Bounds on risk-averse mixed-integer multi-stage stochastic programming problems with mean-CVaR

被引:13
|
作者
Mahmutogullari, Ali Irfan [1 ]
Cavus, Ozlem [1 ]
Akturk, M. Selim [1 ]
机构
[1] Bilkent Univ, Dept Ind Engn, TR-06800 Ankara, Turkey
关键词
Stochastic programming; Mixed-integer multi-stage stochastic; programming; Dynamic measures of risk; CVaR; Bounding; TIME CONSISTENCY; OPTIMIZATION; APPROXIMATIONS;
D O I
10.1016/j.ejor.2017.10.038
中图分类号
C93 [管理学];
学科分类号
12 ; 1201 ; 1202 ; 120202 ;
摘要
Risk-averse mixed-integer multi-stage stochastic programming forms a class of extremely challenging problems since the problem size grows exponentially with the number of stages, the problem is non convex due to integrality restrictions, and the objective function is nonlinear in general. We propose a scenario tree decomposition approach, namely group subproblem approach, to obtain bounds for such problems with an objective of dynamic mean conditional value-at-risk (mean-CVaR). Our approach does not require any special problem structure such as convexity and linearity, therefore it can be applied to a wide range of problems. We obtain lower bounds by using different convolution of mean-CVaR risk measures and different scenario partition strategies. The upper bounds are obtained through the use of optimal solutions of group subproblems. Using these lower and upper bounds, we propose a solution algorithm for risk-averse mixed-integer multi-stage stochastic problems with mean-CVaR risk measures. We test the performance of the proposed algorithm on a multi-stage stochastic lot sizing problem and compare different choices of lower bounds and partition strategies. Comparison of the proposed algorithm to a commercial solver revealed that, on the average, the proposed algorithm yields 1.13% stronger bounds. The commercial solver requires additional running time more than a factor of five, on the average, to reach the same optimality gap obtained by the proposed algorithm. (C) 2017 Elsevier B.V. All rights reserved.
引用
收藏
页码:595 / 608
页数:14
相关论文
共 50 条
  • [31] Simultaneous Batching and Scheduling in Multi-product Multi-stage Batch Plants through Mixed-Integer Linear Programming
    Kopanos, Georgios M.
    Puigjaner, Luis
    PRES 2010: 13TH INTERNATIONAL CONFERENCE ON PROCESS INTEGRATION, MODELLING AND OPTIMISATION FOR ENERGY SAVING AND POLLUTION REDUCTION, 2010, 21 : 505 - 510
  • [32] Optimal Control of Microgrids with Multi-stage Mixed-integer Nonlinear Programming Guided Q-learning Algorithm
    Yoldas, Yeliz
    Goren, Selcuk
    Onen, Ahmet
    JOURNAL OF MODERN POWER SYSTEMS AND CLEAN ENERGY, 2020, 8 (06) : 1151 - 1159
  • [33] Optimal Control of Microgrids with Multi-stage Mixed-integer Nonlinear Programming Guided Q-learning Algorithm
    Yeliz Yoldas
    Selcuk Goren
    Ahmet Onen
    JournalofModernPowerSystemsandCleanEnergy, 2020, 8 (06) : 1151 - 1159
  • [34] Constrained dynamic programming of mixed-integer linear problems by multi-parametric programming
    Rivotti, Pedro
    Pistikopoulos, Efstratios N.
    COMPUTERS & CHEMICAL ENGINEERING, 2014, 70 : 172 - 179
  • [35] Multi-stage risk-averse operation of integrated electric power and natural gas systems
    Sun, Guoqiang
    Sun, Jinghong
    Chen, Sheng
    Wei, Zhinong
    INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS, 2021, 126
  • [36] An interval-parameter mean-CVaR two-stage stochastic programming approach for waste management under uncertainty
    C. Dai
    X. H. Cai
    Y. P. Cai
    Q. Huo
    Y. Lv
    G. H. Huang
    Stochastic Environmental Research and Risk Assessment, 2014, 28 : 167 - 187
  • [37] Postoptimality for mean-risk stochastic mixed-integer programs and its application
    Zhiping Chen
    Feng Zhang
    Li Yang
    Mathematical Methods of Operations Research, 2011, 74 : 445 - 465
  • [38] Postoptimality for mean-risk stochastic mixed-integer programs and its application
    Chen, Zhiping
    Zhang, Feng
    Yang, Li
    MATHEMATICAL METHODS OF OPERATIONS RESEARCH, 2011, 74 (03) : 445 - 465
  • [40] An algorithm for two-stage stochastic mixed-integer nonlinear convex problems
    E. Mijangos
    Annals of Operations Research, 2015, 235 : 581 - 598