A global piecewise smooth Newton method for fast large-scale model predictive control

In this paper, the strictly convex quadratic program (QP) arising in model predictive control (MPC) for constrained linear systems is reformulated as a system of piecewise affine equations. A regularized piecewise smooth Newton method with exact line search on a convex, differentiable, piecewise-quadratic merit function is proposed for the solution of the reformulated problem. The algorithm has considerable merits when applied to MPC over standard active set or interior point algorithms. Its performance is tested and compared against state-of-the-art QP solvers on a series of benchmark problems. The proposed algorithm is orders of magnitudes faster, especially for large-scale problems and long horizons. For example, for the challenging crude distillation unit model of Pannocchia, Rawlings, and Wright (2007) with 252 states, 32 inputs, and 90 outputs, the average running time of the proposed approach is 1.57?ms.

[1]  Wu Li.  Error Bounds for Piecewise Convex Quadratic Programs and Applications , 1995 .

[2]  L. Biegler,et al.  QPSchur: A dual, active-set, Schur-complement method for large-scale and structured convex quadratic programming , 2006 .

[3]  M. Ç. Pinar,et al.  On Newton's method for Huber's robust M-estimation problems in linear regression , 1998 .

[4]  Stefan Schäffler,et al.  Applied Mathematics and Parallel Computing , 1996 .

[5]  Wu Li,et al.  Regularized Newton Methods for Minimization of Convex Quadratic Splines with Singular Hessians , 1998 .

[6]  B. Eaves.  On Quadratic Programming , 1971 .

[7]  Dimitri P. Bertsekas,et al.  Nonlinear Programming , 1997 .

[8]  Eduardo F. Camacho,et al.  On the stability of constrained MPC without terminal constraint , 2006, IEEE Transactions on Automatic Control.

[9]  Olvi L. Mangasarian,et al.  A Finite Newton Method for Classi cation Problems , 2001 .

[10]  Timothy A. Davis,et al.  Modifying a Sparse Cholesky Factorization , 1999, SIAM J. Matrix Anal. Appl..

[11]  Z.-Q. Luo,et al.  Error bounds and convergence analysis of feasible descent methods: a general approach , 1993, Ann. Oper. Res..

[12]  Y. Nesterov.  A method for solving the convex programming problem with convergence rate O(1/k^2) , 1983 .

[13]  Manfred Morari,et al.  Real-time input-constrained MPC using fast gradient methods , 2009, Proceedings of the 48h IEEE Conference on Decision and Control (CDC) held jointly with 2009 28th Chinese Control Conference.

[14]  Gene H. Golub,et al.  Methods for modifying matrix factorizations , 1972, Milestones in Matrix Computation.

[15]  M. Kojima,et al.  EXTENSION OF NEWTON AND QUASI-NEWTON METHODS TO SYSTEMS OF PC^1 EQUATIONS , 1986 .

[16]  Stephen J. Wright,et al.  Numerical Optimization , 2018, Fundamental Statistical Inference.

[17]  Daniel Axehill,et al.  A dual gradient projection quadratic programming algorithm tailored for model predictive control , 2008, 2008 47th IEEE Conference on Decision and Control.

[18]  Stephen J. Wright.  Primal-Dual Interior-Point Methods , 1997, Other Titles in Applied Mathematics.

[19]  Donald Goldfarb,et al.  A numerically stable dual method for solving strictly convex quadratic programs , 1983, Math. Program..

[20]  B. Kouvaritakis,et al.  Efficient MPC Optimization using Pontryagin's Minimum Principle , 2006, Proceedings of the 45th IEEE Conference on Decision and Control.

[21]  Olvi L. Mangasarian,et al.  A finite newton method for classification , 2002, Optim. Methods Softw..

[22]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[23]  Sanjay Mehrotra,et al.  On the Implementation of a Primal-Dual Interior Point Method , 1992, SIAM J. Optim..

[24]  Bart De Schutter,et al.  Equivalence of hybrid dynamical models , 2001, Autom..

[25]  Jack Dongarra,et al.  LINPACK Users' Guide , 1987 .

[26]  F. Facchinei,et al.  Finite-Dimensional Variational Inequalities and Complementarity Problems , 2003 .

[27]  Wu Li,et al.  A New Algorithm for Solving Strictly Convex Quadratic Programs , 1997, SIAM J. Optim..

[28]  C. Mészáros.  The BPMPD interior point solver for convex quadratic problems , 1999 .

[29]  M. Fukushima,et al.  "Reformulation: Nonsmooth, Piecewise Smooth, Semismooth and Smoothing Methods" , 2010 .

[30]  Stephen J. Wright,et al.  Fast, large-scale model predictive control by partial enumeration , 2007, Autom..

[31]  Komei Fukuda,et al.  An output-sensitive algorithm for multi-parametric LCPs with sufficient matrices , 2008, 0807.2318.

[32]  M. Best.  An Algorithm for the Solution of the Parametric Quadratic Programming Problem , 1996 .

[33]  Stephen P. Boyd,et al.  Fast Model Predictive Control Using Online Optimization , 2010, IEEE Transactions on Control Systems Technology.

[34]  P. Tseng,et al.  On the linear convergence of descent methods for convex essentially smooth minimization , 1992 .

[35]  Wu Li,et al.  An Implementation of the QSPLINE Method for Solving Convex Quadratic Programming Problems With Simple Bound Constraints , 2003 .

[36]  Stephen J. Wright,et al.  Application of Interior-Point Methods to Model Predictive Control , 1998 .

[37]  David Q. Mayne,et al.  Constrained model predictive control: Stability and optimality , 2000, Autom..

[38]  Alberto Bemporad,et al.  The explicit linear quadratic regulator for constrained systems , 2003, Autom..

[39]  John E. Dennis,et al.  Numerical methods for unconstrained optimization and nonlinear equations , 1983, Prentice Hall series in computational mathematics.

[40]  H. B. Nielsen,et al.  Finite alogorithms for robust linear regression , 1990 .

[41]  H. J. Ferreau,et al.  An online active set strategy to overcome the limitations of explicit MPC , 2008 .