayesian trust within matching， incorporating strategic coupling into a potential game framework， and accelerating equilibrium solving through graph learning and generative decision-making jointly improve the social welfare， low-carbon performance， and manipulation resistance of cross-park computing-electricity trading without raising clearing prices. Moreover， the carbon reduction value of the mechanism is amplified as the power generation mixes become cleaner， and its solution advantages become more prominent as markets grow larger， forming a synergistic reinforcement with the evolution of the power generation mixes under the "dual carbon" goals. The proposed mechanism complements the three-level Stackelberg-evolution-alternating direction method of multipliers coordination mechanism for a single park in terms of research objects， mechanism design， and solution methods， together forming a complete methodological framework for computing-electricity coordination.

关键词

Keywords

references

Electricity ， IEA . Electricity 2024： Analysis and forecast to 2026 ［R］. Paris ： International Energy Agency ， 2024 .

中国信息通信研究院 . 中国数字经济发展研究报告（2024年）［R］. 北京：中国信息通信研究院， 2024 .

WIERMAN A ， LIU Z H ， LIU I ， et al . Opportunities and challenges for data center demand response ［C］// Proceedings of International Green Computing Conference . IEEE ， 2015 ： 1 - 10 .

国家发展和改革委员会 . 国家发展改革委等部门关于同意京津冀地区启动建设全国一体化算力网络国家枢纽节点的复函（发改高技〔2022〕 212号）［Z］. 北京， 2022 .02.07. https：//www.gov.cn/zhengce/zhengceku/2022-02/17/content_5674330.htm https://www.gov.cn/zhengce/zhengceku/2022-02/17/content_5674330.htm .

仝杰，丁浩，陈洪银，等 . 电力-算力协同：内涵、架构、技术与实践［J］. 中国电机工程学报， 2026 ， 46 （ 14 ）： 5687 - 5707 .

TONG Jie ， DING Hao ， CHEN Hongyin ， et al . Electricity-computing power synergy： Connotation， architecture， technologies， and practice ［J］. Proceedings of the CSEE ， 2026 ， 46 （ 14 ）： 5687 - 5707 .

陈颉，周子龙，张巧龙，等 . 基于时空关联定价机制的跨园区综合能源系统优化调度策略［J］. 山东电力技术， 2025 ， 52 （ 8 ）： 1 - 15 .

CHEN Jie ， ZHOU Zilong ， ZHANG Qiaolong ， et al . Optimization dispatch strategy for multi-park integrated energy system based on spatiotemporal correlation pricing mechanism ［J］. Shandong Electric Power ， 2025 ， 52 （ 8 ）： 1 - 15 .

娄伟，张征凯，赵学理，等 . 考虑用户差异化的园区综合能源系统多主体博弈方法［J］. 山东电力技术， 2026 ， 53 （ 4 ）： 26 - 40 .

LOU Wei ， ZHANG Zhengkai ， ZHAO Xueli ， et al . Multi‑agent game method for park integrated energy system considering differentiated characteristics of users ［J］. Shandong Electric Power ， 2026 ， 53 （ 4 ）： 26 - 40 .

PAUDEL A ， CHAUDHARI K ， LONG C ， et al . Peer‑to‑peer energy trading in a prosumer‑based community microgrid： A game‑theoretic model ［J］. IEEE Transactions on Industrial Electronics ， 2019 ， 66 （ 8 ）： 6087 - 6097 .

MORSTYN T ， MCCULLOCH M D . Multiclass energy management for peer‑to‑peer energy trading driven by prosumer preferences ［J］. IEEE Transactions on Power Systems ， 2019 ， 34 （ 5 ）： 4005 - 4014 .

曾卓为，邓叶恒，王晓鹏，等 . 计及设备灵活运行与储能配置的园区综合能源系统低碳经济调度［J］. 电力工程技术， 2026 ， 45 （ 7 ）： 62 - 70， 128 .

ZENG Zhuowei ， DENG Yeheng ， WANG Xiaopeng ， et al . Low‑carbon economic dispatch of park‑level integrated energy system with flexible equipment operation and energy storage configurations ［J］. Electric Power Engineering Technology ， 2026 ， 45 （ 7 ）： 62 - 70， 128 .

