Solving Mathematical Programs with Complementarity Constraints Arising in Nonsmooth Optimal Control

被引:1
作者
Nurkanovic, Armin [1 ]
Pozharskiy, Anton [1 ]
Diehl, Moritz [1 ,2 ]
机构
[1] Univ Freiburg, Dept Microsyst Engn IMTEK, Syst Control & Optimizat Lab, Freiburg, Germany
[2] Univ Freiburg, Dept Math, Freiburg, Germany
关键词
MPECs; Nonlinear programming; Optimal control; Benchmark; INTERIOR-POINT METHOD; GLOBAL CONVERGENCE; REGULARIZATION SCHEME; DYNAMICAL-SYSTEMS; RELAXATION SCHEME; PENALTY METHOD; OPTIMIZATION; STATIONARITY; ALGORITHM; CONVEX;
D O I
10.1007/s10013-024-00704-z
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
This paper examines solution methods for mathematical programs with complementarity constraints (MPCC) obtained from the time-discretization of optimal control problems (OCPs) subject to nonsmooth dynamical systems. The MPCC theory and stationarity concepts are reviewed and summarized. The focus is on relaxation-based methods for MPCCs, which solve a (finite) sequence of more regular nonlinear programs (NLP), where a regularization/homotopy parameter is driven to zero. Such methods perform reasonably well on currently available benchmarks. However, these results do not always generalize to MPCCs obtained from nonsmooth OCPs. To provide a more complete picture, this paper introduces a novel benchmark collection of such problems, which we call NOSBENCH. The problem set includes 603 different MPCCs and we split it into a few representative subsets to accelerate the testing. We compare different relaxation-based methods, NLP solvers, homotopy parameter update and relaxation parameter steering strategies. Moreover, we check whether the obtained stationary points allow first-order descent directions, which may be the case for some of the weaker MPCC stationarity concepts. In the best case, the Scholtes' relaxation (SIAM J. Optim. 11, 918-936, 2001) with IPOPT (Math. Program. 106, 25-57, 2006) as NLP solver manages to solve 73.8% of the problems. This highlights the need for further improvements in algorithms and software for MPCCs.
引用
收藏
页码:659 / 697
页数:39
相关论文
共 103 条
[41]   A New Augmented Lagrangian Method for MPCCs-Theoretical and Numerical Comparison with Existing Augmented Lagrangian Methods [J].
Guo, Lei ;
Deng, Zhibin .
MATHEMATICS OF OPERATIONS RESEARCH, 2022, 47 (02) :1229-1246
[42]  
Gurobi Optimization LLC, 2023, Gurobi optimizer reference manual
[43]  
Hall J., 2023, IN PRESS
[44]   A Sequential Convex Programming Approach to Solving Quadratic Programs and Optimal Control Problems With Linear Complementarity Constraints [J].
Hall, Jonas ;
Nurkanovic, Armin ;
Messerer, Florian ;
Diehl, Moritz .
IEEE CONTROL SYSTEMS LETTERS, 2022, 6 :536-541
[45]  
Halm M., 2021, ARXIV
[46]  
Hatz K., 2013, ANLMCSP40760613
[47]   Projected dynamical systems in a complementarity formalism [J].
Heemels, WPMH ;
Schumacher, JM ;
Weiland, S .
OPERATIONS RESEARCH LETTERS, 2000, 27 (02) :83-91
[48]   Theoretical and numerical comparison of relaxation methods for mathematical programs with complementarity constraints [J].
Hoheisel, Tim ;
Kanzow, Christian ;
Schwartz, Alexandra .
MATHEMATICAL PROGRAMMING, 2013, 137 (1-2) :257-288
[49]   Trajectory Optimization with Optimization-Based Dynamics [J].
Howell, Taylor A. ;
Le Cleac'h, Simon ;
Singh, Sumeet ;
Florence, Pete ;
Manchester, Zachary ;
Sindhwani, Vikas .
IEEE ROBOTICS AND AUTOMATION LETTERS, 2022, 7 (03) :6750-6757
[50]   Convergence of a penalty method for mathematical programming with complementarity constraints [J].
Hu, XM ;
Ralph, D .
JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS, 2004, 123 (02) :365-390