| dc.description.abstract | In industry there is a constant need to make solutions that are more efficient, reduces risk and reduces the impact on the environment. In order to achieve this we see a rapid increase in use of automatic production. Automation of industrial processes typically uses robot manipulators in operations such as drilling, welding, painting, laser cutting or assembly. In the pursuit of improved performance from industrial robot manipulators we see an increasing interest in development of optimal motion on different performance indexes such as velocity, accuracy or energy efficiency. This paper concerns energy efficient scenarios where even minor improvements in performance can lead to substantial impacts on both energy use and profit.
This master thesis includes development of a mathematical model for industrial robot manipulator IRB 1600, by ABB. This includes estimation of masses, centers of mass and inertias. Using a Nikon Metrology camera measurement system the dynamic friction of the robot is characterized, and repeated experiments indicate that friction compensation using this characterization improves performance of the robot in terms of accuracy by 25- 50% for joint 1 thru 3 for different scenarios.
Further, a scheme for development of optimal motion profiles is made, in terms of minimizing mechanical energy consumed by the robot. The motion profiles are made to follow a predefined path, and are derived through numerical optimization. Repeated experiments are performed using an inverse dynamics PD+ controller. By comparing the performance of the robot manipulator to trajectories developed using the standard motion planner (RobotStudio/Rapid), a decrease of about 7% in consumed energy is obtained. | |
