RBF-neural network adaptive control of mobile manipulator

The mobile manipulators have attracted a lot of interest in the last few decades. The accurate and reliable control of them seems to be an essential and important requirement for many communication, sensing and control applications. However, some key issues exist that restrict the applications: the presence of external disturbances as well as the model dynamic uncertainties greatly increase the difficulty when we perform the designing of a controller for the mobile manipulator. Aimed at easy the suffering in designing the controller in practice and improve the robot performance, in this paper, an intelligent robust controller based on neural network is proposed for the coordinated control of a mobile manipulator. This method does not require an accurate model of the robot. The unknown dynamic parameters of the mobile platform and the manipulator are identified and compensated in closed-loop control using RBF (Radial Basis Function) neural network. The output errors due to the disturbances can be completely eliminated by this method. The weighting matrices, centers and widths of the RBF structure in the proposed method can be updated on-line. Simulations are presented to show the effectiveness of the presented method.