|
| 基于DDPG和电价感知的中长期双边竞价策略优化模型 |
| Mid- to Long-Term Bilateral Bidding Strategy Optimization Using DDPG With Price Perception |
| 投稿时间：2025-08-21 修订日期：2025-09-03 |
| DOI： |
| 中文关键词: 中长期双边交易 多周期电价感知 DDPG算法 强化学习 策略优化 |
| 英文关键词: Mid- to Long-Term Bilateral Transactions Mul-ti-Period Price Perception Deep Deterministic Policy Gradient (DDPG) Reinforcement Learning Strategy Optimization |
| 基金项目:低碳高可靠城市配电系统示范工程 |
|
| 摘要点击次数: 0 |
| 全文下载次数: 0 |
| 中文摘要: |
| 随着电力市场改革的不断深化，用户侧参与中长期电力市场交易活跃度不断提升，其对电价变化的感知与响应行为已成为影响市场资源配置效率的重要因素。本文针对传统中长期竞价模型在刻画用户动态行为与应对连续策略优化方面的不足，提出了一种融合多周期电价感知机制与深度确定性策略梯度（DDPG）算法的发电商中长期双边竞价策略优化方法。首先，构建基于Sigmoid结构的多周期非线性电价感知模型，引入周期调节与不确定扰动因素，刻画用户在不同电价周期下的响应阈值与行为差异；其次，基于DDPG算法设计发电商的策略学习框架，并引入差异化奖励函数，兼顾利润、社会福利与市场公平性，实现连续动作空间下的策略自适应优化；最后，通过仿真实验验证所提模型在提升市场收益、促进负荷调节与实现多目标均衡方面的有效性。研究结果表明，本文方法能够更精准地模拟电力用户行为，优化发电商竞价策略，增强电力市场的运行稳定性与社会福利水平。 |
| 英文摘要: |
| With the deepening of electricity market reforms, the partic-ipation of end-users in mid- to long-term electricity market transactions has been steadily increasing. Their sensitivity and behavioral responses to price fluctuations have become critical factors influencing the efficiency of market resource alloca-tion. To address the limitations of conventional mid- to long-term bidding models in capturing dynamic user behavior and optimizing continuous strategies, this paper proposes a bilateral bidding strategy optimization method for generators that integrates a multi-period price-perception mechanism with the Deep Deterministic Policy Gradient (DDPG) algo-rithm. First, a multi-period nonlinear price-perception model with a Sigmoid structure is developed, incorporating cycle adjustments and stochastic disturbances to characterize users’ response thresholds and behavioral heterogeneity across dif-ferent pricing periods. Second, a strategy-learning framework for generators is designed based on the DDPG algorithm, where a differentiated reward function is introduced to sim-ultaneously account for profit, social welfare, and market fairness, thereby enabling adaptive optimization of strategies in continuous action spaces. Finally, simulation experiments demonstrate the effectiveness of the proposed model in im-proving market revenues, facilitating load regulation, and achieving multi-objective equilibrium. The results confirm that the proposed approach more accurately captures elec-tricity users’ behavioral responses, optimizes generator bid-ding strategies, and enhances both the operational stability and social welfare of electricity markets. |
|
View Fulltext
查看/发表评论 下载PDF阅读器 |
| 关闭 |