Title :
Unique polyhedral representations of continuous selections for convex multiparametric quadratic programs
Author :
Spjotvold, J. ; Tondel, P. ; Johansen, T.A.
Author_Institution :
Dept. of Eng. Cybern., Norwegian Univ. of Sci. & Technol., Trondheim, Norway
Abstract :
A method which ensures uniqueness of the polyhedral partition of the parameter space for convex multiparametric quadratic programs (mp-QP) is presented. When the mp-QP has non-unique solutions, the norm of the solution vector is minimized and a unique affine optimizer function is identified for each region. It is proven, that under certain assumptions on the problem data, a globally continuous piecewise affine optimizer function is selected.
Keywords :
predictive control; quadratic programming; constrained optimal control; continuous selections; convex multiparametric quadratic programs; explicit model predictive control; globally continuous piecewise affine optimizer function; polyhedral partition; quadratic programming; unique polyhedral representations; Cybernetics; Electronic mail; Optimal control; Optimization methods; Partitioning algorithms; Predictive control; Predictive models; Quadratic programming; Space technology; Stability;
Conference_Titel :
American Control Conference, 2005. Proceedings of the 2005
Conference_Location :
Portland, OR, USA
Print_ISBN :
0-7803-9098-9
Electronic_ISBN :
0743-1619
DOI :
10.1109/ACC.2005.1470060