TUSHAR W ， SAHA T K ， YUEN C ， et al . Peer‑to‑peer trading in electricity networks： An overview ［J］. IEEE Transactions on Smart Grid ， 2020 ， 11 （ 4 ）： 3185 - 3200 .

MORSTYN T ， TEYTELBOYM A ， MCCULLOCH M D . Bilateral contract networks for peer‑to‑peer energy trading ［J］. IEEE Transactions on Smart Grid ， 2019 ， 10 （ 2 ）： 2026 - 2035 .

SORIN E ， BOBO L ， PINSON P . Consensus‑based approach to peer‑to‑peer electricity markets with product differentiation ［J］. IEEE Transactions on Power Systems ， 2019 ， 34 （ 2 ）： 994 - 1004 .

BELGIOIOSO G ， ANANDUTA W ， GRAMMATICO S ， et al . Operationally‑safe peer‑to‑peer energy trading in distribution grids： A game‑theoretic market‑clearing mechanism ［J］. IEEE Transactions on Smart Grid ， 2022 ， 13 （ 4 ）： 2897 - 2907 .

KHORASANY M ， PAUDEL A ， RAZZAGHI R ， et al . A new method for peer matching and negotiation of prosumers in peer‑to‑peer energy markets ［J］. IEEE Transactions on Smart Grid ， 2021 ， 12 （ 3 ）： 2472 - 2483 .

RAJA AALI ， GRAMMATICO S . Bilateral peer‑to‑peer energy trading via coalitional games ［J］. IEEE Transactions on Industrial Informatics ， 2023 ， 19 （ 5 ）： 6814 - 6824 .

TALARI S ， KHORASANY M ， RAZZAGHI R ， et al . Mechanism design for decentralized peer‑to‑peer energy trading considering heterogeneous preferences ［J］. Sustainable Cities and Society ， 2022 ， 87 ： 104182 .

RAMCHURN S D ， HUYNH D ， JENNINGS N R . Trust in multi‑agent systems ［J］. The Knowledge Engineering Review ， 2004 ， 19 （ 1 ）： 1 - 25 .

WANG Y ， VASSILEVA J . Trust and reputation model in peer‑to‑peer networks ［C］// Proceedings of Proceedings Third International Conference on Peer‑to‑Peer Computing （P2P）2003 . IEEE ， 2003 ： 150 - 157 .

刘景矿，张欣怡，汪振双，等 . 数字经济赋能下建筑废弃物资源化产业数字化转型多主体行为策略研究［J］. 广州大学学报（自然科学版）， 2025 ， 24 （ 6 ）： 38 - 56 .

LIU Jingkuang ， ZHANG Xinyi ， WANG Zhenshuang ， et al . Multi‑agent behavior strategy of digital transformation of construction waste recycling industry enabled by digital economy ［J］. Journal of Guangzhou University （Natural Science Edition）， 2025 ， 24 （ 6 ）： 38 - 56 .

刘雷，栾佳美，辛媛媛 . 碳配额交易下智能制造系统的微分博弈［J/OL］. 工业工程， 1 - 12 （ 2026-07-23 ）［ 2026-08-27 ］. https：//link.cnki.net/urlid/44.1429.TH.20260722. 1543.002 https://link.cnki.net/urlid/44.1429.TH.20260722.1543.002 .

LIU Lei ， LUAN Jiamei ， YIN Yuanyuan . Differential game research of intelligent manufacturing system under carbon quota trading ［J］. Industrial Engineering Journal ， 1 - 12 （ 2026-07-23 ）［ 2026-08-27 ］. https：//link.cnki.net/urlid/44.1429.TH.20260722.1543.002 https://link.cnki.net/urlid/44.1429.TH.20260722.1543.002 .

CHENG L F ， YU F ， HUANG P R ， et al . Game‑theoretic evolution in renewable energy systems： Advancing sustainable energy management and decision optimization in decentralized power markets ［J］. Renewable and Sustainable Energy Reviews ， 2025 ， 217 ： 115776 .

