Compressive power systems: applications of compressive sensing and sparse recovery in the analysis of smart power grids

Abstract

During the last two decades, an intelligence revolution intensively changed the technology of electrical power networks, forming a new generation of power systems called smart grids (SGs). In general, “the term SG refers to electricity networks that can intelligently integrate the behavior and actions of all parameters and users connected to them”. This revolution transformed the traditional power grid from a single-layer physical system into a huge cyber-physical network, using a layer of information that flows through the system. This information includes the status of several parameters in the network such as bus voltages, branch currents, and load consumption. In addition to the actuating or controlling commands fed back to the network from controllers and decision making units. Nowadays, the Supervisory Control and Data Acquisition (SCADA) system in addition to the Wide Area Mentoring System (WAMS) provide electrical data for each local system in near real time. The most popular sensing technology used widely in SG data collection systems is the high sampling rate synchronous Phasor Measurement Unit (PMU). Nevertheless, collecting, storing, transferring, and analyzing the huge amount of data flowing through the information layer of the SG, together with the uncertainty caused by renewable-based distributed generators and unpredictable load characteristics, challenge the standard methods for security, monitoring, and control. In this thesis, we aim to exploit the inherent sparse nature of both structure and data in SGs to introduce new, fast and reliable techniques to address the challenges related to real time data analysis, monitoring and security in smart power systems. Our work is primarily inspired by a new paradigm in the field of signal processing widely known as the theory of Compressive Sensing and Sparse Recovery (CS-SR). Generally, CS-SR implies that a sufficiently sparse phenomenon can be recovered from a small set of randomly collected measurements. In our early chapters, combining the sparse sampling theory from the field of CS with concepts borrowed from graph theory, we introduce a set of sparse recovery-based mathematical formulations to address famous global monitoring challenges such as power line outage localization, network topology identification, network dynamic behavior modeling and tracking. Lastly, we develop a modified sparse representation-based classification approach to deal with a challenging local monitoring problem widely known as power quality events recognition.