Autonomous Indoor Vehicle Navigation Using Modified Steering Velocity Obstacles

摘要

Wheeled mobile robot as an autonomous vehicle that operate in indoor environment must be equipped with the ability to navigate from the initial position to the target. In the workspace with narrow passage such as office, capability to plan and track the path have to be completed with precise reaction to avoid other objects. This paper presents the development of navigation approach of an autonomous indoor vehicle in office environment. For the global path planning, Probabilistic Road Map is used to generate waypoints. Pure Pursuit is devised to track these points. For the local path planning, we propose to modify steering property of Velocity Obstacles based on distances read by LIDAR sensor. For evaluating our approach, the proposed indoor navigation is implemented in simulations for a two-wheeled differential-steering mobile robot. The results shows that our approach was able to avoid static obstacles when navigating the robot to cover a distance of 13.04 meters from the initial position to reach the target position in an environment with a limited area within 3.65 minutes using an average angular velocity of 0.02 rad/s.