Monitoring and Optimization for Power Grids: A Signal Processing Perspective

Abstract

I. INTRODUCTIONAlbeit the North American power grid has been recognized as the most important engineering achievement of the 20th century, the modern power grid faces major challenges [87]. Increasingly complex interconnections even at the continent size render prevention of the rare yet catastrophic cascade failures a strenuous concern. Environmental incentives require carefully revisiting how electrical power is generated, transmitted, and consumed, with particular emphasis on the integration of renewable energy resources. Pervasive use of digital technology in grid operation demands resiliency against physical and cyber attacks on the power infrastructure. Enhancing grid efficiency without compromising stability and quality in the face of deregulation is imperative. Soliciting consumer participation and exploring new business opportunities facilitated by the intelligent grid infrastructure hold a great economic potential.The smart grid vision aspires to address such challenges by capitalizing on state-of-the-art information technologies in sensing, control, communication, and machine learning [2], [24]. The resultant grid is envisioned to have an unprecedented level of situational awareness and controllability over its services and infrastructure to provide fast and accurate diagnosis/prognosis, operation resiliency upon contingencies and malicious attacks, as well as seamless integration of distributed energy resources. A. Basic Elements of the Smart GridA cornerstone of the smart grid is the advanced monitorability on its assets and operations. Increasingly pervasive installation of the phasor measurement units (PMUs) allows the so-termed synchrophasor measurements to be taken roughly 100 times faster than the legacy supervisory control and data acquisition (SCADA) measurements, time-stamped using the global positioning system (GPS) signals to capture the grid dynamics. In addition, the availability of low-latency two-way communication networks will pave the way to high-precision real-time grid state estimation and detection, remedial actions

The challenging paradigm of interrelated energy systems towards a more sustainable future

“…This triad is also valuable for organising research efforts shedding light into these challenges. Technological adaptation-oriented research has contributed with knowledge on the integration of electricity storage in distribution systems [37][38][39][40][41], integration of distributed generation sources from wind [42][43][44], solar [45][46][47][48][49], CHP [50][51][52], and micro-CHP [53][54][55][56], integration of EVs [57][58][59][60], integration of smart meters [61][62][63][64][65][66], implementation of DR [67][68][69][70][71], deployment of active distribution management systems [72][73][74][75][76], and advanced grid monitoring and control [77][78][79], as well as the use of artificial intelligence methods [80][81][82], and machine learning applicati...…”

Randomized Block Frank–Wolfe for Convergent Large-Scale Learning

“…(22). However, since finding upper bounds on g(x t ) is difficult [8], [15], [14], bounds on the minimal expected duality gap until iteration t, defined as [8], [14] …”

“…The objective function of (6) is nonconvex in certain applications, such as constrained multilinear decomposition [21] and power system state estimation [22], [23]. Yet, convergence of RB-FW has never been investigated for this case.…”

Advanced optimization methods for power systems

“…1) Semidefinite programming primer: ACOPF can be formulated as a quadratic constrained quadratic program (QCQP); see (22) below and the paper [8] in the invited session of PSCC2014 for more details. Convex relaxation of quadratic programs has been applied to many engineering problems [9] (see also [10] for some applications in power systems). There is a rich theory, mature algorithms, and extensive empirical experiences; see, e.g., [11], [12], [9].…”