Distributionally robust stochastic optimal control

被引：0
作者：
Shapiro, Alexander [1 ]
Li, Yan [2 ]
机构：
[1] Georgia Inst Technol, Atlanta, GA 30332 USA
[2] Texas A&M Univ, College Stn, TX 77843 USA
来源：
OPERATIONS RESEARCH LETTERS | 2025年 / 60卷
关键词：
Stochastic optimal control; Dynamic equations distributional robustness; Game formulation; Risk measures;
D O I：
10.1016/j.orl.2024.107234
中图分类号：
C93 [管理学]; O22 [运筹学];
学科分类号：
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要：
The main goal of this paper is to discuss construction of distributionally robust counterparts of stochastic optimal control problems. Randomized and non-randomized policies are considered. In particular necessary and sufficient conditions for existence of non-randomized policies are given.
引用
收藏
页数：6
相关论文
共 18 条
[1]  
[Anonymous], 2000, Foundations of Inventory Management
[2]  
Bertsekas D, 1978, MATH SCI ENG
[3]  
BILLINGSLEY P, 1986, PROBABILITY MEASURE
[4]  
Bonnans J. F., 2013, Springer Series in Operations Research
[5]   "Dice"-sion-Making Under Uncertainty: When Can a Random Decision Reduce Risk? [J].
Delage, Erick ;
Kuhn, Daniel ;
Wiesemann, Wolfram .
MANAGEMENT SCIENCE, 2019, 65 (07) :3282-3301
[6]  
Dellacherie C., 1988, Probabilities and Potential C
[7]  
Guide A. H., 2006, Infinite dimensional analysis
[8]   Robust dynamic programming [J].
Iyengar, GN .
MATHEMATICS OF OPERATIONS RESEARCH, 2005, 30 (02) :257-280
[9]  
Jaskiewicz A., 2016, Handbook of Dynamic Game Theory
[10]  
Li Y, 2024, Arxiv, DOI arXiv:2308.11139
12