elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Weighted A* Search - Unifying View and Application

Ebendt, Rüdiger and Drechsler, Rolf (2009) Weighted A* Search - Unifying View and Application. Artificial Intelligence, 173 (14), pp. 1310-1342. Elsevier. DOI: 10.1016/j.artint.2009.06.004 ISSN 0004-3702

[img] PDF - Registered users only
748kB

Official URL: http://dx.doi.org/10.1016/j.artint.2009.06.004

Abstract

The A* algorithm is a well-known heuristic best-first search method. Several performance-accelerated extensions of the exact A* approach are known. Interesting examples are approximate algorithms where the heuristic function used is inflated by a weight (often referred to as weighted A*). These methods guarantee a bounded suboptimality. As a technical contribution, this paper presents the previous results related to weighted A* from authors like Pohl, Pearl, Kim, Likhachev and others in a more condensed and unifying form. By use of this unified view, a novel general bound on suboptimality of the result is derived: in the case of avoiding any reopening of expanded states, for �µ>0, this bound is (1+�µ)^â��N/2â�� where N is (an upper bound on) the maximum depth of the search. Binary Decision Diagrams (BDDs) are well-known to AI, e.g. from set-based exploration of sparse-memory and symbolic manipulation of state spaces. The problem of exact or approximate BDD minimization is introduced as a possible new challenge for heuristic search. Like many classical AI domains, this problem is motivated by real-world applications. Several variants of weighted A* search are applied to problems of BDD minimization and the more classical domains like blockworlds and sliding-tile puzzles. For BDD minimization, the comparison of the evaluated methods also includes previous heuristic and simulation-based methods such as Rudell's hill-climbing based sifting algorithm, Simulated Annealing and Evolutionary Algorithms. A discussion of the results obtained in the different problem domains gives our experiences with weighted A*, which is of value for the AI practitioner.

Item URL in elib:https://elib.dlr.de/59485/
Document Type:Article
Additional Information:Eine Kopie der Arbeit im PDF-Format kann per Email an Ruediger.Ebendt@dlr.de angefordert werden. A copy of the paper (PDF) can be requested by email to Ruediger.Ebendt@dlr.de.
Title:Weighted A* Search - Unifying View and Application
Authors:
AuthorsInstitution or Email of AuthorsAuthors ORCID iD
Ebendt, RüdigerUNSPECIFIEDUNSPECIFIED
Drechsler, RolfUniversität BremenUNSPECIFIED
Date:August 2009
Journal or Publication Title:Artificial Intelligence
Refereed publication:Yes
Open Access:No
Gold Open Access:No
In SCOPUS:Yes
In ISI Web of Science:Yes
Volume:173
DOI :10.1016/j.artint.2009.06.004
Page Range:pp. 1310-1342
Publisher:Elsevier
ISSN:0004-3702
Status:Published
Keywords:Shortest-Path, Dijkstra, A*, heuristic search
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Transport
HGF - Program Themes:Transport System
DLR - Research area:Transport
DLR - Program:V VS - Verkehrssystem
DLR - Research theme (Project):V - Mittel- und langfristige Entwicklung der Personenverkehrsnachfrage (old)
Location: Berlin-Adlershof
Institutes and Institutions:Institute of Transportation Systems > Traffic Management
Deposited By: Ebendt, Dr.rer.nat. Rüdiger
Deposited On:17 Aug 2009 14:31
Last Modified:06 Sep 2019 15:17

Repository Staff Only: item control page

Browse
Search
Help & Contact
Information
electronic library is running on EPrints 3.3.12
Copyright © 2008-2017 German Aerospace Center (DLR). All rights reserved.