Energy-efficient industrial robot design by redundant actuation

Abstract

A method for improving the energy efficiency of articulated robots with redundant actuation is presented in this work. A lot of energy is dissipated as electrical energy loss when consumed electrical energy is converted to the mechanical energy of an actuator that operates the robot. In addition, mechanical energy loss can occur in the process of using the mechanical energy of the actuator during operation of the robot. The purpose of this work is to improve the energy efficiency of the robot by reducing these two types of energy loss. When operating a general robot, electrical energy loss from the actuators required current and mechanical energy loss from the mechanical requirements of actuators occurs. These two types of energy loss can be reduced by extension to a redundantly-actuated mechanism that uses more actuators than the number of the structures Degrees of Freedom (DOF). The redundantly-actuated robot can reduce the electrical energy loss by distributing the required currents to multiple actuators. Moreover the mechanical energy loss can also be reduced by rearranging each actuators mechanical power. A 2-DOF test manipulator is manufactured with reference to a commercial manipulator for the verification of energy saving theory using redundant actuation. An automobile welding task trajectory is used to operate the manipulator. By the operation of this trajectory, energy saving effects in accordance with these two types of operation algorithms are verified. The first operation algorithm is minimum-norm torque distribution algorithm. This algorithms purpose is to minimize the sum of the square of each torque, commonly used for operating the redundantly-actuated robot. This algorithm can be realized by manipulating the synchronous position control of each actuator. The second operation algorithm is a minimum-energy consumption algorithm. This algorithms purpose is to maximize the energy saving effect by analyzing a consumed electrical power model. The energy saving effects are verified as 41.4% and 45.0%, according to each operating algorithm.