A combined mathematical model and genetic algorithm solution methodology for tri-levels electrical grid protection problem with multiple types of attacks. (November 2022)

• Record Type: Journal Article
• Authors: Başkan, Orkun; Sağır, Müjgan
• Is Part Of: Computers & industrial engineering. Volume 173(2022)
• Publication Date: 2022-11
• Subjects: Interdiction problems -- Electrical grid problem -- Tri-level programing -- Multiple attacks -- Genetic algorithm; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108718
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: Different attack types and defense types are considered for the first time in interdiction problems. The Purpose Function is designed to perform load shedding from points with lower criticality level. Two new heuristics (random and rule-based) based on Genetic Algorithm have been developed. Abstract: As an important component of energy industry, electric systems have a key role for the continuity of many critical industries and any problem encountered in these systems, such as random failures, deliberate attacks like terrorist acts have a potential to have an impact on many other sectors. We study an interdiction problem in this context in which two sides with opposite interests; a defender protects a subset of some sides and the attacker targets a subset of some unprotected sides. These sides could be some electrical components such as power plants, transmission lines etc. The problem, therefore determining which of the Plant, Substation, and Transmission Line components of an electrical grid will be protected with the limited means in such manner that it uninterruptedly supplies power to a certain zone. The study addresses the interdiction focused on an "attack" that is carried out via terrorists acts and similar interventions. The study's contribution is the consideration of different attack types that were overlooked in previous studies and of the criticality of load shedding zone. A three-level mathematical model is developed for the given problem. Thereafter, we proposed a genetic algorithm-based heuristics where the initial population is generated randomly (r_GA) and on the basis of a rule (k_GA). The performances of r_GA and k_GA were compared depending on the critical systems' protection levels as strict, medium and, loose.