Autonomous Navigation in Environments with Arbitrary Non-convex Obstacles
Mayur B Sawant, Abdelhamid Tayebi, Ilia G. Polushin · 2022 IEEE 61st Conference on Decision and Control (CDC) · 2022
We develop an autonomous navigation algorithm for a robot operating in two-dimensional environments cluttered with obstacles which can have arbitrary non-convex shapes and can be in close proximity with each other, under the assumption that there exists a safe path connecting the initial and the target locations. We propose a hybrid feedback controller, with Zeno-free switching between the move-to-target mode and the obstacle-avoidance mode, guaranteeing global asymptotic stability of the target equilibrium. To handle the obstacle non-convexity, we introduce a transformation that modifies (virtually) the obstacles’ shapes, in a non-conservative manner, to generate a modified free-space suitable for the design of a reliable obstacle avoidance strategy. Finally, we validate the efficacy of the proposed hybrid feedback controller through simulations.