| 163 | 0 | 15 |
| 下载次数 | 被引频次 | 阅读次数 |
【背景】电动汽车的快速普及使充电站间的空间竞争效应日益显著，充电站占用率预测对电网调度与用户体验至关重要。与传统充电总负荷预测相比，占用率更直接地反映了用户的离散选择行为，受预测站与周边站点价格、距离等多源时空因素耦合影响。【目标】有效捕捉多源信息复杂的时空竞争关系，实现更高效的数据融合。【方法】提出一种基于多源条件去噪扩散概率模型的充电站占用率预测方法，整合预测站历史占用率、动态电价、气象数据及经指数衰减加权的周边充电站价格信息，通过门控融合机制融合动态和静态特征，并设计中位数先验微调策略以提升概率预测区间的质量。【数据】利用公开UrbanEV数据集进行实验。【结果】多源条件去噪扩散概率模型在处理多源异构数据和复杂空间竞争关系时显著优于自回归积分移动平均模型、长短期记忆网络、Transformer、时空图卷积网络及去噪扩散概率模型等基线模型，有效克服了传统方法在处理多源异构数据与复杂空间竞争关系时的局限性。【应用】本研究为精准预测充电站占用率提供了创新解决方案，有助于优化电网调度与提升用户充电体验。
Abstract：[Background] The rapid growth of electric vehicles has intensified spatial competition among charging stations, making accurate occupancy rate prediction essential for grid scheduling and improving user experience. Compared with conventional forecasting of the total charging load,occupancy rates better capture users' discrete choice behavior, which is shaped by multi-source spatiotemporal factors, such as pricing and the distance between the target station and surrounding locations. [Objective] This study aims to effectively capture complex spatiotemporal competitive relationships across heterogeneous information sources and enable more efficient data fusion. [Method]This study proposes a charging station occupancy prediction approach based on a multi-source conditional denoising diffusion probability model. The model integrates the target station's historical occupancy data, dynamic electricity pricing, meteorological information, and exponentially weighted pricing information from surrounding stations. A gated fusion mechanism is used to combine dynamic and static features, and a median-prior fine-tuning strategy is incorporated to improve the quality of the predicted probability intervals. [Data] Experiments were conducted using the publicly available UrbanEV dataset. [Result] The experimental results show that the proposed multi-source conditional denoising diffusion probabilistic model significantly outperforms the autoregressive integrated moving average model, long short-term memory networks, transformers, spatiotemporal graph convolutional networks, and denoising diffusion probabilistic models. This improvement addresses the limitations of traditional approaches in handling multisource, heterogeneous data and complex spatial competition. [Application] The proposed method provides an effective solution for accurate charging station occupancy prediction, which can support grid scheduling and improve the user charging experience.
[1]IEA, Electric vehicles initiative, cleanecergy, ministe rial,Glabalev outlook 2022.[EB/OL].(2022-05-31)[2025-08-10]. https://www.iea.org/reports/global-ev-outlook-2022.
[2]REN Y, SUN X, WOLFRAM P, et al. Hidden delays of climate mitigation benefits in the race for electric vehicle deployment[J]. Nature Communications, 2023, 14:3164.
[3]FANG C, LU H, HONG Y, et al. Dynamic pricing for electric vehicle extreme fast charging[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22(1):531-541.
[4]WANG H, JIA Y, SHI M, et al. A mutually beneficial operation framework for virtual power plants and electric vehicle charging stations[J]. IEEE Transactions on Smart Grid, 2023, 14(6):4634-4648.
[5]LI S, XIONG H, CHEN Y. DiffPLF:a conditional diffusion model for probabilistic forecasting of EV charging load[J]. Electric Power Systems Research, 2024, 235:110723.
[6]CROZIER C, MORSTYN T, MCCULLOCH M. The opportunity for smart charging to mitigate the impact of electric vehicles on transmission and distribution systems[J]. Applied Energy, 2020, 268:114973.
[7]ALIZADEH M, WAI H T, CHOWDHURY M, et al. Optimal pricing to manage electric vehicles in coupled power and transportation networks[J]. IEEE Transactions on Control of Network Systems, 2017, 4(4):863-875.
[8]ZHAO M, WANG D, LI W, et al. Unraveling influencing factors of public charging station utilization[J]. Transportation Research Part D:Transport and Environment,2024, 137:104506.
[9]LIU C, PENG Z, LIU L, et al. Analysis of spatiotemporal characteristics and influencing factors of electric vehicle charging based on multisource data[J]. ISPRS International Journal of Geo-Information, 2024, 13(2):37.
