Loading...

Solving Nonlinear Multicommodity Network Flow Problem by Penalty Function Methods

Darroudi, Ali | 2010

728 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 40569 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Zokaee Aashtiani, Hedayat
  7. Abstract:
  8. Multicommodity network flow problem arises in the areas of transportation, logistics, and, telecommunication. The cost of each arc might be a linear or nonlinear function of the sum of commodity flows on that arc. Most of researches have worked on linear multicommodity flow problem. In addition, they consider the shared capacity constraints as the side constraints. In this thesis, we present a penalty-based algorithm for solving the multicommodity network flow problem with a linear or nonlinear objective function and considering the shared capacity constraints as the side. To solve multicommodity network flow problem, we relax the shared capacity constraints, and add a penalty term for the violation of the side constraints to the objective function. This algorithm uses three different penalty functions, Hyperbolic penalty function, Liu penalty function, and, Dynamic penalty function, to eliminate side constraints and produce a convex problem. A Frank-Wolf method is subsequently used to decompose the problem into a sequence of linear network flow problems. Computational results show that these approaches provide an efficient way for solving the multicommodity network flow problem
  9. Keywords:
  10. Multi-Commodity Network Flow ; Nonlinear Objective Function ; Hyperbolic Penalty Function ; Liu Penalty Function ; Dynamic Penalty Function

 Digital Object List

  • محتواي پايان نامه
  •   view

 Bookmark

No TOC