基于态势感知的配电网空间负荷预测系统  

Spatial load forecasting system for distribution network based on situational awareness

作　　者：周忠强 黄育松 梁铃 施诗 覃海 陈胜 ZHOU Zhong-qiang;HUANG Yu-song;LIANG Ling;SHI Shi;QIN Hai;CHEN Sheng(China Southern Power Grid Guizhou Power Grid Co.,Ltd.,Power Dispatching Control Center,Guiyang 550002,China)

机构地区：[1]中国南方电网贵州电网有限责任公司电力调度控制中心,贵阳550002

出　　处：《信息技术》2024年第3期116-121,共6页Information Technology

基　　金：中国南方电网有限责任公司科技研究项目(06650-0KK52180017);贵州电网有限责任公司科技研究项目(0665002019070305FS00009)。

摘　　要：针对动态地区配电网空间负荷优化预测问题,设计了一种基于实时自控态势感知的配电网空间负荷预测系统。首先构建配电网空间负荷预测目标决策要素经验池;然后利用深度长短期神经网络对配电网历史运行数据集进行处理,实现配电网实时自控态势感知;最后利用深度确定性策略梯度算法构建空间负荷预测与目标决策要素经验池之间的耦合模型,实现配电网空间负荷高效精准预测。对模型展开了实际算例分析,多类型配网供电区域场景下配电网空间负荷预测精度达到91.03%,多类型配网供电区域场景下配电网空间负荷预测效率提高了23.71%。To solve the optimization and forecasting problem of spatial load of distribution network in dynamic areas,the spatial load forecasting system of distribution network based on real-time self-control situational awareness is designed.Firstly,the experience pool of objective decision-making elements of distribution network spatial load forecasting is constructed.Then,the deep long-term and short-term neural network is used to process the historical operation data set of distribution network,so as to realize the real-time automatic situational awareness of distribution network.Finally,the coupling model between spatial load forecasting and the experience pool of target decision-making elements is constructed by using the deep deterministic strategy gradient algorithm to realize the efficient and accurate spatial load forecasting of distribution network.The actual example analysis of the model shows that the accuracy of distribution network spatial load forecasting under the scenario of multi type distribution network power supply area is 91.03%,and the efficiency of distribution network spatial load forecasting under the scenario of multi type distribution network power supply area is improved by 23.71%.

关 键 词：自控态势感知 配电网系统 空间负荷预测 限制映射关系 耦合预测模型 

分 类 号：TM714[电气工程—电力系统及自动化]