Authors:
Raulcezar M. F. Alves
1
and
Carlos R. Lopes
2
Affiliations:
1
Federal University of Uberlandia, Brazil
;
2
Federal University of Uberlândia, Brazil
Keyword(s):
Planning Systems, EHC, BFS, LRTA*, FF Planner.
Related
Ontology
Subjects/Areas/Topics:
Agents
;
Artificial Intelligence
;
Artificial Intelligence and Decision Support Systems
;
Enterprise Information Systems
;
Intelligent Agents
;
Internet Technology
;
Problem Solving
;
Scheduling and Planning
;
Web Information Systems and Technologies
Abstract:
A number of new heuristic search methods have been employed in the development of planning systems over the last years. Enforced Hill Climbing (EHC) combined with a complete search strategy, such as Best First Search (BFS), is a method that has been frequently used in several AI planning systems. Although this method presents an enhanced performance when compared to alternative methods used in many of the other planning domains, it does all the same show some weaknesses. In this paper the authors propose to replace the use of EHC and BFS with LRTA*, which is a search algorithm guided by heuristics like EHC and is as complete as BFS. Moreover, the authors implemented some optimizations on LRTA*, such as a heap with a maximum capacity to store the states during the search, along with pruning of successors after state expansion. Experiments show significant improvements compared to the standard form of the FF planner, which is a representative planning system based on EHC and BFS.