DYNAMIC OBSTACLE AVOIDANCE FOR AN ACKERMAN VEHICLE - A Vector Field Approach

Abstract

A vector field navigation system was shown to avoid dynamic obstacles and reach a goal with a prespecified position and heading using a simulated Ackerman vehicle. The navigation system was divided into two distinct vector fields, an environmental field which was created for goal oriented navigation and obstacle field which was designed for obstacle avoidance. Discussed in this paper were the methods of obstacle avoidance and combining the two fields of the navigation system. The obstacle avoidance method created a rotational vector field centred on a single obstacle. Algorithms were created to select the obstacle that would be the centre of the field and the direction of rotation of the field. A parameter based method was used to combine the obstacle field and the environmental field. A simulation workspace was created to show the navigation behaviours created by combining these methods and a sample of these results were presented in this paper.