[10]GALEANO-SUÁREZ D, TOQUICA D, HENAO N,et al. Impact of distribution locational marginal pricing and cost-sharing pricing mechanisms on fairness, efficiency, and voltage quality in transactive energy systems[J]. Utilities Policy, 2025, 93:101890.
[11]ARIAS M B, BAE S. Electric vehicle charging demand forecasting model based on big data technologies[J]. Applied Energy, 2016, 183:327-339.
[12]QU H, KUANG H, WANG Q, et al. A physics-informed and attention-based graph learning approach for regional electric vehicle charging demand prediction[J]. IEEE Transactions on Intelligent Transportation Systems,2024, 25(10):14284-14297.
[13]WANG S, LI Y, SHAO C, et al. An adaptive spatio-temporal graph recurrent network for short-term electric vehicle charging demand prediction[J]. Applied Energy,2025, 383:125320.
[14]XU X, WU J, LU Y, et al. A spatio-temporal prediction approach for charging load of clustered electric vehicles in dynamic traffic flow environment of highway[J]. Sustainable Energy, Grids and Networks, 2024, 40:101593.
[15]苏粟,王建祥,王磊,等.基于动态哈夫模型及双边匹配的电动汽车充电引导策略[J].电力系统自动化,2024, 48(7):181-189.SU Su, WANG Jianxiang, WANG Lei, et al. Guidance strategy for electric vehicle charging based on dynamic huff model and bilateral matching[J]. Automation of Electric Power Systems, 2024, 48(7):181-189.
[16]胡晓伟,吴则洋,卢泓博,等.基于遗传算法优化深度神经网络的站点客流预测[J].交通运输工程与信息学报, 2025, 23(1):72-84.HU Xiaowei, WU Zeyang, LU Hongbo, et al. Passenger flow prediction for stations using genetic algorithmoptimized deep neural networks[J]. Journal of Transportation Engineering and Information, 2025, 23(1):72-84.
[17]易术,黄丹阳.融合多源数据与元胞传输模型的高速公路交通状态估计方法[J].交通运输工程与信息学报, 2023, 21(4):103-114.YI Shu, HUANG Danyang. Freeway traffic state estimation based on multi-source data and cell transmission model[J]. Journal of Transportation Engineering and Information, 2023, 21(4):103-114.
[18]AHSAN M M, RAMAN S, LIU Y, et al. A comprehensive survey on diffusion models and their applications[J]. Applied Soft Computing, 2025, 181:113470.
[19]RASUL K, SEWARD C, SCHUSTER I, et al. Autoregressive denoising diffusion models for multivariate probabilistic time series forecasting[C]//38th International Conference on Machine Learning, Proceedings of Machine Learning Research. PMLR. 2021:8857-68
[20]RASUL K, SHEIKH AS, SCHUSTER I, et al.Multi-variate probabilistic time series forecasting via conditioned normalizing flow[DB/OL].(2021-01-13)[2025-08-10].https://openreview.net/forum id=WiGQBFuVRv.
[21]TASHIRO Y, SONG J, SONG Y, et al. CSDI:conditional score-based diffusion models for probabilistic time series imputation[DB/OL].(2021-11-10)[2025-08-12]. https://openreview.net/forum id=VzuIzbRDrum.
[22]ROMBACH R, BLATTMANN A, LORENZ D, et al.High-resolution image synthesis with latent diffusion models[C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). New Orleans,USA:IEEE, 2022:10674-10685.
[23]SONG Y, SOHL-DICKSTEIN J, KINGMA D P, et al.Score-based generative modeling through stochastic differential equations[EB/OL].(2021-02-10)[2025-08-12].https://arxiv.org/abs/2011.13456.
[24]HO J, JAIN A, ABBEEL P. Denoising diffusion probabilistic models[EB/OL].(2020-06-19)[2025-08-10]. https://arxiv.org/abs/2006.11239.
[25]LI H, QU H, TAN X, et al. UrbanEV:an open benchmark dataset for urban electric vehicle charging demand prediction[J].Scientific Data, 2025, 12:523.
基本信息:
DOI：10.19961/j.cnki.1672-4747.2025.09.024
中图分类号:TM910.6;U491.8
引用信息:
[1]勉海荣,焦小刚,毕利.基于多源条件去噪扩散模型的电动车充电站占用率预测[J].交通运输工程与信息学报,2026,24(02):94-105.DOI:10.19961/j.cnki.1672-4747.2025.09.024.
基金信息:
国家自然科学基金项目（62266034）; 宁夏重点研发项目（引才专项）（2023BSB03015）; 宁夏大学研究生创新项目（CXXM2025-041）
2025-10-15
2025-10-15
2025-10-15