Byeonghun Na

Design of a Direct-Driven Linear Actuator for a High-Speed Quadruped Robot, Cheetaroid-I

Several legged robots have been developed in recent years and proved that they are an effective transportation system on an uneven terrain. In particular, quadruped robots are regarded as a new trend in robotics due to their superior gait stability and robustness to disturbances. More recently, many robotics researchers are making their best efforts to improve the locomotion speed, as well as the stability and robustness, of quadruped robots. The high-speed locomotion creates various challenges in the development of actuators, mechanical design, and control algorithms of the robot. In this paper, a linear actuation system for the high-speed locomotion of a quadruped robot is introduced. The proposed actuator is designed based on the principle of brushed direct-current electric motor systems. For the minimal impedance and improved power capacity, the actuator is designed with dual layers of cores, which are aligned parallel to permanent magnets. The mechanical and electrical properties of the actuation system, such as back-drivability, controllability, and response time, are verified by experimental results. A robotic leg, which is the rear leg of a Cheetah-like robot, is designed with the proposed actuator, and experimental results for trajectory tracking performance are presented.

Design of a direct-driven linear actuator for development of a cheetaroid robot

Quadruped robots are regarded as a new trend in robotics due to their superior gait stability and robustness to disturbances. More recently, many robotics researchers are making their best efforts to improve the locomotion speed, as well as the stability and robustness, of quadruped robots. The high-speed locomotion creates various challenges in the development of actuators, mechanical design, and control algorithms of the robot. In this paper, a linear actuation system for the high-speed locomotion of a quadruped robot is introduced. The proposed actuator is designed based on the principle of brushed direct-current electric motor systems. For the minimal impedance and improved force capacity, the actuator is designed with dual layers of cores, which are aligned parallel to permanent magnets. The mechanical and electrical properties of the actuation system, such as back-drivability, controllability, and response time, are verified by experimental results. A robotic leg, which is the rear leg of a cheetah-like robot, is designed with the proposed actuator, and is introduced briefly in this paper also.

Control Power Reduction and Frequency Bandwidth Enlargement of Robotic Legs by Nonlinear Resonance

Legged robots are regarded as a new means for transportation at uneven terrains and dangerous situations. In particular, quadruped robots that are capable of high-speed running are receiving great attention due to their superior mobility. The legs of such high-speed running robots rapidly repeat the swing and stance motions. Therefore, the legs of high-speed running robots should exhibit low impedance and friction for fast swing motions, while it is required to produce a significantly large actuation force for propulsion of the robot body in a stance phase. For this purpose, a direct-driven actuation mechanism was proposed for the Cheetaroid robot in our previous work. In this paper, in order to further enlarge the frequency bandwidth of the leg module and to reduce the required control (i.e., actuation) power, a dual-stage spring is designed to introduce a motion-adaptive resonance into the system. The structure and parameters of the dual-stage spring are determined in an optimal manner to minimize the control power. The resonance due to the dual-stage spring occurs only during high-speed locomotion, since the overall mechanism is designed such that the spring force is applied to the leg for high-speed locomotion only. The proposed system is verified by simulation studies and experimental results in this paper.

WalkON Suit: A Medalist in the Powered Exoskeleton Race of Cybathlon 2016

The powered exoskeleton race in Cybathlon 2016 consisted of six challenging tasks that required a pilot with complete paraplegia to walk on a level floor, uphill, downhill, and on stairs; stand up and sit down; step on stones; and even pass through a tilted path. All of these tasks addressed exactly the requirements for a powered exoskeleton designed to assist with activities of daily living (ADL) for paraplegics. Every team brought unique technologies to achieve this goal.

Design of a compact rotary series elastic actuator for improved actuation transparency and mechanical safety

Actuators for human-interactive robot systems require transparency and guaranteed safety. An actuation system is called transparent, when it is able to generate an actuation force as desired without any actuator dynamics. The requirements for the transparent actuation include high precision and large frequency bandwidth in actuation force generation, zero mechanical impedance, and so on. In this paper, a compact rotary series elastic actuator (cRSEA) is designed considering the actuation transparency and the mechanical safety; the mechanical parameters of a cRSEA are optimally selected for the controllability, the input and output torque transmissibility, and the mechanical impedance. A mechanical clutch that automatically disengages the transmission is devised such that the human is mechanically protected from an excessive actuation torque due to any possible controller malfunction or any external impact from a collision. The proposed cRSEA with a mechanical clutch is applied to develop a wearable robot for incomplete paraplegic patients. Experimental results of a manufactured cRSEA system are introduced in this paper also.

