Independent thesis Advanced level (professional degree), 20 credits / 30 HE credits
The aim of this project is to add manipulation functionality to one of the NASA's Mars rovers through the integration of a robotic arm. The functionality is needed to enable retrieval of deployed communication bricks from the Martian surface. For this, a simple hook end-eector is designed and tested using a controller implemented on a computer.
The manipulator is a ve degree of freedom robotic arm with revolute joints. It is a nonlinear MIMO system but can be approximated to be five highly decoupled linear SISO systems. The manipulator is controlled in joint space, where each joint follows a trajectory. The trajectory is generated as four sequential sub-trajectories, where the manipulator comes to rest at each transition point.
We assume that the Martian surface is free of obstacles and that the manipulator can work in any position on the workspace. We also assume that the position of the hole on the communication brick is known. For controlling the end-eector in joint space, a trajectory-following PD-controller is constructed.
When testing the controller the rst four joints (waist, shoulder, elbow and twist) follow the desired trajectory well. The last joint, wrist, follows the trajectory well for the rst period of time and then start to oscillate around the path curve. The controller was tuned by decreasing Kp value for the wrist joint. The reason for deciding to decrease Kp was to slow the system down and therefore gain an increased stability. Tuning resulted in a better behaving wrist joint controller which follows the desired trajectory in a satisfactory way. Oscillations in wrist have disappeared and the controller works as intended.
We designed and manufactured an end-eector that was within the IRG price range. We established baseline communication with the manipulator, developed a standalone control software and implemented a software for collection of communication bricks. This controller was working properly and fullled the goal set at the beginning of the thesis to design a communication brick collecting manipulator.
2008. , 35 p.