Motion control of a mobile robot with an onboard manipulator

Navigation and control of autonomous vehicles with onboard manipulator systems are currently being investigated for intelligent manufacturing applications. A multilayer neural network controller which feedback linearizes a mobile base with an onboard arm is presented. Feedback linearization using neural networks is used to take into account the complete dynamics with non-holonomic constraints. The neural network feedback linearization provides an inner loop that accounts for possible motion of the onboard arm. The neural network considered in this study does not require any a priori training but rather exhibits a learning while functioning feature. The weights of the network are dynamically updated on line. The cases of maintaining a desired course and speed and the case of disturbance rejection due to the interactive forces between the mobile base and the onboard arm are considered