Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Graph Embedding Dynamic Feature-based Supervised Contrastive Learning of Transient Stability for Changing Power Grid Topologies

Zijian Lv, Xin Chen, Zijian Feng

TL;DR

Test result demonstrated that the GEDF-SCL model can achieve high accuracy in transient stability prediction and adapt well to changing power grid topologies.

Abstract

Accurate online transient stability prediction is critical for ensuring power system stability when facing disturbances. While traditional transient stablity analysis replies on the time domain simulations can not be quickly adapted to the power grid toplogy change. In order to vectorize high-dimensional power grid topological structure information into low-dimensional node-based graph embedding streaming data, graph embedding dynamic feature (GEDF) has been proposed. The transient stability GEDF-based supervised contrastive learning (GEDF-SCL) model uses supervised contrastive learning to predict transient stability with GEDFs, considering power grid topology information. To evaluate the performance of the proposed GEDF-SCL model, power grids of varying topologies were generated based on the IEEE 39-bus system model. Transient operational data was obtained by simulating N-1 and N-$\bm{m}$-1 contingencies on these generated power system topologies. Test result demonstrated that the GEDF-SCL model can achieve high accuracy in transient stability prediction and adapt well to changing power grid topologies.

Graph Embedding Dynamic Feature-based Supervised Contrastive Learning of Transient Stability for Changing Power Grid Topologies

TL;DR

Test result demonstrated that the GEDF-SCL model can achieve high accuracy in transient stability prediction and adapt well to changing power grid topologies.

Abstract

Accurate online transient stability prediction is critical for ensuring power system stability when facing disturbances. While traditional transient stablity analysis replies on the time domain simulations can not be quickly adapted to the power grid toplogy change. In order to vectorize high-dimensional power grid topological structure information into low-dimensional node-based graph embedding streaming data, graph embedding dynamic feature (GEDF) has been proposed. The transient stability GEDF-based supervised contrastive learning (GEDF-SCL) model uses supervised contrastive learning to predict transient stability with GEDFs, considering power grid topology information. To evaluate the performance of the proposed GEDF-SCL model, power grids of varying topologies were generated based on the IEEE 39-bus system model. Transient operational data was obtained by simulating N-1 and N--1 contingencies on these generated power system topologies. Test result demonstrated that the GEDF-SCL model can achieve high accuracy in transient stability prediction and adapt well to changing power grid topologies.
Paper Structure (12 sections, 7 equations, 5 figures, 6 tables)

This paper contains 12 sections, 7 equations, 5 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Physics-informed Graph Embedding Dynamic Features
  3. Contrastive Learning with Graph Embedding Dynamic Features
  4. Transient Stability Prediction and Alternating Grid Topology
  5. Topology Change Dataset
  6. Topology Data
  7. PMU Data
  8. N-1 Contingency Rotor Angle Stability
  9. Labeling Methodology
  10. Training Results
  11. Transfer Learning for N-$m$-1 Contingency
  12. Conclusion

Figures (5)

  • Figure 1: The framework of the transient stability GEDF-based supervised contrastive learning (GEDF-SCL).
  • Figure 2: IEEE 39-bus system.
  • Figure 3: Illustration of grid toplogy changes. Each new topology is generated by randomly removing four edges (shown as blue dashed lines in Fig. \ref{['fig2:subfig:b']}\ref{['fig2:subfig:c']}\ref{['fig2:subfig:d']}) and simultaneously adding four new edges (shown as red solid lines).
  • Figure 4: IEEE 39-bus N-1 simulations.
  • Figure 5: Simulation results and its corresponding GEDFs.