Sparse preconditioning for model predictive control

  title={Sparse preconditioning for model predictive control},
  author={A. Knyazev and A. Malyshev},
  journal={2016 American Control Conference (ACC)},
We propose fast O(N) preconditioning, where N is the number of gridpoints on the prediction horizon, for iterative solution of (non)-linear systems appearing in model predictive control methods such as forward-difference Newton-Krylov methods. The Continuation/GMRES method for nonlinear model predictive control, suggested by T. Ohtsuka in 2004, is a specific application of the Newton-Krylov method, which uses the GMRES iterative algorithm to solve a forward difference approximation of the… Expand
Preconditioned warm-started Newton-Krylov methods for MPC with discontinuous control
We present Newton-Krylov methods for efficient numerical solution of optimal control problems arising in model predictive control, where the optimal control is discontinuous. As in our earlier work,Expand
Preconditioned Krylov iterations and condensing in interior point MPC method
Abstract We investigate using Krylov subspace iterative methods in model predictive control (MPC), where the prediction model is given by linear or linearized systems with linear inequalityExpand
Least squares dynamics in Newton-Krylov Model Predictive Control
Modification of Newton-Krylov methods for nonlinear Model Predictive Control by using least squares to fit the state to the dynamics and some constraints on the state, if they are inconsistent. Expand
Preconditioning of conjugate gradient iterations in interior point MPC method
It is proposed reusing matrix decompositions of direct solvers as preconditioners in Krylov-subspace methods applied to subsequent iterations of the interior point method, which results in at least halving its asymptotic computational complexity. Expand


Preconditioning for continuation model predictive control
This work suggests simplifying the construction of the preconditioner, by approximately solving a forward recursion for the state and a backward recursions for the costate, or simply reusing previously computed solutions. Expand
Preconditioned Continuation Model Predictive Control
CNMPC is adopted for challenging minimum-time problems, and performance is improved by introducing efficient preconditioning, utilizing parallel computing, and substituting MINRES for GMRES. Expand
Application of Interior-Point Methods to Model Predictive Control
We present a structured interior-point method for the efficient solution of the optimal control problem in model predictive control. The cost of this approach is linear in the horizon length,Expand
A fast well-conditioned interior point method for predictive control
Numerical results indicate that the computational complexity of the proposed method scales quadratically with the number of states and linearly with the horizon length, and the computational advantage is obtained by theuse of an inexact Newton method along with the use of novel preconditioners. Expand
Optimal preconditioning and iteration complexity bounds for gradient-based optimization in model predictive control
  • P. Giselsson
  • Computer Science, Mathematics
  • 2013 American Control Conference
  • 2013
This paper shows how to optimally precondition the optimization data by solving a semidefinite program, where optimally refers to the preconditionsing that minimizes an explicit iteration complexity bound. Expand
Continuation model predictive control on smooth manifolds
An algorithm for on-line controller implementation is proposed and its numerical effectiveness for a test problem on a hemisphere is demonstrated and it is extended to a case where the state is implicitly constrained to a smooth manifold. Expand
A continuation/GMRES method for fast computation of nonlinear receding horizon control
The proposed algorithm for nonlinear receding horizon control is applied to a two-link arm whose dynamics is highly nonlinear and is shown to be bounded under some conditions. Expand
Efficient particle continuation model predictive control
An algorithm for on-line controller implementation of continuation MPC for ensembles of predictions corresponding to various anticipated changes is proposed and its numerical effectiveness for a test minimum time problem arriving to a destination is demonstrated. Expand
Preconditioners for inexact interior point methods for predictive control
This paper presents a new method for solving a linear discrete-time finite horizon optimal control problem (FHOCP) with quadratic cost and linear constraints on the states and inputs. Such a FHOCPExpand
First-order methods in embedded nonlinear model predictive control
This paper investigates the suitability of Nesterov's fast gradient method algorithms in a nonlinear MPC setup by combining different first-order methods with the real-time iteration (RTI) scheme for non linear MPC and using them to control a model of an inverted pendulum. Expand