Showing 1–1 of 1 results for author: Ekenberg, K

Distributionally Robust RRT with Risk Allocation

Authors: Kajsa Ekenberg, Venkatraman Renganathan, Björn Olofsson

Abstract: An integration of distributionally robust risk allocation into sampling-based motion planning algorithms for robots operating in uncertain environments is proposed. We perform non-uniform risk allocation by decomposing the distributionally robust joint risk constraints defined over the entire planning horizon into individual risk constraints given the total risk budget. Specifically, the determini… ▽ More An integration of distributionally robust risk allocation into sampling-based motion planning algorithms for robots operating in uncertain environments is proposed. We perform non-uniform risk allocation by decomposing the distributionally robust joint risk constraints defined over the entire planning horizon into individual risk constraints given the total risk budget. Specifically, the deterministic tightening defined using the individual risk constraints is leveraged to define our proposed exact risk allocation procedure. Our idea of embedding the risk allocation technique into sampling based motion planning algorithms realises guaranteed conservative, yet increasingly more risk feasible trajectories for efficient state space exploration. △ Less

Submitted 13 May, 2023; v1 submitted 17 September, 2022; originally announced September 2022.