Detail publikace

Obstacle avoidance for non-point mobile robots

Originální název

Obstacle avoidance for non-point mobile robots

Anglický název

Obstacle avoidance for non-point mobile robots

Jazyk

en

Originální abstrakt

The obstacle avoidance algorithm described in this paper is based on the generalized potential fields, which depend on robot's velocity as well as its position. Unlike classical potential fields, the robot is not repulsed by an obstacle if it is not moving toward it. The key problem when using generalized potentials is the computation of the time it would take to bring the robot moving with constant velocity to collision with the nearest obstacle in the given direction. It is easy for mobile robots with circular footprint, when orientation does not play a role. This paper presents new technique for computation of time remaining to the collision based on linear programming, which can be used in the case when robot's shape can not be approximated by a circle and rotational degree of freedom must be taken into account.

Anglický abstrakt

The obstacle avoidance algorithm described in this paper is based on the generalized potential fields, which depend on robot's velocity as well as its position. Unlike classical potential fields, the robot is not repulsed by an obstacle if it is not moving toward it. The key problem when using generalized potentials is the computation of the time it would take to bring the robot moving with constant velocity to collision with the nearest obstacle in the given direction. It is easy for mobile robots with circular footprint, when orientation does not play a role. This paper presents new technique for computation of time remaining to the collision based on linear programming, which can be used in the case when robot's shape can not be approximated by a circle and rotational degree of freedom must be taken into account.

BibTex


@inproceedings{BUT6908,
  author="Bohumil {Honzík}",
  title="Obstacle avoidance for non-point mobile robots",
  annote="The obstacle avoidance algorithm described in this paper is based on the generalized potential fields, which depend on robot's velocity as well as its position. Unlike classical potential fields, the robot is not repulsed by an obstacle if it is not moving toward it. The key problem when using generalized potentials is the computation of the time it would take to bring the robot moving with constant velocity to collision with the nearest obstacle in the given direction. It is easy for mobile robots with circular footprint, when orientation does not play a role. This paper presents new technique for computation of time remaining to the collision based on linear programming, which can be used in the case when robot's shape can not be approximated by a circle and rotational degree of freedom must be taken into account.",
  address="ARGESIM",
  booktitle="Proceedings of the IMACS Symposium on Mathematical Modelling",
  chapter="6908",
  institution="ARGESIM",
  year="2000",
  month="january",
  pages="887",
  publisher="ARGESIM",
  type="conference paper"
}