Dynamic control in path-planning with real-time heuristic search

摘要: Real-time heuristic search methods, such as LRTA*, are used by situated agents in applications that require the amount of planning per action to be constant-bounded regardless problem size. LRTA* interleaves and execution, with a fixed depth being achieve progress towards goal. Here we generalize algorithm allow for dynamically changing (sub-)goal. Evaluation path-planning on videogame maps shows new significantly outperforms fixed-depth, fixed-goal LRTA*. The can same quality solutions but nine times less computation, or use produce four better solutions. These extensions make real-time practical choice computer video-games.