Robust Finite-State Controllers for Uncertain POMDPs

被引:0
作者
Cubuktepe, Murat [1 ]
Jansen, Nils [2 ]
Junges, Sebastian [3 ]
Marandi, Ahmadreza [4 ]
Suilen, Marnix [2 ]
Topcu, Ufuk [1 ]
机构
[1] Univ Texas Austin, Dept Aerosp Engn & Engn Mech, Austin, TX 78712 USA
[2] Radboud Univ Nijmegen, Dept Software Sci, Nijmegen, Netherlands
[3] Univ Calif Berkeley, Dept Elect Engn & Comp Sci, Berkeley, CA 94720 USA
[4] Eindhoven Univ Technol, Dept Ind Engn & Innovat Sci, Eindhoven, Netherlands
来源
THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE | 2021年 / 35卷
关键词
MARKOV DECISION-PROCESSES;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Uncertain partially observable Markov decision processes (uPOMDPs) allow the probabilistic transition and observation functions of standard POMDPs to belong to a so-called uncertainty set. Such uncertainty, referred to as epistemic uncertainty, captures uncountable sets of probability distributions caused by, for instance, a lack of data available. We develop an algorithm to compute finite-memory policies for uPOMDPs that robustly satisfy specifications against any admissible distribution. In general, computing such policies is theoretically and practically intractable. We provide an efficient solution to this problem in four steps. (1) We state the underlying problem as a nonconvex optimization problem with infinitely many constraints. (2) A dedicated dualization scheme yields a dual problem that is still nonconvex but has finitely many constraints. (3) We linearize this dual problem and (4) solve the resulting finite linear program to obtain locally optimal solutions to the original problem. The resulting problem formulation is exponentially smaller than those resulting from existing methods. We demonstrate the applicability of our algorithm using large instances of an aircraft collision-avoidance scenario and a novel spacecraft motion planning case study.
引用
收藏
页码:11792 / 11800
页数:9
相关论文
共 31 条
  • [1] Ahmadi M, 2018, ANN ALLERTON CONF, P115, DOI 10.1109/ALLERTON.2018.8635859
  • [2] Second-order cone programming
    Alizadeh, F
    Goldfarb, D
    [J]. MATHEMATICAL PROGRAMMING, 2003, 95 (01) : 3 - 51
  • [3] Optimizing fixed-size stochastic controllers for POMDPs and decentralized POMDPs
    Amato, Christopher
    Bernstein, Daniel S.
    Zilberstein, Shlomo
    [J]. AUTONOMOUS AGENTS AND MULTI-AGENT SYSTEMS, 2010, 21 (03) : 293 - 320
  • [4] [Anonymous], 2018, Successive convexification: A superlinearly convergent algorithm for non-convex optimal control problems
  • [5] [Anonymous], 2015, DECISION MAKING UNCE
  • [6] Ben-Tal A, 2009, PRINC SER APPL MATH, P27
  • [7] Theory and Applications of Robust Optimization
    Bertsimas, Dimitris
    Brown, David B.
    Caramanis, Constantine
    [J]. SIAM REVIEW, 2011, 53 (03) : 464 - 501
  • [8] Sampling-based motion planning with sensing uncertainty
    Burns, Brendan
    Brock, Oliver
    [J]. PROCEEDINGS OF THE 2007 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION, VOLS 1-10, 2007, : 3313 - +
  • [9] Optimal cost almost-sure reachability in POMDPs
    Chatterjee, Krishnendu
    Chmelik, Martin
    Gupta, Raghav
    Kanodia, Ayush
    [J]. ARTIFICIAL INTELLIGENCE, 2016, 234 : 26 - 48
  • [10] A STORM is Coming: A Modern Probabilistic Model Checker
    Dehnert, Christian
    Junges, Sebastian
    Katoen, Joost-Pieter
    Volk, Matthias
    [J]. COMPUTER AIDED VERIFICATION (CAV 2017), PT II, 2017, 10427 : 592 - 600