Abstract

The rapid expansion and integration of Distributed Generations (DG) into power systems plays an increasingly important role in their planning, operation, and control. The rules used to design and operate current systems are being altered by the DGs incorporation. This may jeopardize the system's reliability and security. Private owners of large DGs should not be restricted to a particular time schedule to connect/disconnect their generation to/from the system. This feature dynamically changes the typical power system with unidirectional power flow from generation to the loads. A smart Central Protection Unit (CPU) is needed to take proper measures in case of DGs arbitrarily disconnection, isolation or any other type of fault. On the other hand, recent major blackouts resulting from pushing the power systems to the edge has revealed the need for a smarter supervisory system for enhanced reliability and stability. Hence, there is a high demand for a robust and smart supervisory system which can diagnose power systems disturbances in real-time and prevent aggravation and expansion.This thesis is focused on studying the impacts of DG integration on the power systems. Phasor Measurement Units (PMUs) play an important role on the monitoring of power systems. Multiple major data analysis techniques including K-means, Smart K-means clustering, and DBSCAN clustering of the PMU output data have been implemented. Higher order moments of Kurtosis and Skewness indices were also employed in order to estimate the system state.

Date of publication

Fall 1-15-2014

Document Type

Thesis

Language

english

Persistent identifier

http://hdl.handle.net/10950/191

COinS