BACKSTEPPING-ADAPTIVE CONTROL OF MOBILE MANIPULATORS FOR TRAJECTORY TRACKING

Abstract

A mobile manipulator is a manipulator arm mounted on the mobile platform to offer more flexibilities to reach to the target position in the space. In recent years, research activities in this area have expanded because of the mobility merges with the manipulation. A good amount of research and improvements have been carried out in the field of nonholonomic control of mobile vehicles and motion control of manipulator arms. In literatures, control methods such as state-feedback, output-feedback, dynamic coupling, model-based, adaptive tracking control, etc. have been applied to the control of mobile manipulators.

One of the objectives of this research work is to construct the systematic modeling of kinematics and dynamics of the mobile manipulator using the Lagrangian dynamics under the nonholonomic constraints. Then design a controller for n+m degree of freedom mobile manipulator with the aim of simultaneous control of the velocity of the mobile platform and the motion of the end-effector. Using the idea of kinematic backstepping control and adaptive torque control, a two-step control is presented for the nonholonomic mobile manipulator. The kinematic velocity control is designed in the first step such that all the desired trajectories are achieved. In the second step, the adaptive torque controller based on the dynamics of the mobile manipulator is designed such that the mobile platform velocity and the end-effector position converge to the reference trajectories designed in the first step. The parameters in this case are assumed to be completely unknown. The parameter update law is formed and used along with the designed controller to update the parameters. Those parameters are used as the estimates of the real parameters and a control signal is produced. This control scheme provides an efficient solution to the motion control problem. System dynamics is modeled in the MATLAB/Simulink and an appropriate controller is modelled for the specified task. Simulink results validate that the designed control method guarantees that the mobile manipulator states converge to the desired trajectories.