Constrained generalized predictive control using relaxation with quadratic penalization

Abstract
In this paper an iterative algorithm will be derived to find a numerical solution for the generalized predictive control of linear systems with constraints on controlled and manipulated variables. The algorithm uses the relaxation method with quadratic penalization. Conditions are proposed for the convergence of the iterative solution and results of the application to a second order linear system are presented.
Description
Keywords
Predictive control, Prediction algorithms, Algorithm design and analysis, Vectors, Linear programming, Europe, Convergence
Citation