Graph Abstraction in Real-time Heuristic Search

V. Bulitko, N. Sturtevant, J. Lu, T. Yau
2007 The Journal of Artificial Intelligence Research  
Real-time heuristic search methods are used by situated agents in applications that require the amount of planning per move to be independent of the problem size. Such agents plan only a few actions at a time in a local search space and avoid getting trapped in local minima by improving their heuristic function over time. We extend a wide class of real-time search algorithms with automatically-built state abstraction and prove completeness and convergence of the resulting family of algorithms.
more » ... e then analyze the impact of abstraction in an extensive empirical study in real-time pathfinding. Abstraction is found to improve efficiency by providing better trading offs between planning time, learning speed and other negatively correlated performance measures.
doi:10.1613/jair.2293 fatcat:yyh2eiw325avpoclrm54kt2ugq