전기적-기계적 특성 기반 로봇 관절용 전기 구동 액츄에이터의 사양 및 사이즈 결정

Abstract

Recently, with the development of artificial intelligence (AI) and information communication technology (ICT) and the growing demand for improving quality of life, robots have been applied to various fields, such as vehicles, services and medical care, as well as industrial application. Therefore, although the demand for high-performance robots and robot parts is increasing, the development of core parts such as the reduction gear and electric motors is mostly dependent on foreign technology. Also, to determine the design specifications of the actuator, the mass of the actuator must be considered, but the actuator is designed and weighted when the design specifications are determined. Due to this conflicting relationship, when determining the design specifications of the actuator, the mass of the actuator depends on the designer's experience or assumptions. Therefore, it leads to the over-sized actuator. Most research on the electric motors for robot joints only suggests a method of selecting a conventional commercial motor to drive a given load. Therefore, this thesis proposes a method to effectively determine the design specification and size of the actuator for robot joints based on the actual movement of the robot. The design objective is to minimize the mass of the actuator, and the process is as follows. First, through dynamic analysis of the robot system, the load torque and speed required for the actuators are calculated when the end-effector moves along the trajectory. Second, the gear candidates (with minimum mass) that can drive the calculated load torque and speed are determined according to the gear ratio. Third, considering the given voltage, current, and temperature limit as well as gear ratio and gear efficiency, an electric motor is designed to drive each joint. At this time, the electrical characteristics of the electric motor are calculated through the analytical method considering magnetic nonlinearity. In addition, the temperature of the electric motor is predicted by performing a thermal analysis reflecting the driving pattern of the electric motor. Also, based on the analysis results, the reduction gear and the electric motor, which are the minimum mass for driving each joint, are determined. Finally, the proposed method is validated by comparing the test with a commercial motor.