Hierarchical Clustering based optimal PMU placement for power system fault observability

Optimal number and location of phasor measurement units (PMUs) in the power system networks faces challenges for achieving the full network observability during fault conditions. Achieving fault observability approach requires more constraints than normal system observability and consequently suffer...

Full description

Bibliographic Details
Main Authors: Moustafa Eissa, Amr Kassem
Format: Article
Language:English
Published: Elsevier 2018-08-01
Series:Heliyon
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2405844018333292
Description
Summary:Optimal number and location of phasor measurement units (PMUs) in the power system networks faces challenges for achieving the full network observability during fault conditions. Achieving fault observability approach requires more constraints than normal system observability and consequently suffers from complex analysis and heavy computational burden for the large-scale networks. A new algorithm for determining the optimal PMU placement considering the network fault observability is introduced. The proposed algorithm is achieved through four stages. The first stage is achieved through the network fault simulation to obtain the post fault change in voltage (ΔV) at each bus. Then, the post fault change in voltage (ΔV) is used to build the network connectivity matrix (CM) and forming a new developed Faulted Connectivity Matrix (FCM) that describes the power system topology during the fault conditions. The correlation between the buses is obtained, in the second stage, by applying Pearson correlation coefficient. Hierarchical Clustering technique is given, in the third stage, to cluster the network into coherent zones to find the most correlated buses. Finally, the optimal location of the PMUs is identified within each zone based on simple proposed placement rules. The proposed algorithm is tested under a variety of fault events applied on different standard test systems. The results show the simplicity and the effectiveness of the proposed algorithm.
ISSN:2405-8440