Smooth Trajectory Generation with 4D Space Analysis for Dynamic Obstacle Avoidance

Suhyeon Gim (1, 2) , Lounis Adouane (1) , Sukhan Lee (2) , Jean-Pierre Derutin (3)

1 IP - Institut Pascal
2 SKKU - Sungkyunkwan University [Suwon]
3 LASMEA - Laboratoire des sciences et matériaux pour l'électronique et d'automatique

This paper presents smooth trajectory generation scheme for obstacle avoidance in static and dynamic environment. The smooth trajectory has successive two steps where smooth path is generated firstly and then corresponding velocity is planned along the path. Smooth path of continuous curvature is composed by para-metrically adjusted clothoids with proposed algorithm and then the safe velocity planning is carried out in the 4D configuration framework. Two circles are used to completely surround the used nonholonomic car-like vehicle, this permit to check the probable future vehicle's collisions and to have space-time analysis. Some demonstrative simulations show the strong potential of the proposed smooth and flexible methodology for future experimentations with actual vehicles.

Mots clés: velocity planning; dynamic obstacle avoidance; Continuous curvature path

Domaines: Automatique / Robotique; Automatique; Intelligence artificielle [cs.AI]; Robotique [cs.RO]; Systèmes et contrôle [cs.SY]

Suhyeon Gim, Lounis Adouane, Sukhan Lee, Jean-Pierre Derutin. Smooth Trajectory Generation with 4D Space Analysis for Dynamic Obstacle Avoidance. 11th International Conference on Informatics in Control, Automation and Robotics (ICINCO'14), Special Session on Intelligent Vehicle Controls and Intelligent Transportation Systems - IVC-ITS 2014, Sep 2014, Vienne, Austria. ⟨10.5220/0005148808020809⟩. ⟨hal-01712938⟩