13.07.2015 Views

Assessment and Future Directions of Nonlinear Model Predictive ...

Assessment and Future Directions of Nonlinear Model Predictive ...

Assessment and Future Directions of Nonlinear Model Predictive ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Robust <strong>Model</strong> <strong>Predictive</strong> Control for ObstacleAvoidance: Discrete Time CaseSaša V. Raković <strong>and</strong> David Q. MayneImperial College London, London SW7 2BT, United Kingdomsasa.rakovic@imperial.ac.uk, d.mayne@imperial.ac.uk1 IntroductionThe importance <strong>of</strong> the obstacle avoidance problem is stressed in [4]. Computation<strong>of</strong> reachability sets for the obstacle avoidance problem is addressed, for continuoustimesystems in [4, 5] <strong>and</strong> for discrete-time systems in [12]; further results appear in,for instance [2, 17, 18]. The obstacle avoidance problem is inherently non–convex.Most existing results are developed for the deterministic case when external disturbancesare not present. The main purpose <strong>of</strong> this paper is to demonstrate thatthe obstacle avoidance problem in the discrete time setup has considerable structureeven when disturbances are present. We extend the robust model predictiveschemes using tubes (sequences <strong>of</strong> sets <strong>of</strong> states) [9, 11, 14] to address the robustobstacle avoidance problem <strong>and</strong> provide a mixed integer programming algorithmfor robust control <strong>of</strong> constrained linear systems that are required to avoid specifiedobstacles. The resultant robust optimal control problem that is solved on–line hasmarginally increased complexity compared with that required for model predictivecontrol for obstacle avoidance in the deterministic case.This paper is organized as follows. Section 2 discusses the general idea <strong>of</strong> robustcontrol invariant tubes for the obstacle avoidance problem. Section 3 considersin more detail the case when system being controlled is linear. Section 4 presentsa simple tube controller, establishes its properties <strong>and</strong> provides an illustrativeexample. Finally, Section 5 gives conclusions <strong>and</strong> indicates further extensions.Notation: Let N {0, 1, 2,...} <strong>and</strong> N q {0, 1,...,q} for q ∈ N. Apolyhedronis the (convex) intersection <strong>of</strong> a finite number <strong>of</strong> open <strong>and</strong>/or closedhalf-spaces, a polytope is the closed <strong>and</strong> bounded polyhedron <strong>and</strong> a closed(open) polygon is the union <strong>of</strong> a finite number <strong>of</strong> polytopes (polyhedra). Giventwo sets U ⊂ R n <strong>and</strong> V ⊂ R n , the Minkowski set addition is defined byU⊕V {u + v | u ∈U, v ∈V}, the Minkowski/Pontryagin set differenceis: U⊖V {x | x ⊕V ⊆U}. The distance <strong>of</strong> a point z from a set X is denotedby d(z,X) inf{|z − x| |x ∈ X}.2 Problem FormulationWe consider the following discrete-time, time-invariant system:x + = f(x, u, w) (1)R. Findeisen et al. (Eds.): <strong>Assessment</strong> <strong>and</strong> <strong>Future</strong> <strong>Directions</strong>, LNCIS 358, pp. 617–627, 2007.springerlink.com c○ Springer-Verlag Berlin Heidelberg 2007

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!