A Hybrid Control Framework for Robust Maneuver-based motion planning

Call Number:

NULL

Accession Number:

NULL

Other Number:

NULL

URL:

https://hybrid.soe.ucsc.edu/files/preprints/26.pdf

Keywords:

hybrid systems

Abstract:

We introduce a modeling framework for robustness of maneuver-based motion planning algorithms for nonlinear systems with symmetries. Our framework implements a hybrid controller that robustly combines motion primitives from a pre-defined library. The closed-loop system is viewed as a hybrid system with flows given by a differential equation, jumps given by a difference equation, and two sets where these dynamics are allowed. We show that our hybrid controller for implementation of motion planning algorithms confers to the closed-loop system robustness properties to a large class of perturbations.

NULL