Model predictive control: A multi-parametric programming approach

doi:10.1016/S1570-7946(00)80052-8

Computer Aided Chemical Engineering

Volume 8, 2000, Pages 301-306

https://doi.org/10.1016/S1570-7946(00)80052-8 Get rights and content

In this paper, linear model predictive control problems are formulated as multi-parametric quadratic programs, where the control variables are treated as optimization variables and the state variables as parameters. It is shown that the control variables are affine functions of the state variables and each of these affine functions is valid in a certain polyhedral region in the space of state variables. An approach for deriving the explicit expressions of all the affine functions and their corresponding polyhedral regions is presented. The key advantage of this approach is that the control actions are computed off-line: the on-line computation simply reduces to a function evaluation problem.

References (8)

M. Morari et al.
Comput. Chem. Engng.
(1999)
E. Zafiriou
Comput. Chem. Engng.
(1990)
A. Bemporad et al.
Automatica
(1999)
J. Acevedo et al.
Ind. Eng. Chem. Res.
(1997)

There are more references available in the full text version of this article.

Cited by (14)

Optimal operation and control of intensified processes — challenges and opportunities
2019, Current Opinion in Chemical Engineering
Citation Excerpt :
To address the remaining challenges related to large-scale problems, continuous efforts on the development of parallel computing, software and algorithms should be encouraged. The computational complexity of model predictive controllers can be reduced with the implementation of multi-parametric mp-MPC, which enables the offline solution of MPC problem by reformulating it into a multi-parametric optimization problem and obtaining explicit control laws [27,28]. Such strategy has been applied to pressure swing adsorption systems by Khajuria and Pistikopoulos [29] and extended in Khajuria and Pistikopoulos [30].
Process systems engineering (PSE) tools can be utilized to enable the optimal operation and control of intensified processes. In this work, the challenges in the control of intensified processes are summarized, including the difficulties in modelling and performing online optimization of these highly complex dynamic systems. Nevertheless, PSE progress in the areas of nonlinear programming, reduced-order modeling, development of software tools, parallel computing and artificial intelligence are enabling the implementation and optimization of large-scale intensified systems. Future opportunities for research are identified, including the simultaneous optimization of scheduling of operation and control, a promising strategy that has proven its value in regular operations, and can expand the operating windows of intensified processes.
Embedded linear model predictive control for 8-bit microcontrollers via convex lifting
2017, IFAC-PapersOnLine
This paper demonstrates an efficient approach for the implementation of parametric programming solutions to linear model predictive control problems. The emphasis is put namely on embedded hardware applications where feasible size and online evaluation of the controller code can be mission critical. As the explicit solutions are known to become quickly intractable for such a purpose once the amount of problem data grows, we exploit the concept of convex lifting of a given parameter space partition, and use it to avoid its storage and to replace the traditional point location with a significantly less costly procedure. In this light we practically illustrate the deployment of generated code on an 8-bit microcontroller unit, representing the class of low-cost embedded controllers built into mass-produced devices, in an active vibration suppression case study accompanied by a detailed performance, memory and timing analysis.
On multi-parametric programming and its applications in process systems engineering
2016, Chemical Engineering Research and Design
Citation Excerpt :
Therefore, most algorithms which are applicable to mp-QP problems are inherently also applicable to mp-LP problems. Based on the results from the Basic Sensitivity Theorem, a string of papers in 2000 (Bemporad et al., 2000c,b; Pistikopoulos et al., 2000), as well as the famous paper by Bemporad et al. (2002a), described a geometrical approach for the solution of mp-QP problems, which relies on the exploration of the parameter space by moving from one critical region to another. As the initial algorithm was shown to generate a large number of artificial cuts (Tøndel et al., 2003b), new algorithms based on the geometrical principle were described, based on variable step-size (Baotic, 2002), inference of the active set of adjacent critical regions (Tøndel et al., 2003b,c) and combination of the inference of the active set with the original algorithm (Spjøtvold et al., 2006; Spjotvold et al., 2006a).
Embedded explicit model predictive vibration control
2016, Mechatronics
Citation Excerpt :
The other main class of efficient MPC formulations preserves the optimality of the problem, while still removing some of the burden of solving the task in real time [11]. The best known representative formulation of the latter class rests on the idea of transferring the computational complexity from on-line control into off-line, and is called multi-parametric MPC (MPMPC) or explicit MPC (EMPC) [2,3]. In EMPC, the optimization problem is essentially solved using parametric programming beforehand, formulating the MPC control problem as a set of regions in state space to which linear control laws are assigned.
This article presents an embedded active vibration suppression system featuring real-time explicit model predictive control (EMPC) that is implemented on a microcontroller unit (MCU). The EMPC controller minimizes the tip deflection of an aluminum cantilever beam driven by piezoceramic actuators, gaining its feedback from direct position measurements. The output and input performance of the EMPC method is compared to an analogously tuned positive position feedback (PPF) controller. An extensive analysis is provided on the cycle timing and memory needs of the explicit predictive vibration control scheme. The results demonstrate that the EMPC controller may achieve the same vibration suppression results compared to PPF with less input effort, while inherently respecting process constraints. Furthermore, we show that EMPC task execution timing is comparable in the random access memory (RAM) and read only memory (ROM) alternatives, suggesting that numerous current microcontrollers are suitable for EMPC-based active vibration control, in case the prediction model is kept simple.
Integrated production scheduling and process control: A systematic review
2014, Computers and Chemical Engineering
Production scheduling and process control have the common aim of identifying (economically) optimal operational decisions, and it is reasonable to expect that significant economic benefits can be derived from their integration. Yet, the scheduling and control fields have evolved quite independently of each other, and efforts aimed at integrating these two decision-making activities are quite recent. In this paper, we review progress made thus far in this direction. We identify key elements of control and scheduling, and carry out a systematic investigation of their use as building blocks for the formulation and solution of the integrated scheduling/control problem. On the basis of our review, we define several necessary directions for future development as well as a complement of promising applications.
Dynamic optimization of an industrial evaporator using graph search with embedded nonlinear programming
2006, Analysis and Design of Hybrid Systems 2006
In this paper, we consider the task to start the operation of an industrial evaporation system. Rigorous modelling gives rise to a hybrid automaton with large nonlinear DAE-models that describe the continuous evolution in the discrete locations. The optimization problem is solved by a hierarchical procedure that consists of a branch-and-bound algorithm with embedded nonlinear dynamic optimization over a finite look-ahead horizon. Important elements of the algorithm are the introduction of a dynamic choice of the time intervals over which the controls are constant and of tailored penalty functions in order to obtain solutions which are close to infeasible trajectories. Copyright © 2006 IFAC

View all citing articles on Scopus

View full text

Model predictive control: A multi-parametric programming approach

Comput. Chem. Engng.

Comput. Chem. Engng.

Automatica

Ind. Eng. Chem. Res.