The mobile manipulator’s navigation based on fuzzy control and netual network

With the development of robot to intelligent direction, the research of robotic control has become the focus. In this paper, a navigation system of mobile manipulator is designed, fuzzy control theory and artificial neural network are applied. And study basic arithmetic of image processing, use various image pre-treatment and segmentation technology. The main research result of the paper is as follows: (1) The vision system of mobile manipulator is designed. Using the color information to recognize the road simplifies the character picking-up of image and improves the robust of road recognition. (2) The mobile manipulatorpsilas road following system is designed, It design two inputs and single outputpsilas fuzzy controller according to the movement characteristics of the mobile manipulator, the error and error change are regarded as inputs, the voltage difference of between left and right wheels is regarded as output. Using the fuzzy control strategy avoids the complicated process of building model in the traditional control. The systems of road following can be used in the high nonlinear and uncertain environment. (3) Man-made guiding road and road sign are laid on the ground, the number road sign is identified by BP neural network. In this paper, the factors, which influence the capability of road following system, are analyzed. At the same time, the lag of vision system is revised. The strategy of road following has already been successfully used in the HEBUT-2 intelligent mobile manipulator.