This study explores two issues in the transition from mp-MPC theory to the implementation of an industrial controller: offset-free output-feedback tracking, and controller tuning based on local linear analysis of the closed-loop system. It is shown that the disturbance-estimation based offset-free tracking schemes involving an observer/estimator, known from on-line MPC, are also applicable in mp-MPC. Further, that a "joint"-scheme, in which the MPC controller integrates the functions of constrained dynamic control and offset-free tracking without using a separate target calculator, may be efficiently used for control of small-scale multivariable processes with redundant control inputs. Such schemes facilitate tuning for efficient disturbance rejection and robustness using local linear analysis, which was found to be extremely valuable tuning tool. An experimental case study on a two-input single-output system for pressure control in the vacuum chamber of a wire annealer is presented.
COBISS.SI-ID: 25808423
The recently developed methods of explicit (multi-parametric) model predictive control (e-MPC) for hybrid systems provide an interesting opportunity for solving a class of nonlinear control problems. With this approach, the nonlinear process is approximated by a piecewise affine (PWA) hybrid model containing a set of local linear dynamics. Compared to linear-model-based MPC, a performance improvement is expected with the reduction of the plant-to-model mismatch; however at a cost of controller computation complexity. In order to reduce the computational load, so that desired horizon lengths may be used, we present an efficient sub-optimal solution. The feasibility of the approach for the application was evaluated in an experimental case study, where an output feedback, offset-free-tracking hybrid e-MPC controller was considered as a replacement for a PID-controller-based scheme for the control of the pressure in a wire-annealing machine.
COBISS.SI-ID: 23705895
In this work we explore advanced control algorithms for the vertical stabilization of plasma in the ITER tokamak for the case where a combination of ohmic in-vessel and superconducting poloidal actuators is used for effective response to disturbances subject to thermal constraints. We apply constrained linear-quadratic optimal control, which is a hybrid between conventional linear quadratic optimal control and model predictive control (MPC). We discuss the issues of practical implementation in the form of simplified explicit MPC, which allows application to fast processes by avoiding the use of on-line optimization. A computationally tractable explicit MPC controller for the plasma VS system capable of practically useful constraint-handling was demonstrated. Softening of the output constraints was required for feasibility, while computational tractability was reached by using move blocking and sparse placement of constraints.
COBISS.SI-ID: 26592551
The paper addresses the problem of numerical issues and degeneracies in the parametric quadratic programming (pQP) algorithm, used for computing partitions of explicit model predictive controllers (eMPC) with the Multi-Parametric Toolbox (MPT). We summarise the pQP problem setup and the basic algorithm, analyse its implementation in MPT, expose the numerical issues and suggest a series of improvements for more reliable operation, which are relevant also for other pQP solvers.
COBISS.SI-ID: 25004839
The paper addresses the problem of numerical issues and degeneracies arising when solving the multi-parametric linear complementarity problem (pLCP) for the purpose of computing partitions of explicit model predictive controllers (eMPC) based on constrained linear models and a 2-norm cost function. After summarizing the basic pLCP algorithm, several numerical issues relevant for reliable computation of eMPC partitions are exposed. The improved performance is illustrated with an eMPC controller example, which poses a problem for the available multi-parametric quadratic programming (pQP) solvers.
COBISS.SI-ID: 25715751