A novel partitioning algorithm for power grid using community detection

This paper proposes a novel partitioning algorithm based on angles between vectors using community detection theory. Considering the physical properties of the power system, the weighted modularity matrix of power grid is constructed and modified to a non-negative definite matrix which can be an orthogonal matrix by orthogonal decomposition. Thus transform the community detection problem as a vector partitioning problem which can be solved without decomposing the modularity matrix or even computing the vertex vector explicitly either. A novel partitioning algorithm based on merging vectors with least angle is designed. The algorithm is applied to partition IEEE 39-bus system, and the results demonstrate the validity of our algorithm.