DocumentCode :
3479153
Title :
Bounded Incremental Real-Time Dynamic Programming
Author :
Fan, Changjie ; Chen, Xiaoping
Author_Institution :
Dept. of Comput. Sci., Univ. of Sci. & Technol. of China, Hefei
fYear :
2007
fDate :
11-13 Oct. 2007
Firstpage :
637
Lastpage :
644
Abstract :
A real-time multi-step planning problem is characterized by alternating decision-making and execution processes, whole online decision-making time divided in slices between each execution, and the pressing need for policy that only relates to current step. We propose a new criterion to judge the optimality of a policy based on the upper and lower bound theory. This criterion guarantees that the agent can act earlier in a real-time decision process while an optimal policy with sufficient precision still remains. We prove that, under certain conditions, one can obtain an optimal policy with arbitrary precision using such an incremental method. We present a bounded incremental real-time dynamic programming algorithm (BIRTDP). In the experiments of two typical real-time simulation systems, BIRTDP outperforms the other state-of-the-art RTDP algorithms tested.
Keywords :
Markov processes; decision making; dynamic programming; bound theory; bounded incremental real-time dynamic programming algorithm; decision-making; multi-step planning problem; Algorithm design and analysis; Computer science; Decision making; Dynamic programming; Heuristic algorithms; Information technology; Process planning; Real time systems; Technology planning; Uncertainty;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Frontiers in the Convergence of Bioscience and Information Technologies, 2007. FBIT 2007
Conference_Location :
Jeju City
Print_ISBN :
978-0-7695-2999-8
Type :
conf
DOI :
10.1109/FBIT.2007.14
Filename :
4524180
Link To Document :
بازگشت