Napredna pretraga

Determination of optimal capacitor bank allocation by an adapted evolutionary algorithm with use of constraint fuzzification

Barukčić, Marinko; Hederić, Željko; Hadžiselimović, Miralem
Determination of optimal capacitor bank allocation by an adapted evolutionary algorithm with use of constraint fuzzification // Journal of Energy Technology, 4 (2011), 1; 23-38 (međunarodna recenzija, članak, znanstveni)

Naslov
Determination of optimal capacitor bank allocation by an adapted evolutionary algorithm with use of constraint fuzzification

Autori
Barukčić, Marinko ; Hederić, Željko ; Hadžiselimović, Miralem

Izvornik
Journal of Energy Technology (1855-5748) 4 (2011), 1; 23-38

Ključne riječi
Capacitor bank allocation ; distributive network ; evolutionary algorithm ; fuzzy logic ; optimizatio

Sažetak
This paper presents the determination of optimal capacitor placement in radial distribution networks, during which a mathematical record of the optimization problem in the form of the objective function is carried out. To solve the previously set optimization problem, a customized evolutionary algorithm has been used. Adjustment is done by modification of the way of individual coding and presentation of the new reproduction operator. During the performance of evolutionary algorithms, individuals that exceed the set limits of the objective function are rejected or penalized. Problems can occur when the algorithm rejects quality individuals from the population due to the slightest excess of limitation. In succeeding generations, these individuals may provide good offspring ; it is therefore advisable to retain them in the population. To ensure the survival of individuals who exceed the limits of the objective function by small amounts, a fuzzification of restrictions is used. This paper presents the fuzzification of voltage limits and the determination of the voltage constraint coefficient, using fuzzy set theory and fuzzy logic. For individuals that exceed the voltage limits obtained, the objective function value is multiplied by the voltage constraint coefficient. The presented algorithm is implemented on a test network with 34 buses.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika

Projekt / tema
008-0362978-1485 - Dijagnostika električnih strojeva i strojevi s povećanim stupnjem djelovanja (Ante Elez, )

Ustanove
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek

Uključenost u ostale bibliografske baze podataka:

• INSPEC
• Cambridge Scientific Abstracts: Abstracts in New Technologies and Engineering (CSA ANTE)
• ProQuest's Technology Research Databas