A lagrange relaxation method for the network design problem considering facility location

Ming Zhe Chen, Naoto Katayama, Hiroaki Mohri

Research output: Contribution to journalArticle

Abstract

The network design problem, which attempts to simultaneously take into account facility location and network topology, has a number of applications in transportation network design, distribution systems, telecommunication network planning, and other areas. In this paper, we investigate a new model that simultaneously optimizes facility locations for distribution centers and concentrators, network topology for line-haul trucks and fiber-optic communication lines, and routes of commodity flow for freight and communication data. Our model consists of uncapacitated facility location, uncapaci-tated arc design, general network design, an unlimited number of freight transshipment at facilities, and multi-commodity flow routing that has a large number of origin-destination pairs. It has been noted that this model is not suitable for solving problems that exist in the real world with current OR software on the market. We present a new formulation of this problem that minimizes the sum of flow costs, arc design costs, and facility location costs, and a new Lagrange relaxation problem with relaxed flow conservation constraints and some forcing constraints. We present an exact solution method for this Lagrange relaxation problem, and a subgradient method for Lagrange multipliers. In addition, we present an approximate solution method for obtaining good upper bound using solutions of a Lagrange relaxation problem, which is called Lagrange heuristics. Numerical examples up to 100 nodes are given to show the effectiveness of our formulation and solution methods for lower and upper bound.

Original languageEnglish
Pages (from-to)330-338
Number of pages9
JournalJournal of Japan Industrial Management Association
Volume56
Issue number5
Publication statusPublished - 2005

Fingerprint

Lagrange Method
Facility Location Problem
Relaxation Method
Network Design
Facility Location
Lagrange
Network Topology
Arc of a curve
Costs
Fiber Optics Communications
Topology
Subgradient Method
Network Planning
Distribution Center
Multicommodity Flow
Concentrator
Telecommunication Network
Transportation Networks
Formulation
Lagrange multipliers

Keywords

  • Facility location
  • Lagrange relaxation
  • Mathematical programming
  • Network design
  • Optimization

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering
  • Applied Mathematics
  • Management Science and Operations Research
  • Strategy and Management

Cite this

A lagrange relaxation method for the network design problem considering facility location. / Chen, Ming Zhe; Katayama, Naoto; Mohri, Hiroaki.

In: Journal of Japan Industrial Management Association, Vol. 56, No. 5, 2005, p. 330-338.

Research output: Contribution to journalArticle

@article{613abdb5b22a4327af43ad6e671b0860,
title = "A lagrange relaxation method for the network design problem considering facility location",
abstract = "The network design problem, which attempts to simultaneously take into account facility location and network topology, has a number of applications in transportation network design, distribution systems, telecommunication network planning, and other areas. In this paper, we investigate a new model that simultaneously optimizes facility locations for distribution centers and concentrators, network topology for line-haul trucks and fiber-optic communication lines, and routes of commodity flow for freight and communication data. Our model consists of uncapacitated facility location, uncapaci-tated arc design, general network design, an unlimited number of freight transshipment at facilities, and multi-commodity flow routing that has a large number of origin-destination pairs. It has been noted that this model is not suitable for solving problems that exist in the real world with current OR software on the market. We present a new formulation of this problem that minimizes the sum of flow costs, arc design costs, and facility location costs, and a new Lagrange relaxation problem with relaxed flow conservation constraints and some forcing constraints. We present an exact solution method for this Lagrange relaxation problem, and a subgradient method for Lagrange multipliers. In addition, we present an approximate solution method for obtaining good upper bound using solutions of a Lagrange relaxation problem, which is called Lagrange heuristics. Numerical examples up to 100 nodes are given to show the effectiveness of our formulation and solution methods for lower and upper bound.",
keywords = "Facility location, Lagrange relaxation, Mathematical programming, Network design, Optimization",
author = "Chen, {Ming Zhe} and Naoto Katayama and Hiroaki Mohri",
year = "2005",
language = "English",
volume = "56",
pages = "330--338",
journal = "Journal of Japan Industrial Management Association",
issn = "0386-4812",
publisher = "Nihon Keikei Kogakkai",
number = "5",

}

TY - JOUR

T1 - A lagrange relaxation method for the network design problem considering facility location

AU - Chen, Ming Zhe

AU - Katayama, Naoto

AU - Mohri, Hiroaki

PY - 2005

Y1 - 2005

N2 - The network design problem, which attempts to simultaneously take into account facility location and network topology, has a number of applications in transportation network design, distribution systems, telecommunication network planning, and other areas. In this paper, we investigate a new model that simultaneously optimizes facility locations for distribution centers and concentrators, network topology for line-haul trucks and fiber-optic communication lines, and routes of commodity flow for freight and communication data. Our model consists of uncapacitated facility location, uncapaci-tated arc design, general network design, an unlimited number of freight transshipment at facilities, and multi-commodity flow routing that has a large number of origin-destination pairs. It has been noted that this model is not suitable for solving problems that exist in the real world with current OR software on the market. We present a new formulation of this problem that minimizes the sum of flow costs, arc design costs, and facility location costs, and a new Lagrange relaxation problem with relaxed flow conservation constraints and some forcing constraints. We present an exact solution method for this Lagrange relaxation problem, and a subgradient method for Lagrange multipliers. In addition, we present an approximate solution method for obtaining good upper bound using solutions of a Lagrange relaxation problem, which is called Lagrange heuristics. Numerical examples up to 100 nodes are given to show the effectiveness of our formulation and solution methods for lower and upper bound.

AB - The network design problem, which attempts to simultaneously take into account facility location and network topology, has a number of applications in transportation network design, distribution systems, telecommunication network planning, and other areas. In this paper, we investigate a new model that simultaneously optimizes facility locations for distribution centers and concentrators, network topology for line-haul trucks and fiber-optic communication lines, and routes of commodity flow for freight and communication data. Our model consists of uncapacitated facility location, uncapaci-tated arc design, general network design, an unlimited number of freight transshipment at facilities, and multi-commodity flow routing that has a large number of origin-destination pairs. It has been noted that this model is not suitable for solving problems that exist in the real world with current OR software on the market. We present a new formulation of this problem that minimizes the sum of flow costs, arc design costs, and facility location costs, and a new Lagrange relaxation problem with relaxed flow conservation constraints and some forcing constraints. We present an exact solution method for this Lagrange relaxation problem, and a subgradient method for Lagrange multipliers. In addition, we present an approximate solution method for obtaining good upper bound using solutions of a Lagrange relaxation problem, which is called Lagrange heuristics. Numerical examples up to 100 nodes are given to show the effectiveness of our formulation and solution methods for lower and upper bound.

KW - Facility location

KW - Lagrange relaxation

KW - Mathematical programming

KW - Network design

KW - Optimization

UR - http://www.scopus.com/inward/record.url?scp=84864062636&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84864062636&partnerID=8YFLogxK

M3 - Article

VL - 56

SP - 330

EP - 338

JO - Journal of Japan Industrial Management Association

JF - Journal of Japan Industrial Management Association

SN - 0386-4812

IS - 5

ER -