Sobresaliente "Cum Laudem"
2015
The proposed method is based on a sensor-based global motion planning paradigm. This is a
planning approach based on a fast sense-model-plan scheme able to integrate sensor information
into a simple grid based environment model and to calculate a globally consistent, smooth and
safe trajectory fast enough for use as a reactive navigation method. This approach has certain
advantages: one is the ability of global planning methodologies to guarantee a path between a given
point and the goal point, if a path exists. The others are the smoothness (the term smooth is used
to denote C ¥) and safety of the obtained solution. This solution eliminates the local minima trap
problem and the oscillations in narrow places suffered in other methods. Besides, the methodology
indirectly eliminates the need for a supervision system to detect local minima failures (obstructed
paths) to determine a new, feasible global path from the current position to the goal point. Our
algorithm builds a unique C ¥ potential function V(x) that can be used as a control law.
This enables simplification of the mobile robot architectures, while maintaining good time
response, smooth and safe planned trajectories with continuous curvature. The trajectory generated
by the planner is the fastest possible to reach the goal position, considering the best path according
to the maximum acceptable velocity at each point in the trajectory (path plus velocity).
University:
Carlos III University of Madrid
URL: