TY - GEN
T1 - Heuristic Search in Dual Space for Constrained Fixed-Horizon POMDPs with Durative Actions
AU - Khonji, Majid
AU - Khalifa, Duoaa
N1 - Publisher Copyright:
Copyright © 2023, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
PY - 2023/6/27
Y1 - 2023/6/27
N2 - The Partially Observable Markov Decision Process (POMDP) is widely used in probabilistic planning for stochastic domains. However, current extensions, such as constrained and chance-constrained POMDPs, have limitations in modeling real-world planning problems because they assume that all actions have a fixed duration. To address this issue, we propose a unified model that encompasses durative POMDP and its constrained extensions. To solve the durative POMDP and its constrained extensions, we first convert them into an Integer Linear Programming (ILP) formulation. This approach leverages existing solvers in the ILP literature and provides a foundation for solving these problems. We then introduce a heuristic search approach that prunes the search space, which is guided by solving successive partial ILP programs. Our empirical evaluation results show that our approach outperforms the current state-of-the-art fixed-horizon chance-constrained POMDP solver.
AB - The Partially Observable Markov Decision Process (POMDP) is widely used in probabilistic planning for stochastic domains. However, current extensions, such as constrained and chance-constrained POMDPs, have limitations in modeling real-world planning problems because they assume that all actions have a fixed duration. To address this issue, we propose a unified model that encompasses durative POMDP and its constrained extensions. To solve the durative POMDP and its constrained extensions, we first convert them into an Integer Linear Programming (ILP) formulation. This approach leverages existing solvers in the ILP literature and provides a foundation for solving these problems. We then introduce a heuristic search approach that prunes the search space, which is guided by solving successive partial ILP programs. Our empirical evaluation results show that our approach outperforms the current state-of-the-art fixed-horizon chance-constrained POMDP solver.
UR - http://www.scopus.com/inward/record.url?scp=85167986757&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85167986757
T3 - Proceedings of the 37th AAAI Conference on Artificial Intelligence, AAAI 2023
SP - 14927
EP - 14936
BT - AAAI-23 Special Tracks
A2 - Williams, Brian
A2 - Chen, Yiling
A2 - Neville, Jennifer
T2 - 37th AAAI Conference on Artificial Intelligence, AAAI 2023
Y2 - 7 February 2023 through 14 February 2023
ER -