| Fully autonomous vehicles have been a dream for control engineers and robotics experts since the development of early computers. A full gambit of agencies from the United States Department of Defense and NASA to academic institutions such as Carnegie Mellon and University of California at Berkeley, have been looking into the issue of creating stable autonomous vehicles, many of which have been successful to varied degrees. Unfortunately, all successful attempts have been incredibly cost prohibitive and thus, impractical to implement in public vehicles. Therefore a fresh perspective of this system needs to be taken into account. Autonomous vehicles have been proven to be possible but these systems are cumbersome and expensive to implement making them an impractical choice. Many more practical systems are not sufficiently stable. A compromise needs to be made between these two systems to create a practical yet stable system that can be readily implemented in modern vehicles. The purpose of this endeavor is to create a cost effective, reliable control system capable of maintaining a vehicle along a path. The controller created will rely solely on an image processing algorithm with a single reference point. This image processing algorithm will be created in LabVIEW to acquire or calculate any and all information needed to stabilize a vehicle within a given lane and analyzed in MATLAB. While initial attempts at such a control method are promising, improvements to the image preprocessing need addressed for future prototypes to be made. |
