Vehicle Routing Problem in Competitive Environment
Student: Mohammad Mahdi-Mohtadi
Advisor : Dr. Kazem-Noghondarian
Co-Advisors : Dr. Ahmad-Makuyi
Internal Committee Members: Siamak-Noori- S.Jafar- Sadjadi
External Committee Members: Alireza-Hajji- Fariborz-jolai
This thesis discusses vehicle routing problem in a competitive environment. In real world, distribution conditions for most products are not monopolistic; competitors simultaneously distribute products. A distributer who meets the customer earlier than others will meet all or part of his demand. Therefore each distributer's service time for each node exerts influence on sales volume and profit. This fact should also be taken into account that, to increase their sale and profit, competitors will react to any action by distributers. These conditions can be explained through game theory which is applied in this article as the basis of modeling for the problem.
Since it is necessary to understand complex conditions of the problem for accurate modeling, previous literature has been reviewed before modeling. Applying scenario planning, we have tried to describe competitive atmosphere of the problem and determine different competitive patterns through the extraction of key variables. One of these scenarios has been chosen as the basic pattern being discussed here.
As no solution was found to reach equilibrium in big games while we reviewed literature of game theory, an innovative algorithm has been developed for this purpose.
Two models have been offered for the problem. In first model each competitor has one vehicle with no capacity limit. In second model, which is closer to real conditions, each competitor has some vehicles with specific capacity. These models have been solved and tested for reliability using the proposed algorithms and supra-innovative algorithms in the form of a sample problem.