穆清, 程昱, 王春明, 汪旸, 严剑峰, 徐得超, 庞润甜. 基于分群解耦的电力系统机电电磁混合仿真的分网技术[J]. 现代电力, 2021, 38(4): 355-361. DOI: 10.19725/j.cnki.1007-2322.2020.0335
引用本文: 穆清, 程昱, 王春明, 汪旸, 严剑峰, 徐得超, 庞润甜. 基于分群解耦的电力系统机电电磁混合仿真的分网技术[J]. 现代电力, 2021, 38(4): 355-361. DOI: 10.19725/j.cnki.1007-2322.2020.0335
MU Qing, CHENG Yu, WANG Chunming, WANG Chang, YAN Jianfeng, XU Dechao, PANG Runtian. Power System Tearing Technology for EMT-TS Hybrid Simulation Based on Group Decoupled Interface Division[J]. Modern Electric Power, 2021, 38(4): 355-361. DOI: 10.19725/j.cnki.1007-2322.2020.0335
Citation: MU Qing, CHENG Yu, WANG Chunming, WANG Chang, YAN Jianfeng, XU Dechao, PANG Runtian. Power System Tearing Technology for EMT-TS Hybrid Simulation Based on Group Decoupled Interface Division[J]. Modern Electric Power, 2021, 38(4): 355-361. DOI: 10.19725/j.cnki.1007-2322.2020.0335

基于分群解耦的电力系统机电电磁混合仿真的分网技术

Power System Tearing Technology for EMT-TS Hybrid Simulation Based on Group Decoupled Interface Division

  • 摘要: 多条耦合直流输电的响应只能通过电磁暂态仿真程序模拟。为了提高仿真精度,机电-电磁混合仿真已用于电力系统运行的研究。为解决目前机电-电磁混合仿真速度太慢的问题,提出了一种电力系统混合仿真分网技术。首先把机电网络分成尽可能多的网络孤岛,把电磁暂态网络通过传输线解耦方式分网,然后基于分群解耦的机电-电磁混合仿真算法形成单时步内解耦的机电-电磁子网群。继而又提出一种新的并行程序结构,利用机电-电磁子网群实现并行仿真,加快仿真速度。最后对含有7条直流输电和4000个节点的电磁网络的国调方式计算案例进行的研究表明,该方法可以在不降低精度的情况下大大提高仿真速度。

     

    Abstract: The response for multiple coupled HVDCs can only be simulated through EMT programs. To increase the simulation accuracy, the EMT-TS Hybrid simulation has been used in studies of the power system operation. Currently, the EMT-TS hybrid simulation speed is too slow. This paper presents a novel power system tearing technology to greatly improve the simulation speed in EMT-ST network islands as many as possible, while the EMT network is teared through the combined transmission line. An algorithm is supposed to form the EMT-ST subnetwork groups, which has no coupling with each other. And the ST-EMT subnet group decoupled in a single time step base on this algorithm. Then an novel parallel program structure is proposed to use this EMT-ST subnetwork groups to realize the parallel simulation and speedup the simulation. The case studies of the grids with 7 HVDCs and 4000 nodes show that the proposed method could greatly improve the simulation speed without lowering precision.

     

/

返回文章
返回