Dynamic event identification and wide -area response -based control in large systems
Abstract
The expanding use of synchronized Phasor Measurement Units (PMUs) in power networks has brought the necessity and many opportunities to develop wide-area monitoring and control systems. In this research, we have analyzed wide-area phasor measurements to investigate three related problems for power system operation and control. First, the Haar wavelet is employed to analyze event frequency signals, and the envelope of the output detection signal is approximated by the best Least Squares fit of a first-order exponential with a variable-width window. The estimated post-event damping is used to determine the end of the event, and a detection logic is used to separate overlapping events. Second, a hierarchical clustering method is investigated to cluster a power network into coherent groups based on wide-area frequency signals. The measurements from representative PMUs are utilized online to find the most likely group from which an event originates. Last, a decision tree (DT) based method is investigated to design a wide-area response-based (WARB) control scheme. The scheme involves triggering a combination of WARB controls when a DT trained for stability prediction indicates impending loss of synchronism, and it is able to prevent loss of synchronism across the intertie in many situations and to preserve stability for the entire network sometimes. The event identification and wide-area control schemes are simple, response-based and computationally efficient; they could be implemented in PMU-based controllers for online application.
Degree
Ph.D.
Advisors
Ong, Purdue University.
Subject Area
Electrical engineering|Artificial intelligence|Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.