A first step towards anytime invariant quasi static feedback for real time tracking

Control laws synthesis are not, by far, the only task that has to be computed in an embedded motion system, such as a mobile robot, a car, a train, a plane, and a submarine. Many other duties have to be fulfilled, just to name a few: intelligent sensor synthesis (feature extraction, detection and recognition), data exchange and synchronization, failure and obstacle monitoring. We embed anytime features in flatness based tracking control scheme, aimed at motion control systems. The error dynamics is approximated through a multiresolution-like process, yielding a sequence of controllers; the more precise an element of this sequence is, the more computational cost it has, the better the stabilization performances are. The techniques in this paper are aimed at automotive applications with practical implementations on an electrical vehicle.