Data-driven distributionally robust scheduling method for park-level integrated energy system considering electric vehicle integration

Authors

• Chuangxin Guo, College of Electrical Engineering ， Zhejiang University ， Hangzhou 310027 ， China
• Kun Chen, College of Electrical Engineering ， Zhejiang University ， Hangzhou 310027 ， China
• Yumian Lin, College of Electrical Engineering ， Zhejiang University ， Hangzhou 310027 ， ChinaFollow
• Houbo Xiong, College of Electrical Engineering ， Zhejiang University ， Hangzhou 310027 ， China
• Lei Yan, College of Electrical Engineering ， Zhejiang University ， Hangzhou 310027 ， China
• Liwei Du, Fuzhou Power Supply Company ， State Grid Fujian Electric Power Co .， Ltd.， Fuzhou 350116 ， China

Keywords

park-level integrated energy system, electric vehicle, charging management, data-driven, distributionally robust optimization

Abstract

Fluctuations in renewable energy output and the growing and stochastic electric vehicle （EV） charging demand introduce source- and load-side uncertainties that significantly affect the scheduling plan of the park-level integrated energy system （PIES ）.To improve the economy and robustness of PIES operation，this paper proposes a data-driven distributionally robust scheduling method for PIES considering EV integration.First，given that the uncertainty associated with EV integration is difficult to characterize accurately，five features of EV charging behavior are selected.A Gaussian mixture model is used to characterize the multimodal behavior of EV charging，and an uncertainty set is constructed by combining Neyman sampling with comprehensive norm constraints.Second，a coordinated delayed and staggered charging management scheme is designed to achieve peak shaving and valley filling while meeting users ’ charging requirements.Then，EV charging management is incorporated into the economic scheduling framework of PIES to construct a data-driven distributionally robust optimization （DRO ） model.Finally，an improved column-and-constraint generation algorithm is proposed to solve the DRO scheduling problem.The case study results show that the proposed delayed charging management scheme smooths the PIES electric load curve，significantly increases the renewable energy accommodation rate，and reduces dependence on purchased electricity.Meanwhile，compared with a conventional two-stage robust optimization model，the proposed method reduces the total cost by 10.26%，verifying its feasibility and economic efficiency and providing a new pathway for efficient PIES scheduling under high EV penetration.

DOI

10.19781/j.issn.1673-9140.2026.04.018

First Page

223

Last Page

234

Recommended Citation

Guo, Chuangxin; Chen, Kun; Lin, Yumian; Xiong, Houbo; Yan, Lei; and Du, Liwei (2026) "Data-driven distributionally robust scheduling method for park-level integrated energy system considering electric vehicle integration," Journal of Electric Power Science and Technology: Vol. 41: Iss. 4, Article 18.
DOI: 10.19781/j.issn.1673-9140.2026.04.018
Available at: https://jepst.researchcommons.org/journal/vol41/iss4/18