首页|期刊导航|现代电力|基于数据驱动响应约束的配电网可靠性评估方法

基于数据驱动响应约束的配电网可靠性评估方法OA

A Data-driven Response Constraint-based Reliability Assessment Method for Distribution Networks

中文摘要英文摘要

为量化具有能流互济的综合能源系统(integrated energy system,IES)接入对配电网可靠性的影响,提出一种基于数据驱动响应约束的配电网-综合能源系统可靠性评估方法.首先,分析综合能源系统接入对配电网可靠性的影响,提出配电网-综合能源系统可靠性评估框架;建立基于数据驱动的 IES优化运行模型,采用深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法实现综合能源系统最优经济调度;基于大 M法和分段线性化法将训练好的 DDPG模型线性化,构建数据驱动响应约束;最后结合数据驱动响应约束,以配电网失负荷量最少为目标函数,建立混合整数线性规划的配电网最优切负荷模型,通过 IEEE RBTS-BUS6算法验证其有效性.算例表明,所提模型能够实现不完全信息下配电网的可靠性评估,IES通过能流互补为配电网提供可靠性支撑.

To quantify the impact of IES access with interconnected energy flows on the reliability of distribution networks,a data-driven response constraint-based reliability assessment method is proposed for distribution networks and IES.Firstly,the impact of IES access on the reliability of distribution networks is analyzed,and a reliability assessment framework is proposed for distribution networks and IES.A data-driven optimization operation model for IES is developed,and a deep deterministic policy gradient(DDPG)algorithm is utilized to achieve the optimal economic dispatch of the IES.Subsequently,the trained DDPG model is linearized using the large M method and segmented linearization method to construct data-driven response constraints.Finally,with minimizing load loss in the distribution network as the objective,the data-driven response constraints are combined to establish the optimal load-slicing model of the distribution network with mixed-integer linear programming,and its validity is verified through IEEE RBTS-BUS6 system.Examples demonstrate that the proposed model can realize the reliability assessment for distribution networks under the condition of incomplete information,while IES provides reliability support for distribution networks through interconnected energy flow.

张帅;刘文霞;刘佳怡;王雅姝;成锐

新能源电力系统国家重点实验室(华北电力大学),北京市 昌平区 102206新能源电力系统国家重点实验室(华北电力大学),北京市 昌平区 102206新能源电力系统国家重点实验室(华北电力大学),北京市 昌平区 102206新能源电力系统国家重点实验室(华北电力大学),北京市 昌平区 102206新能源电力系统国家重点实验室(华北电力大学),北京市 昌平区 102206

信息技术与安全科学

配电网综合能源系统可靠性评估深度学习数据驱动

distribution networkintegrated energy system(IES)reliability assessmentdeep learningdata driven

《现代电力》 2026 (3)

401-412,12

中国博士后科学基金资助项目(GZC20230785).Project Supported by China Postdoctoral Science Foundation(GZC20230785).

10.19725/j.cnki.1007-2322.2024.0042

评论