数据驱动的电力系统动态模型参数误差溯源方法  

Data-driven method for error tracing of dynamic model parameters in power system

作　　者：姜赫 安军[1] 刘征帆 李德鑫 Jiang He;An Jun;Liu Zhengfan;Li Dexin(Key Laboratory of Modern Power System Simulation and Control&Renewable Energy Technology of the Ministry of Education,Northeast Electric Power University,Jilin 132012,Jilin,China;Electric Power Research Institute of State Grid Jilin Electric Power Company,Changchun 130021,China)

机构地区：[1]东北电力大学现代电力系统仿真控制与绿色电能新技术教育部重点实验室,吉林吉林132012 [2]国网吉林省电力有限公司电力科学研究院,长春130021

出　　处：《电测与仪表》2022年第11期63-70,共8页Electrical Measurement & Instrumentation

基　　金：国家重点研发计划项目(2018YFB0904500);国家电网有限公司总部科技项目(18-GW-05)。

摘　　要：动态仿真中模型及参数的不准确直接导致仿真轨迹与实测轨迹不一致。电力系统区域大,包含元件众多,如何准确的识别出致差区域是一个难题。文章提出了一种数据驱动的大规模电力系统误差溯源方法,定义了扰动深度指标以及轨迹差异度指标,用以评价元件受扰程度和轨迹差异度,基于指标及其约束条件,给出了主要致差区域的识别方法,用于误差溯源。文章以IEEE 10机39节点仿真系统验证了所提方法的有效性,减少了动态仿真验证的工作代价,为分块解耦提供支持。The models and parameters of dynamic simulation components will lead to inconformity between measured trajectory and simulated trajectory.The power system has a large area and many components.Accurately identifying the error tracing area is a challenge.This paper proposed a data-driven method for error tracing in large-sacle power system.It defines a disturbance depth indicator and the variance indicator to evaluate the disturbance degree and trajectory variance of components.The two indicators and their constraints help to identify the error tracing area.This paper verifies the effectiveness of the proposed method by an IEEE 10-machine 39-node simulation system.The proposed method reduces cost of dynamic simulation validation task and support the block decoupling process.

关 键 词：电力系统 仿真验证 扰动深度 差异度指标 致差区域 

分 类 号：TM743[电气工程—电力系统及自动化]