Addis Ababa University Libraries Electronic Thesis and Dissertations: AAU-ETD! >
College of Natural and Computational Sciences >
Thesis - Earth Sciences >

Please use this identifier to cite or link to this item: http://hdl.handle.net/123456789/3028

Authors: Gebreslasie, Gebremedhin
Advisors: Dr. K. V. Suryabhagavan
Keywords: GIS and Remote Sensing, Expert systems, Knowledge base, Decision tree, Weighted Overlay analysis, Off-road trafficability, Military operations
CExpert systems, Knowledge base, Decision tree, Weighted Overlay analysis, Off-road trafficability,
Decision making tools
Copyright: Jun-2009
Date Added: 11-May-2012
Publisher: AAU
Abstract: The successes of military operations depend on how decision makers and planners have evaluated the battle field prior to the deployment of armed forces on the ground. The study of off-road trafficability of the terrain is one of the key military operations that should be provided for military commanders at all levels in a real time scenario. This can rarely be provided using paper maps. This thesis comparatively evaluates the state-of-the-art of spatial modeling techniques for off-road trafficability analysis of wheeled military vehicles using GIS and remote sensing techniques. The Expert systems and Weighted Overlay Analysis were comparatively studied for modeling off-road trafficability. Similar data layers of Lu/Lc, Soil, Slope, Rivers and Manmade obstacles were used to generate the off-road trafficability maps from the two methods. The goal of comparison of these decision making tools was to test whether data in an ordinal scale from the Weighted Overlay Analysis produce comparable result with the Expert system that use hierarchy of decision tree. The results showed that there was a strong spatial correspondence between the outputs from the two methods with a spatial correlation of 0.78. Besides, a zonal cross tabulation between results showed that the two methods strongly accord to each other in the SLOW-GO and NO-GO trafficability classes with 86% and 75% summarized in the same zone respectively. However, results of comparison also showed that there was a significant disagreement between the two methods in the GO and VERY SLOW-GO classes with only 53% and 31% summarized in the same zone respectively. The disagreement between the two methods was mainly due to reclassification of data, and the factor weights (wj) used in the WOA method which can’t be employed in the Expert system and the detail knowledge of experts used in the Expert system to yield expert level performance that can’t be entertained in the WOA.
URI: http://hdl.handle.net/123456789/3028
Appears in:Thesis - Earth Sciences

Files in This Item:

File Description SizeFormat
949631854673203668232007159915488736602 MBAdobe PDFView/Open

Items in the AAUL Digital Library are protected by copyright, with all rights reserved, unless otherwise indicated.


  Last updated: May 2010. Copyright © Addis Ababa University Libraries - Feedback