BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260523T215402EDT-9637uMa07h@132.216.98.100
DTSTAMP:20260524T015402Z
DESCRIPTION:Abstract\n\nThe ongoing deepening penetration of renewable ener
 gy sources is posing significant challenges to electric power system opera
 tion and planning. High levels of renewable energy\, particularly from win
 d and solar sources\, introduce significant variability and uncertainties 
 that system operators must integrate into operational planning problems to
  ensure secure and cost-effective operation. Solving NP-hard (Non Determin
 istic Polynomial-time) operational planning problems such as the unit comm
 itment\, security-constrained unit commitment\, and ac optimal power flow 
 (AC-OPF) repeatedly is essential for ensuring reliable and economical dail
 y operations. However\, their numerous constraints and the inclusion of un
 certainties and variability can substantially prolong solution times and m
 ay render these problems intractable for large systems. Nevertheless\, exi
 sting empirical evidence and prior research indicate that these problems o
 ften include numerous unnecessary constraints.\n\nThis thesis seeks to adv
 ance the state-of-the-art of optimization-based constraint screening appro
 aches for power system operational planning problems. We leverage constrai
 nt learning to achieve efficient constraint screening outcomes. Constraint
  learning embeds trained machine learning models directly into constraint 
 screening approaches. Constraint learning is primarily led by discovering 
 insights from previously solved operations planning instances which inheri
 t the economical aspect of their objective functions as well as the observ
 ed demand patterns.\n\nIn optimization-based constraint screening for oper
 ational planning problems\, robust optimization is often employed to ensur
 e the operator’s ability to handle a broad range of possible scenarios\, w
 here scenarios refer to probable realizations of the system’s sources of u
 ncertainty. In this realm\, we propose polyhedral uncertainty sets that ar
 e capable of capturing spatial correlations in the uncertainty space of va
 riable renewable energy and demand\, called net load. Polyhedral uncertain
 ty sets offer coverage levels similar to those of convex hulls without the
  over conservatism of multidimensional bounding boxes. Afterward\, we exte
 nd the optimization-based approach known as umbrella constraint discovery 
 (UCD)\, in the context of polyhedral uncertainty sets integrated in unit c
 ommitment problems. The classical UCD approach identifies non-redundant co
 nstraints by enforcing a consistency logic on the set of constraints. Furt
 hermore\, we augment UCD with an upper bound cost-driven constraint derive
 d by fitting an appropriate regression model using past solved instances o
 f the unit commitment problem. This new formulation called techno-economic
  UCD screens out redundant and inactive constraints that are not necessary
  to achieve optimal solutions for unit commitment with significant computa
 tional enhancement. This is a key departure from UCD\, which is only capab
 le of screening out redundant constraints.\n\nFurthermore\, we extend the 
 optimization-based bound tightening technique for constraint screening pro
 blem in the context of the AC-OPF problem using constraint learning. Due t
 o the non-convexity of the AC-OPF problem\, we investigate how different c
 onvex relaxations of the AC-OPF perform when performing line constraint sc
 reening. Next\, we propose an interpretable machine learning algorithm for
  real-time constraint generation for the security-constrained unit commitm
 ent problem. Our proposed approach is simplifying and accelerates the conv
 entional constraint generation approach by leveraging machine learning app
 roaches to learn the active set of pre- and post-contingency constraints. 
 Those are then used to warm-start the constraint generation process used i
 n the practical solution of security-constrained unit commitment problems 
 to reveal the constraints set necessary and sufficient to guarantee the fe
 asibility and optimality of the solution at a fraction of the computationa
 l cost needed by state-of-the-art approaches.\n\nFinally\, we develop a no
 vel approach to determine the distance of an optimization problem solution
  to its non-redundant constraints defining its feasible region\, or even i
 ts violated constraints in cases when problems are infeasible. Here\, the 
 notion of “distance” to a problem’s constraints is associated with the abi
 lity of the power system to respond to uncertain events\, i.e.\, how flexi
 ble it is. For this purpose\, we propose novel system flexibility metrics 
 which are calculated by solving an associated inverse optimization\n	proble
 m. We reveal that when applying this approach to the loadability set of a 
 power system\, it can accurately determine the feasibility of uncertain ne
 t load vector\, and it is able to identify which constraints are closest t
 o that uncertain net load vector.\n
DTSTART:20231012T143000Z
DTEND:20231012T163000Z
LOCATION:Room 603\, McConnell Engineering Building\, CA\, QC\, Montreal\, H
 3A 0E9\, 3480 rue University
SUMMARY:PhD defence of Mohamed Awadalla – Data-Driven Constraint Learning a
 nd Screening for Operations of Sustainable Power Systems
URL:/ece/channels/event/phd-defence-mohamed-awadalla-d
 ata-driven-constraint-learning-and-screening-operations-sustainable-351763
END:VEVENT
END:VCALENDAR