Adaptive path planning [electronic resource] : Algorithm and analysis
- Washington, D.C. : United States. Dept. of Energy, 1993.
Oak Ridge, Tenn. : Distributed by the Office of Scientific and Technical Information, U.S. Dept. of Energy.
- Physical Description:
- 10 pages : digital, PDF file
- Additional Creators:
- Sandia National Laboratories, United States. Department of Energy, and United States. Department of Energy. Office of Scientific and Technical Information
- Restrictions on Access:
- Free-to-read Unrestricted online access
- Path planning has to be fast to support real-time robot programming. Unfortunately, current planning techniques are still too slow to be effective, as they often require several minutes, if not hours of computation. To alleviate this problem, we present a learning algorithm that uses past experience to enhance future performance. The algorithm relies on an existing path planner to provide solutions to difficult tasks. From these solutions, an evolving sparse network of useful subgoals is learned to support faster planning. The algorithm is suitable for both stationary and incrementally-changing environments. To analyze our algorithm, we use a previously developed stochastic model that quantifies experience utility. Using this model, we characterize the situations in which the adaptive planner is useful, and provide quantitative bounds to predict its behavior. The results are demonstrated with problems in manipulator planning. Our algorithm and analysis are sufficiently general that they may also be applied to task planning or other planning domains in which experience is useful.
- Report Numbers:
- E 1.99:sand--93-0242c
E 1.99: conf-931055--1
- Other Subject(s):
- Published through SciTech Connect.
6. international symposium of robotics research,Pittsburgh, PA (United States),2 Oct 1993.
Chen, Pang C.
- Funding Information:
View MARC record | catkey: 14112899