基于信息缺口决策理论的车网互动聚合商日前定价策略

刘诗懿; 朱刚毅; 胡捷添; 李明莉; 周映宇; 向月; 刘俊勇

刘诗懿,朱刚毅,胡捷添,李明莉,周映宇,向月,刘俊勇.基于信息缺口决策理论的车网互动聚合商日前定价策略[J].电测与仪表,2026,63(5):79-90.
LIU Shiyi,ZHU Gangyi,HU Jietian,LI Mingli,Zhou Yingyu,XIANG Yue,LIU Junyong.Day-ahead pricing strategy for vehicle-to-grid aggregator based on info-gap decision theory[J].Electrical Measurement & Instrumentation,2026,63(5):79-90.

基于信息缺口决策理论的车网互动聚合商日前定价策略

Day-ahead pricing strategy for vehicle-to-grid aggregator based on info-gap decision theory

DOI：

中文关键词: 车网互动双重不确定性信息缺口决策理论需求响应日前定价

英文关键词:vehicle-to-grid dual uncertainties info-gap decision theory demand response day-ahead pricing

基金项目:贵州省管理创新项目（DW202500210284）；贵州电力交易中心管理科学项目（0682002025080202JY00006）；国家自然科学基金资助项目（52177103）

作者中文名	作者英文名	单位
刘诗懿	LIU Shiyi	贵州电力交易中心有限责任公司
朱刚毅	ZHU Gangyi	贵州电力交易中心有限责任公司
胡捷添	HU Jietian	贵州电力交易中心有限责任公司
李明莉	LI Mingli	贵州电力交易中心有限责任公司
周映宇	Zhou Yingyu	四川大学电气工程学院
向月	XIANG Yue	四川大学电气工程学院
刘俊勇	LIU Junyong	四川大学电气工程学院

摘要点击次数: 64

中文摘要:

在新型电力系统与电力市场深度融合的背景下，车网互动（Vehicle-to-Grid, V2G）聚合商在日前运营中同时面临系统侧结算电价波动和用户侧响应行为随机的双重不确定性。然而，现有研究多基于单一不确定源假设，难以在缺乏完整信息的条件下兼顾经济性与鲁棒性。针对这一问题，本文提出一种基于信息缺口决策理论（Info-Gap Decision Theory, IGDT）的V2G日前鲁棒定价策略，为聚合商在不完全信息环境下提供收益保障型的定价与报量决策支持。所提模型在收益最大化目标下，构建了耦合用户心理感知与收益预期的非线性参与行为机制，并利用McCormick包络将非确定性优化转化为混合整数线性规划以提高求解可行性。算例结果表明，聚合商与用户的收益分成比例存在最优区间，所提方法在不增加激励成本的前提下可显著降低偏差罚占比并提升净收益。该策略有效平衡了收益与风险，为V2G聚合商在信息不足环境下的稳健运营提供了新的决策框架。

英文摘要:

Against the backdrop of the rapid advancement of new power systems and electricity markets, Vehicle-to-Grid (V2G) aggregators face dual uncertainties in day-ahead operations, namely fluctuations in settlement prices on the system side and stochastic response behaviors on the user side. To address this, this paper proposes a robust V2G day-ahead pricing strategy based on Info-Gap Decision Theory (IGDT), aiming to provide reliable pricing and bidding decision support for aggregators under incomplete information. The strategy adopts a benefit-guaranteeing IGDT formulation with the objective of maximizing the aggregator’s expected profit, and incorporates a nonlinear participation behavior model that couples user psychological perception with revenue expectations. By employing McCormick envelopes, the complex non-deterministic optimization problem is reformulated into a solvable Mixed-Integer Linear Programming model. Case studies demonstrate that there exists an optimal range for the profit-sharing ratio between aggregators and users. Compared with multiple heuristic pricing methods, the proposed strategy can significantly reduce the share of deviation penalties and improve net profits without increasing incentive expenditures. Moreover, the aggregator’s prof-itability is found to be most sensitive to peak-period settlement prices, while the impacts of valley-period dispatch volumes and station capacity expansion exhibit diminishing marginal effects.

查看全文查看/发表评论下载PDF阅读器

关闭