Application of intelligent computing to autonomous vehicle control

This paper utilizes a real-valued genetic algorithm (RGA) and fuzzy system to an autonomous vehicle control problem. Obstacle avoidance and parking control are performed by the use of a CCD camera, sonar sensor, and localization system on a wheeled mobile robot (WMR). The camera provides image of the surrounding. Distance between the WMR and object or obstacle can be obtained by image process and distance computation algorithm. Fuzzy system and genetic algorithm are integrated in the control scheme that can effectively drive the WMR without complicated mathematical equations. Sonar sensors detect safety distance for the WMR. The localization system provides coordinate position of the WMR. The WMR can perform obstacle avoidance and garage parking successfully.