Finding Most Vital Links over Time in a Flow Network

This paper deals with finding most vital links of a network which carries flows over time (also called ”dynamic flows”). Given a network and a time horizon T, Single Most Vital Link Over Time (SMVLOT) problem looks for a link whose removal results in greatest decrease in the value of maximum flow ov...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:An international journal of optimization and control Ročník 2; číslo 2; s. 173 - 186
Hlavní autori: MOROWATI-SHALILVAND, Shahram, MEHRI-TEKMEH, Javad
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Balikesir Balikesir University, Faculty of Engineering Department of Industrial Engineering 01.01.2012
Predmet:
Algorithms
Benders decomposition
Combinatorial analysis
Integer programming
Linear programming
Links
Mathematical programming
ISSN:2146-0957, 2146-5703
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:This paper deals with finding most vital links of a network which carries flows over time (also called ”dynamic flows”). Given a network and a time horizon T, Single Most Vital Link Over Time (SMVLOT) problem looks for a link whose removal results in greatest decrease in the value of maximum flow over time (dynamic maximum flow) up to time horizon T between two terminal nodes. SMVLOT problem is formulated as a mixed binary linear programming problem. This formulation is extended to a general case called k-Most Vital Links Over Time (KMVLOT) problem, in which we look for finding those k links whose removal makes greatest decrease in the value of maximum flow over time. A Benders decomposition algorithm is proposed for solving SMVLOT and KMVLOT problems. For the case of SMVLOT problem, the proposed algorithm is improved to a fully combinatorial algorithm by adopting an iterative method for solving existing integer programming problem. However, our experimental results showed the superiority of proposed methods.
Bibliografia:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ISSN:2146-0957
2146-5703
DOI:10.11121/ijocta.01.2012.0098