A Two-Dimensional Path Planning Algorithm

Path planning is a necessary component of any autonomous vehicle. The task is one of nding a course that takes the vehicle from its current location to a destination avoiding obstacles that might damage or divert the vehicle. There are many forms of path planning algorithms such as those that pregenerate a path versus those that perform reactive run-time planning. This paper presents a two-dimensional pregenerating path planning algorithm that models the path as a line and compares a list of obstacles to this line to determine any collisions. Collisions are processed by selecting a point outside of the obstacle and recursively generating two new paths that avoid the collision by going around the obstacle. This paper will rst describe the STESCA control architecture for autonomous vehicles and then consider path planning. A few other path planning algorithms are described. This is followed by details of the algorithm developed in this research. The paper then presents a number of examples.