NURBS-based trajectory design for motion planning in a multi-obstacle environment

This work presents a novel solution for the satisfaction of continuous-time constraints through the use of NURBS curves. Exploiting the NURBS functions properties (local support, convexity and positivity) and applying knot refinement strategies, we improve on the sufficient conditions which guarantee obstacle avoidance and the validation of velocity bounds. We show that feasibility and performance are significantly improved with only a moderate (and controllable) increase in complexity. We highlight the approach for off-line trajectory design with obstacle(s) avoidance.