Fast Grid Splitting Detection for N-1 Contingency Analysis by Graph Computing

TL;DR

This paper introduces a graph computing algorithm for fast grid splitting detection in N-1 contingency analysis, significantly improving speed and efficiency in power system reliability assessments.

Contribution

It presents a novel graph-based method for contingency analysis that outperforms traditional serial algorithms in speed, demonstrated through real-world power system tests.

Findings

Achieved a 6-fold speedup over existing algorithms

Successfully applied to a large-scale power system with 2752 buses

Provides a foundation for real-time contingency analysis

Abstract

In this study, a graph-computing based grid splitting detection algorithm is proposed for contingency analysis in a graph-based EMS (Energy Management System). The graph model of a power system is established by storing its bus-branch information into the corresponding vertex objects and edge objects of the graph database. Numerical comparison to an up-to-date serial computing algorithm is also investigated. Online tests on a real power system of China State Grid with 2752 buses and 3290 branches show that a 6 times speedup can be achieved, which lays a good foundation for advanced contingency analysis.