Planning with Reduced Operator Sets

Classical propositional STRIPSplanning is nothing but the searchfor a path in the state-transition graph induced by the operators in the planning problem. Whatmakes the problem hard is the size and the sometimesadverse structure of this graph. Weconjecture that the search for a plan wouldbe moreefficient if there wereonly a small numberof paths from the initial state to the goal state. Toverify this conjecture, wedefine the notion of reducedoperator sets and describe waysof finding such reduced sets. We demonstrate that somestate-of-the-art planners run faster using reducedoperator sets.

• Type: Conference
• Year: 2000
• Where: AIPS
• Authors: Patrik Haslum, Peter Jonsson
• Keywords: AIPS 2000, Artificial Intelligence, Classical Propositional Stripsplanning, Reducedoperator Sets, State-transition Graph