Uncertain convex programs: randomized solutions and confidence levels

被引:468
作者
Calafiore, G
Campi, MC
机构
[1] Politecn Torino, Dipartimento Automat & Informat, I-10129 Turin, Italy
[2] Univ Brescia, Dipartimento Automat Automaz, I-25123 Brescia, Italy
关键词
D O I
10.1007/s10107-003-0499-y
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Many engineering problems can be cast as optimization problems subject to convex constraints that are parameterized by an uncertainty or 'instance' parameter. Two main approaches are generally available to tackle constrained optimization problems in presence of uncertainty: robust optimization and chance-constrained optimization. Robust optimization is a deterministic paradigm where one seeks a solution which simultaneously satisfies all possible constraint instances. In chance-constrained optimization a probability distribution is instead assumed on the uncertain parameters, and the constraints are enforced up to a pre-specified level of probability. Unfortunately however, both approaches lead to computationally intractable problem formulations. In this paper, we consider an alternative 'randomized' or 'scenario' approach for dealing with uncertainty in optimization, based on constraint sampling. In particular, we study the constrained optimization problem resulting by taking into account only a finite set of N constraints, chosen at random among the possible constraint instances of the uncertain problem. We show that the resulting randomized solution fails to satisfy only a small portion of the original constraints, provided that a sufficient number of samples is drawn. Our key result is to provide an efficient and explicit bound on the measure ( probability or volume) of the original constraints that are possibly violated by the randomized solution. This volume rapidly decreases to zero as N is increased.
引用
收藏
页码:25 / 46
页数:22
相关论文
共 34 条
[1]   Parameterized LMIs in control theory [J].
Apkarian, P ;
Tuan, HD .
SIAM JOURNAL ON CONTROL AND OPTIMIZATION, 2000, 38 (04) :1241-1264
[2]   On avoiding vertexization of robustness problems: The approximate feasibility concept [J].
Barmish, BR ;
Shcherbakov, PS .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2002, 47 (05) :819-824
[3]   Robust convex optimization [J].
Ben-Tal, A ;
Nemirovski, A .
MATHEMATICS OF OPERATIONS RESEARCH, 1998, 23 (04) :769-805
[4]   Robust truss topology design via semidefinite programming [J].
Ben-Tal, A ;
Nemirovski, A .
SIAM JOURNAL ON OPTIMIZATION, 1997, 7 (04) :991-1016
[5]   Robust solutions of uncertain linear programs [J].
Ben-Tal, A ;
Nemirovski, A .
OPERATIONS RESEARCH LETTERS, 1999, 25 (01) :1-13
[6]   On tractable approximations of uncertain linear matrix inequalities affected by interval uncertainty [J].
Ben-Tal, A ;
Nemirovski, A .
SIAM JOURNAL ON OPTIMIZATION, 2002, 12 (03) :811-833
[7]   A probabilistic framework for problems with real structured uncertainty in systems and control [J].
Calafiore, G ;
Dabbene, F .
AUTOMATICA, 2002, 38 (08) :1265-1276
[8]   Stochastic algorithms for exact and approximate feasibility of robust LMIs [J].
Calafiore, G ;
Polyak, BT .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2001, 46 (11) :1755-1759
[9]  
CALAFIORE G, 2002, P IFAC 2002 WORLD C
[10]   CHANCE-CONSTRAINED PROGRAMMING [J].
CHARNES, A ;
COOPER, WW .
MANAGEMENT SCIENCE, 1959, 6 (01) :73-79