程乐峰，孙润宝，倪曼琦，等 . 演化博弈与联邦学习融合驱动的新型电力系统多元主体协同优化［J］. 发电技术， 2026 ， 47 （ 3 ）： 449 - 483 .

CHENG Lefeng ， SUN Runbao ， NI Manqi ， et al . Multi‑agent collaborative optimization in new power systems driven by integration of evolutionary game and federated learning ［J］. Power Generation Technology ， 2026 ， 47 （ 3 ）： 449 - 483 .

程乐峰，刘奕杭，邹涛 . 博弈论视角下智能电网需求响应研究综述［J］. 综合智慧能源， 2025 ， 47 （ 4 ）： 1 - 22 .

CHENG Lefeng ， LIU Yihang ， ZOU Tao . Review of demand response in smart grids from the perspective of game theory ［J］. Integrated Intelligent Energy ， 2025 ， 47 （ 4 ）： 1 - 22 .

窦迅，王俊，王湘艳，等 . 基于演化博弈的区域电-气互联综合能源系统用户需求侧响应行为分析［J］. 中国电机工程学报， 2020 ， 40 （ 12 ）： 3775 - 3785 .

DOU Xun ， WANG Jun ， WANG Xiangyan ， et al . Analysis of user demand side response behavior of regional integrated power and gas energy systems based on evolutionary game ［J］. Proceedings of the CSEE ， 2020 ， 40 （ 12 ）： 3775 - 3785 .

欧阳海滨，陈浩楠，施金豆，等 . 强化学习与代理模型辅助的DE算法研究及其在桁架结构中的应用［J］. 广州大学学报（自然科学版）， 2026 ， 25 （ 2 ）： 35 - 48 .

OUYANG Haibin ， CHEN Haonan ， SHI Jindou ， et al . An adaptive differential evolution with reinforcement learning and surrogate modeling for truss structure optimization ［J］. Journal of Guangzhou University（Natural Science Edition）， 2026 ， 25 （ 2 ）： 35 - 48 .

马帅印，牛晨阳，栾龙源，等 . 人工智能赋能算电协同的关键技术［J/OL］. 综合智慧能源， 1 - 13 （ 2026-08-26 ）［ 2026-08-27 ］. https：//link.cnki.net/urlid/41.1461.TK.20260826.1007.002 https://link.cnki.net/urlid/41.1461.TK.20260826.1007.002 .

MA Shuaiyin ， NIU Chenyang ， LUAN Longyuan ， et al . Key technologies of AI‑empowered computing-electricity coordination ［J］. Integrated Intelligent Energy ， 1 - 13 （ 2026-08-26 ）［ 2026-08-27 ］. https：//link.cnki.net/urlid/41.1461.TK.20260826.1007.002 https://link.cnki.net/urlid/41.1461.TK.20260826.1007.002 .

JANNER M ， DU Y ， TENENBAUM J ， et al . Planning with diffusion for flexible behavior synthesis ［C］// Proceedings of the 39th International Conference on Machine Learning （ICML） . PMLR 162 ， 2022 ： 9902 - 9915 .

WANG Z ， HUNT J ， ZHOU M . Diffusion policies as an expressive policy class for offline reinforcement learning ［J］. arXiv preprint arXiv： 2208. 06193 . 2022 .

CHI C ， XU Z J ， FENG S Y ， et al . Diffusion policy： Visuomotor policy learning via action diffusion ［J］. The International Journal of Robotics Research ， 2025 ， 44 （ 10/11 ）： 1684 - 1704 .

ZHU Z B ， LIU M H ， MAO L Y ， et al . MADiff： offline multi‑agent learning with diffusion models ［C］// Proceedings of Advances in Neural Information Processing Systems 37 . Neural Information Processing Systems Foundation， Inc. （NeurIPS）， 2024 ： 4177 - 4206 .

MA H ， CHEN T ， WANG K ， et al . Efficient online reinforcement learning for diffusion policy ［J］. arXiv preprint arXiv： 2502. 00361 . 2025 .

