| 引用本文: | 李江,仝昊,张子硕.基于自适应 McCormick 方法的储能参与现货市场竞价策略[J].电力系统保护与控制,2026,54(10):103-114. |
| LI Jiang,TONG Hao,ZHANG Zishuo.Bidding strategy for energy storage participation in the spot market based on an adaptive McCormick method[J].Power System Protection and Control,2026,54(10):103-114 |
|
| 摘要: |
| 针对双结算机制下储能电站参与现货市场竞价时难以统筹日前计划与实时偏差电量结算，且双层模型因跨市场双线性项导致求解效率偏低的问题，提出一种基于自适应 McCormick 方法的储能参与现货市场竞价策略。首先，构建包含日前、实时市场出清及偏差电量结算的双层竞价模型，并在上层引入基于条件风险价值 (conditional value-at-risk, CvaR) 的收益风险权衡机制。然后，针对卡鲁什 - 库恩 - 塔克 (Karush-Kuhn-Tucker, KKT) 条件将双层模型单层化后跨日前、实时的双线性耦合项难以通过强对偶等价消除的问题，提出自适应分段 McCormick 线性化方法，通过动态松弛区间收缩与松弛区间回溯机制将原问题转化为可高效求解的混合整数线性规划。最后，在改进 IEEE39 节点系统上开展算例验证，结果表明所提策略在提升储能收益的同时兼顾求解效率与收敛稳定性。 |
| 关键词: 储能电站 电力现货市场 自适应 McCormick 方法 双层竞价模型 |
| DOI：10.19783/j.cnki.pspc.251084 |
| 分类号: |
| 基金项目:国家自然科学基金项目资助 (51977030) |
|
| Bidding strategy for energy storage participation in the spot market based on an adaptive McCormick method |
|
LI Jiang, TONG Hao, ZHANG Zishuo
|
|
Faculty of Electrical Engineering, Shanghai University of Electric Power, Shanghai 200090, China
|
| Abstract: |
| To address the challenges that, under a two-settlement mechanism, an energy storage power station participating in spot-market bidding is difficult to coordinate day-ahead schedules with real-time deviation-energy settlement, and that the associated bilevel model suffers from low computational efficiency due to cross-market bilinear terms, this paper proposes a bidding strategy for energy storage participation in the power spot market based on the adaptive McCormick method. First, a bilevel bidding model is established that incorporates day-ahead and real-time market clearing as well as deviation energy settlement, and a profit-risk trade-off mechanism based on conditional value-at-risk (CVaR) is introduced at the upper level. Then after reformulating the bilevel model into a single-level problem using the Karush-Kuhn-Tucker (KKT) conditions, the cross-market bilinear coupling terms between day-ahead and real-time markets cannot be eliminated through strong duality equivalence. To address this, an adaptive piecewise McCormick linearization method is developed, which transforms the original problem into a mixed-integer linear programming model through dynamic relaxation interval contraction and backtracking mechanisms, enabling efficient solution. Finally, case studies conducted on a modified IEEE39-bus system demonstrate that the proposed strategy improves the profitability of energy storage while achieving satisfactory computational efficiency and convergence stability. |
| Key words: energy storage station electricity spot market adaptive McCormick method bilevel bidding model |
