Authors:
Selim Bora
1
;
Endre Boros
2
;
Lei Lei
2
;
W. Art Chaovalitwongse
3
;
Gino J. Lim
4
and
Hamid R. Parsaei
1
Affiliations:
1
Texas A&M University at Qatar, Qatar
;
2
Rutgers University, United States
;
3
University of Washington, United States
;
4
University of Houston, United States
Keyword(s):
Vessel Scheduling, Bender’s Decomposition, Dynamic Programming.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Dynamic Programming
;
Industrial Engineering
;
Knowledge Discovery and Information Retrieval
;
Knowledge-Based Systems
;
Linear Programming
;
Mathematical Programming
;
Methodologies and Technologies
;
Operational Research
;
Optimization
;
Symbolic Systems
Abstract:
We study a difficult real life scheduling problem encountered in oil and petrochemical industry, involving inventory and distribution operations, which requires integrated scheduling. The problem itself is NP-complete, however we show some special cases, and propose polynomial time solution methods. These could be used as a starting point for a heuristic making use of these simplified cases. This study proposes two alternative approaches for the main problem, one of them making use of one of the special cases using minimum cost flow formulation, and the other one using Benders Decomposition once the problem is reformulated to make it easier to handle. Both results show promising results and computation time. Benders Decomposition approach allows exact solutions to be found in a much faster fashion.