Solving a selective dial-a-ride problem with logic-based Benders decomposition

•We propose logic-based Benders algorithms for a selective dial-a-ride problem.•Subproblems are solved using a combination of CP and MILP algorithms.•Four different strategies for refining Benders cuts are tested.•We speed up the master-problem using heuristic boosting techniques. Today’s society is...

Full description

Saved in:
Bibliographic Details
Published in:Computers & operations research Vol. 96; pp. 30 - 54
Main Authors: Riedler, Martin, Raidl, Günther
Format: Journal Article
Language:English
Published: New York Elsevier Ltd 01.08.2018
Pergamon Press Inc
Subjects:
Algorithms
Benders decomposition
Branch-and-check
Combinatorial Benders cuts
Decomposition
Dial-a-Ride problem
Integer programming
Linear programming
Logic-based Benders decomposition
Mixed integer
Operations research
Public transportation
Studies
Transportation
Travel time
Windows (intervals)
Branch-and-check
Dial-a-Ride problem
Combinatorial Benders cuts
Logic-based Benders decomposition
Transportation
ISSN:0305-0548, 1873-765X, 0305-0548
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•We propose logic-based Benders algorithms for a selective dial-a-ride problem.•Subproblems are solved using a combination of CP and MILP algorithms.•Four different strategies for refining Benders cuts are tested.•We speed up the master-problem using heuristic boosting techniques. Today’s society is facing an ever-growing demand for mobility. To a high degree these needs can be fulfilled by individual and public transport. People that do not have access to the former and cannot use the latter require additional means of transportation. This is where dial-a-ride services come into play. The dial-a-ride problem considers transportation requests of people from pick-up to drop-off locations. Users specify time windows with respect to these points. Requests are served by a given vehicle fleet with limited capacity and tour duration per vehicle. Moreover, user inconvenience considerations are taken into account by limiting the travel time between origin and destination for each request. Previous research on the dial-a-ride problem primarily focused on serving a given set of requests with a fixed-size vehicle fleet at minimal traveling costs. It is assumed that the request set is sufficiently small to be served by the available vehicles. We consider a different scenario in which a maximal number of requests shall be served under the given constraints, i.e., it is no longer guaranteed that all requests can be accepted. For this new problem variant we propose a compact mixed integer linear programming model as well as algorithms based on Benders decomposition. In particular, we employ logic-based Benders decomposition and branch-and-check using mixed integer linear programming and constraint programming algorithms. We consider different variants on how to generate Benders cuts as well as heuristic boosting techniques and different types of valid inequalities. Computational experiments illustrate the effectiveness of the suggested algorithms.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0305-0548
1873-765X
0305-0548
DOI:10.1016/j.cor.2018.03.008