【目的】 针对光储直柔（photovoltaic-storage direct current flexible，PSDF）微电网中可再生能源出力随机性与负荷响应延迟性带来的调度难题，提出一种考虑需求响应延迟特性的随机优化调度策略。【方法】 首先，构建包含能源运营商（energy producer， EP）与负荷聚合商（load aggregator， LA）的双层协同优化模型，引入可平移负荷、可转移负荷和可削减负荷3类柔性负荷的延迟响应特性，刻画用户侧实际调控行为。其次，基于纳什谈判理论建立非对称议价机制，综合考虑双方的议价能力差异，实现公平高效的电能交易定价。为应对风光出力与负荷的不确定性，采用蒙特卡罗场景生成与K-means聚类削减构建典型运行场景，并利用交替方向乘子法（alternating direction method of multipliers， ADMM）进行分布式求解。【结果】 仿真结果表明，所提模型可有效降低负荷聚合商成本1.17%，提升能源运营商收益4.07%，降低园区综合运行成本10.03%。【结论】 所提策略兼顾负荷侧经济性与系统功率平衡，实现了经济性、公平性与隐私性的统一，为高比例新能源接入下的微电网优化调度提供了可行方案。

[Objective] To address the dispatching challenges caused by the randomness of renewable energy output and the delay in demand response within photovoltaic-storage direct current flexible （PSDF） microgrids， this paper proposes a stochastic optimal dispatching strategy that explicitly considers demand response delay characteristics. [Methods] First， a two-layer coordinated optimization model covering the energy producer （EP） and the load aggregator （LA） is established， incorporating three types of flexible loads—shiftable， transferable， and reducible—with delay response characteristics to accurately capture realistic user-side control behaviors. Second， an asymmetric bargaining mechanism based on Nash bargaining theory is developed to account for differences in bargaining power between the two parties， thereby achieving fair and efficient electricity pricing. Furthermore， to handle the uncertainties in photovoltaic and wind power output as well as load demand， a Monte Carlo scenario generation combined with K-means clustering reduction is employed to construct typical operating scenarios. Finally， the alternating direction method of multipliers （ADMM） is applied for distributed optimization to ensure convergence and privacy preservation. [Results] Simulation results indicate that the proposed model effectively reduces the cost of the load aggregator by 1.17%， increases the profit of the energy producer by 4.07%， and decreases the overall operating cost of the microgrid by 10.03%. [Conclusions] The proposed strategy balances the economic performance of the load side and the system power equilibrium， achieving a unified improvement in economy， fairness， and privacy. It provides a practical and scalable solution for optimal dispatching of microgrids with high renewable energy penetration.