Joint trajectory generation and motion control of a wearable robot for complete paraplegics based on forward inflection walking

Wearable robots are regarded as a new transportation system in daily living for complete paraplegics due to spinal cord injury (SCI). For the motion control of a wearable robot, the normal gait pattern of people without disabilities is often applied as a reference input. When the natural dynamics of the human body can be utilized and the displacement of center of gravity (CoG) is possible, the normal gait pattern is effective to increase the gait speed with the minimal energy consumption. Complete paraplegics with high SCI level (i.e., above thoracic spine level), however, are able to voluntarily move neither the legs nor their waist, and thus they cannot control the CoG at all. Consequently, the normal gait pattern is not necessarily the best option for such complete paraplegics. In addition, the degree of freedom (DoF) of wearable robots is less than that of the human body; therefore, it is difficult to expect that the complete paraplegics naturally control the CoG even with the help of wearable robots. In this paper, a new gait pattern, called forward inflection walking (FIW), is proposed for the motion control of a wearable robot for complete paraplegics. The proposed FIW method enables to transfer the CoG to the leg ahead by modification and optimization of the joint angle trajectories for effective walking. Therefore, the FIW method intentionally moves the CoG of complete paraplegics with wearable robots according to the gait phase (i.e., stance and swing). The proposed method is verified by clinical experiments.

Automatic tuning of a mechanical design parameter of a robotic leg by Iterative Learning Mechatronics

The best control performance in mechatronic systems cannot be obtained by only control algorithms, because the control is effective only within the performance range realizable by an actuator system in practice. To obtain the best control performance, therefore, the mechanical system should be designed considering the control performance first. It is, however, difficult to expect the control performance in the mechanical design process, since the control performance is dependent on various unexpectable factors, such as input and output saturations of an actuator, heat problems, sensor limitations and so on, as well as the mechanical design parameters. Therefore, in this paper a recursive mechanical design process based on control experiments, called Iterative Learning Mechatronics (ILM), is proposed. The proposed method seeks a better mechanical design parameter based on a set of control signals, such as a control input and a tracking error, and suggests an update of the design parameter. For verification of the proposed method, the ILM is applied to obtain the best leg stiffness of a robotic leg bearing a load.

Frequency bandwidth enlargement of robotic legs by dual-stage springs

The legs of high-speed running robots rapidly repeat the swing and stance motions. Therefore, the legs of high-speed running robots should exhibit low impedance and friction for fast swing motions, while it is required to produce a significantly large actuation force for propulsion of the robot body in a stance phase. For this purpose, a direct-driven actuation mechanism was proposed for the Cheetaroid robot in our previous work. In this paper, in order to further enlarge the frequency bandwidth of the leg module and to reduce the required control (i.e., actuation) power, a dual-stage spring is designed to introduce an motion-adaptive resonance into the system. The resonance due to the dual-stage spring occurs only during high-speed locomotion, since the overall mechanism is designed such that the spring force is applied to the leg for high-speed locomotion only. The proposed system is verified by simulation studies and experimental results in this paper.

A disturbance observer for robust position tracking control and ground contact detection of a Cheetaroid-I leg

Quadruped robot systems are being intensively investigated by as a new means of transportation systems with multi-purposes. As the operation conditions of such systems, such as a payload, gait speed, and so on, are demanding, the control of the robotic legs is challenging. In addition, the dynamics of such systems is subjected to large disturbances and parameter variations due to the ground contact and the complicated mechanical structure. In this paper, a robust control algorithm based on a disturbance observer (DOB) is proposed for the control of legs of a quadruped robot, Cheetaroid-I. The DOB enables rejecting disturbances caused by the ground contact, payload, etc, and detecting the ground contact without force sensors. The proposed method is verified by experimental results.

Robust Position Tracking Control and Ground Contact Detection of a Cheetaroid-I Leg by a Disturbance Observer

Abstract Quadruped robot systems are being intensively investigated as a new means of transportation systems with multi-purposes. As the operation conditions of such systems, such as a payload and gait speed, are demanding, the control of the robotic legs is challenging. In addition, the dynamics of such systems is subjected to large disturbances and parameter variations due to the ground contact and the complicated mechanical structure. In this paper, a robust control algorithm based on a disturbance observer (DOB) is proposed for the control of legs of a quadruped robot, Cheetaroid-I. The DOB enables rejecting disturbances caused by the ground contact, payload, etc, and detecting the ground contact without force sensors. The proposed method is verified by experimental results.