Elsevier

Automatica

Volume 27, Issue 1, January 1991, Pages 55-74
Automatica

Paper
Continuous-time generalized predictive control (CGPC)

https://doi.org/10.1016/0005-1098(91)90006-NGet rights and content

Abstract

A continuous-time version of the discrete-time Generalized Predictive Controller is presented. The continuous-time formulation arises from a mixture of two kinds of analogy between continuous and discrete-time systems: a physical analogy and an algebraic analogy. Emphasis is placed on the differences arising from a continuous-time formulation, and the relative merits of a continuous and a discrete-time approach are given. Although mainly concerned with the design algorithm itself, the paper also indicates how a self-tuning version can be implemented. Illustrative simulations are given.

References (24)

  • D.W. Clarke et al.

    Generalised Predictive Control. Part 1: the basic algorithm and Part 2: extensions and interpretations

    Automatica

    (1987)
  • D.W. Clarke et al.

    Long-range predictive control using weighting-sequence models

  • Cited by (136)

    • Efficient constrained model predictive control

      2014, European Journal of Control
      Citation Excerpt :

      In [28], MPC was applied to unmanned mobile vehicles, where an additional, variable prediction horizon was combined into a single optimization process, with the variable prediction points automatically distributed by the optimization process to effectively capture the environment. One drawback of using prediction points that are not at the base time step of the system is that discretization issues may occur, particularly over longer time periods, with [9,8] using continuous time models to improve on this subject. Our paper is based on a comparable method to move blocking named Algebraic Model Predictive Control (AMPC) [19], where the typical MPC problem is formulated with variable prediction time intervals, retaining prediction accuracy by using the state transition function evaluated algebraically with the eigenvalue/eigenvector characteristics of the system, thus avoiding inaccuracies due to discrete approximations.

    • Robustness analysis and tuning of generalized predictive control using frequency domain approaches

      2012, Applied Mathematical Modelling
      Citation Excerpt :

      Garriga et al. [15] have studied the effects of tuning parameters on the locations of closed-loop eigenvalues. Some texts are considered Continuous-time GPC (CGPC), e.g., Demircioglu and Gawthrop [16], as a more robust scheme compared to the alternative Discrete-time GPC (DGPC), but the heuristic tuning problem still remained [17]. Demircioglu and Karasu [18] have performed a practical comparative study between CGPC and DGPC and sustained the positive effect of suitable choice T filter in closed loop robustness.

    View all citing articles on Scopus

    The original version of this paper was presented at the IFAC Symposium on Adaptive Systems in Control and Signal Processing which was held in Glasgow, Scotland, U.K. during April 1989. The published Proceedings of this IFAC Meeting may be ordered from: Pergamon Press plc, Headington Hill Hall, Oxford OX3 0BW, U.K. This paper was recommended for publication in revised form by Associate Editor D. W. Clarke under the direction of Editor P. C. Parks.

    View full text