A graph theory based methodology for optimal PMUs placement and multiarea power system state estimation

This paper proposes a novel methodology based on the graph theoretical observability analysis for optimal placement of phasor measurement units (PMUs) in a power system. The proposed method identifies the optimal PMUs number, their strategic location and the minimum set of critical measurements aimed at ensuring the complete system observability. Besides, we demonstrated that the identified monitoring asset allows us to decompose the overall power system into a minimum number of decoupled non-overlapping observable islands. Thanks to this feature each island can assess its state by processing only local variables without the need for a central fusion center acquiring and processing all the PMUs measurements. Consequently the state of the entire power system can be estimated according to a distributed and decentralized computing paradigm. Numerical results obtained on several large case studies, including both IEEE test networks and national power systems, confirm the effectiveness of the proposed methodology.