ABSTRACT
This paper presents a methodology to navigate an Autonomous Ground Vehicle (AGV) going through a cluttered and unknown environment. The approach proposed here employs a scanning laser rangefinder as a sensor to learn continuously about the surroundings, and a reflexive method is used to plan a collision-free path based on the information given by the sensor. To track the path, a simple difference controller is implemented on the AGV platform. The proposed methodology was tested in a typical scenario. Simulation result shows that the designed system provides an effective solution to the navigation problem.
