Vision Robot Moving Control by Supervisory Fuzzy Neural Network

Abstract

Supervisory fuzzy neural network has been developed for the vision robot control. This fuzzy neural network (FNN) is constructed by method of adaptive learning algorithm. The FNN control is combined with supervisory controller (SC) to be a full controller called supervisory fuzzy neural network controller (SFNN). The SFNN is tuned by off‐line method and designed by considering image processing such as to be nearly identical to the perfect controller. The error of system is come from captured image data, so the controlled results should be drive into a suitable switching plane such that the stability can be achieved. This SFNN can achieve good control performance of vision robot. Simulation results show the proposed method possesses satisfied performances for the vision robot.