Neuro-Fuzzy Controller Using LabVIEW

P. Ponce, R. Fernandez-del-Busto, F. Ramirez, and D. Méndez (Mexico)


Neuro-fuzzy controller, mobile robots, fuzzy logic, fuzzy clustering, trigonometric neural networks, LabVIEW.


Mobile robots have become a fundamental area of Robotics research due to the huge variety of applications in modern life; artificial intelligence (AI) techniques have proven very effective when dealing with high complexity tasks, as in Robotics. This paper describes such application, using a neuro-fuzzy controller using LabVIEW which allows a quadruped robot to walk in an unknown environment and learn the correct path. The controller is based in a c-means cluster algorithm and on-line adjustments of the fuzzy membership functions via a trigonometric neural network. Prediction schemes are used to improve the system performance, moving in an unknown environment with static and dynamic obstacles and learning the best trajectory to follow. LabVIEW was the platform on which the controller was designed and implemented. As is well known, LabVIEW is a platform that goes from design to implementation by a minimum number of steps. Although LabVIEW can solve complex problems, it has an intuitive programming environment. Furthermore it includes a graphical user interface.

Important Links:

Go Back