Adaptive Control of Mobile Manipulators with Nonlinear Friction Cone and Switching Constraints

A non-minimal order model for mobile manipulation subject to many nonlinear constraints can handle a multi-objective control problem. In this paper, the task space controller is designed based on a reduced dynamic model. We present the the fundamental properties of non-minimal order dynamic model. According to this dynamics the adaptive control of mobile manipulation subject to these constraints is proposed, which could deal with uncertain dynamics. The input saturation and unknown dynamics are solved by adaptive control which used neural network. Moreover, the non-minimal order dynamics does not need to redesign other controllers while switching from one task to another. We evaluate our method by several experiments, the results demonstrate the validity of the robustness and effectiveness.