MONDERER D ， SHAPLEY L S . Potential games ［J］. Games and Economic Behavior ， 1996 ， 14 （ 1 ）： 124 - 143 .

SUN Y B ， LIU T ， ZHOU R D ， et al . Provably fast convergence of independent natural policy gradient for Markov potential games ［C］// Proceedings of Advances in Neural Information Processing Systems 36 . Neural Information Processing Systems Foundation， Inc. （NeurIPS）， 2023 ： 43951 - 43971 .

KIPF T N ， WELLING M . Semi‑supervised classification with graph convolutional networks ［EB/OL］. 2016 ： arXiv ： 1609 . 02907 . https：//arxiv.org/abs/1609.02907 https://arxiv.org/abs/1609.02907 .

董成祥，魏昕，张坤鹏，等 . 基于图卷积网络的乘客打车需求预测［J］. 工业工程， 2022 ， 25 （ 5 ）： 98 - 105 .

DONG Chengxiang ， WEI Xin ， ZHANG Kunpeng ， et al . Passenger ride‑hailing demand prediction based on graph convolutional networks ［J］. Industrial Engineering Journal ， 2022 ， 25 （ 5 ）： 98 - 105 .

唐玉萍，付英姿，丁烨青 . 基于变分贝叶斯框架下的动态网络演化研究［J］. 昆明理工大学学报（自然科学版）， 2026 ， 51 （ 1 ）： 223 - 230 .

TANG Yuping ， FU Yingzi ， DING Yeqing . Dynamic network evolution based on variational Bayesian framework ［J］. Journal of Kunming University of Science and Technology （Natural Science）， 2026 ， 51 （ 1 ）： 223 - 230 .

徐会杰，马静珂 . 认知异质与心理演化耦合下的网络谣言传播动力学建模与分析［J/OL］. 控制理论与应用， 1 - 13 （ 2026-06-29 ）［ 2026-08-27 ］. https：//link.cnki.net/urlid/44.1240.tp.20260626.1633.024 https://link.cnki.net/urlid/44.1240.tp.20260626.1633.024 .

XU Huijie ， MA Jingke . Modeling and analysis of online rumor propagation dynamics under a framework coupling cognitive heterogeneity and psychological evolution ［J/OL］. Control Theory & Applications ， 1 - 13 （ 2026-06-29 ）［ 2026-08-27 ］. https：//link.cnki.net/urlid/44.1240.tp.2026 0626.1633.024 https://link.cnki.net/urlid/44.1240.tp.20260626.1633.024 .

程乐峰，邹涛，倪曼琦，等 . 演化博弈视角下需求侧参与电力市场的策略均衡稳定性分析综述［J］. 电力工程技术， 2025 ， 44 （ 3 ）： 3 - 17， 96 .

CHENG Lefeng ， ZOU Tao ， NI Manqi ， et al . A review on the strategy equilibrium stability analysis of demand-side participation in the electricity market from an evolutionary game perspective ［J］. Electric Power Engineering Technology ， 2025 ， 44 （ 3 ）： 3 - 17， 96 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Deep reinforcement learning method for energy-computing coordinated scheduling in integrated energy systems considering flexible data center tasks

Review of flexibility and adjustable potential quantification of transportation and computing loads for green power accommodation

Low-carbon economic scheduling for multiple data centers considering spatiotemporal migration characteristics of computing loads

Theoretical foundation， system architecture， and operational mode of integration of computing infrastructure and new power systems

Coordinated capacity optimization of source-grid-load-storage-computing systems in green computing parks in Xinjiang

Related Author

QU Xinmei

ZHONG Junjie

ZHAO Jinjin

SUN Kanghong

ZHANG Zhen

CAO Yijia

ZHU Hongyu

ZHANG Rui

Related Institution

State Key Laboratory for Grid Disaster Prevention and Mitigation，Changsha University of Science and Technology

School of Electrical and Information Engineering，Changsha University of Science and Technology

School of Engineering， Taylor's University

School of New Energy， Inner Mongolia University of Technology

Key Laboratory of Wind Energy and Solar Energy Utilization Technology， Ministry of Education, Inner Mongolia University of Technology

AI脑图

AI问答

⁰