Application of Intelligent Control to the 2007 FIRST Robotics Competition

FIRST (For Inspiration and Recognition of Science and Technology) is a non-profit organization that organizes several robotics competitions for pre-college students. The FIRST robotics competition (FRC) is a high school level competition wherein high school students work with engineers to design and control a device to compete in the competition. The 2007 FRC presented a control problem centered on collision control and target tracking. Our strategy for the competition required the automated acquisition, tracking, controlled contact, and pushing of the opponent robot during the autonomous period. For acquisition of the target, data gathered from ultrasonic range sensors provides information related to the location of target robots. Following processing of the raw input data, and artificial neural network decides which target to pursue. Once this target is acquired, the contact is controlled using a fuzzy controller. Using range to target and closing speed as inputs, a fuzzy controller determines motor drive. This allows intuitive adjustment of the controller, varying the number, center, and shape of membership functions. The fuzzy collision controller has been preliminarily tested. The key to this control is to approach the target with sufficient speed to close the distance quickly, make contact with slow closing speed to avoid ramming, and then maximize the motor drive to push the opponent. Using Gaussian membership functions allows smooth transition between the classes while allowing easy modification of the controller. This method has given the appropriate outputs in controlled calibration and experimentation.