Abstract/Details

Multi-objective optimization using evolutionary algorithms for geographical zone partitioning


2011 2011

Other formats: Order a copy

Abstract (summary)

Transportation planning initiatives at different administrative levels require traffic data at appropriate resolutions. This research presents a methodology for determination of optimal traffic resolution depending on the planning level under consideration. The framework is supported by a proof of concept in the form of a Geographical Zone Partitioning Software (GEZOPS) - a decision-support tool serving to fill an existing gap in the domain of high level traffic analysis tools for transportation planning. The strategy is based on multi-objective optimization using weighted agglomerative ranking & selection and evolutionary algorithms. Facility for handling geospatial constraints for contiguity and primary arterial boundaries is provided with full-scale integration with mapping and software for statistical analysis.

In contributing to the area of evolutionary algorithms, the research demonstrates the application of Restricted Growth Functions (RGF) to encode the set partition. This encoding is used with modern evolutionary algorithms such as the Non-dominated Sorting Genetic Algorithm-II (NSGA-II). Problem-specific issues such as geospatial constraints and representation provide the necessary motivation for design and development of diversity preservation strategies, crossover and mutation approaches to provide the necessary balance between exploration and exploitation. Specific algorithmic aspects and parameterization specifications are implemented to provide a platform for sensitivity analysis and comparative studies.

The Huntsville Metropolitan Planning Organization (HMPO) is used as a case study for testing the algorithms developed in this research initiative. Results from subject-matter expert assessment and validation are provided to demonstrate that the strategy produces partitions that are not only quantitatively fit but also useful from a transportation planning perspective. Scope for further research in the zone partitioning problem using other formulations is identified.

Indexing (details)


Subject
Industrial engineering;
Transportation planning;
Operations research
Classification
0546: Industrial engineering
0709: Transportation planning
0796: Operations research
Identifier / keyword
Social sciences; Applied sciences; Cluster analysis; Data mining; Evolutionary algorithms; Knowledge discovery; Multi-objective optimization; Zone partitioning
Title
Multi-objective optimization using evolutionary algorithms for geographical zone partitioning
Author
Sharma, Nitin S.
Number of pages
209
Publication year
2011
Degree date
2011
School code
0278
Source
DAI-B 73/05, Dissertation Abstracts International
Place of publication
Ann Arbor
Country of publication
United States
ISBN
9781267124241
Advisor
Swain, James J.
University/institution
The University of Alabama in Huntsville
University location
United States -- Alabama
Degree
Ph.D.
Source type
Dissertations & Theses
Language
English
Document type
Dissertation/Thesis
Dissertation/thesis number
3492783
ProQuest document ID
917934521
Copyright
Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works.
Document URL
http://search.proquest.com/docview/917934521
Access the complete full text

You can get the full text of this document if it is part of your institution's ProQuest subscription.

Try one of the following:

  • Connect to ProQuest through your library network and search for the document from there.
  • Request the document from your library.
  • Go to the ProQuest login page and enter a ProQuest or My Research username / password.