Design of Joint Torque Sensors and Torque Feedback Control for Direct-Drive Arms

Energies ◽  
2020 ◽  
Vol 13 (5) ◽  
pp. 1125 ◽  
Author(s):  
Chenyang Zhang

In order to ensure good dynamic characteristics, servo valve is usually adopted as the drive part of Stewart manipulator which causes huge power consumption, while direct drive electro-hydraulic servo system has the advantages of energy saving, simple structure, convenient installation, and low failure rate. But its dynamic characteristics are so poor that it can only be applied to occasions where quick response is not needed. On the consideration above, following works are done in this paper. Since current coupling exists in the control system based on the speed of the servo motor as the control input, the control system of the direct drive Stewart manipulator is established based on the current of the servo motor as the control input in which the current coupling can be solved. In order to improve the dynamic characteristics of the direct drive Stewart manipulator, a Proportion Differentiation (PD) plus dynamic pressure feedback control strategy is also put forward in this paper, which is verified by using a simulated hydraulically driven Stewart manipulator. Simulation results show that both dynamic coupling and current coupling are solved and the control strategy proposed in this paper can significantly increase the bandwidths of all degrees of freedom.


Author(s):  
Kazushi Sanada

A direct drive volume control (DDVC) is applied to fuel injection control for marine diesel engine. The DDVC consists of an AC servomotor, a fixed-displacement hydraulic pump, and a hydraulic cylinder. The hydraulic cylinder pushes a plunger pump and fuel is pressurized. When the fuel pressure becomes greater than injection pressure, fuel is injected to a combustion chamber. A brief introduction of the DDVC is described first in this paper referring to conventional fuel injection systems including a cam mechanism and a common rail system. A mathematical model of the DDVC for simulation is summarized. Experiments of fuel injection shows the control function of the DDVC fuel injection system. The topic of this paper is feedback control of the quantity of fuel injection (fuel mass per injection) of the DDVC. The feedback control system is simulated using the above mathematical model. Fuel injection is stopped by switching a drive signal of the AC servomotor and retracting a piston of the hydraulic cylinder. The timing to stop injection is adjusted based on crank angle. An algorithm of updating the crank angle to stop injection is proposed so that the quantity of fuel injection follows the target value. Simulation study shows that the update algorithm works successfully.


1987 ◽  
Vol 109 (2) ◽  
pp. 122-127 ◽  
Author(s):  
C. W. deSilva ◽  
T. E. Price ◽  
T. Kanade

This paper describes the development of a joint torque sensor for the second direct-drive manipulator at Carnegie-Mellon University (CMU DD Arm II). The approach taken is to develop the sensor using static design considerations and then test it to verify its dynamic performance. Several design considerations applicable to semiconductor strain-gage torque sensors are presented. These are strain capacity limit, nonlinearity, sensitivity, and stiffness specifications. Associated design equations have been developed in the present work. A numerical example is given to illustrate the use of these design considerations. The development of a circular-shaft torque sensor for the CMU DD Arm II, that employs semiconductor strain gages, is described. Typical results from a static calibration test and from step and impulse tests are presented. Test show that the torque sensor performs well under dynamic conditions in a bandwidth of 100 Hz.


2000 ◽  
Vol 66 (651) ◽  
pp. 3625-3631
Author(s):  
Naoyuki TAKESUE ◽  
Ken'ichi KOYANAGI ◽  
Guoguang ZHANG ◽  
Masamichi SAKAGUCHI ◽  
Junji FURUSHO

2005 ◽  
Vol 277-279 ◽  
pp. 142-147
Author(s):  
Suk Yung Park ◽  
Fay B. Horak ◽  
Arthur D. Kuo

We examined how the central nervous system adjusts postural responses to an increased postural challenge due to an initial lean. Postural feedback responses scale to accommodate biomechanical constraints, such as an allowable ankle joint torque. Initial forward leaning, which is observed among the elderly who are inactive or afraid of falling, brings subjects near to the limit of stability and makes the biomechanical constraints more difficult to obey. We hypothesized that the central nervous system is aware of body dynamics and restrains postural responses when subjects initially lean forward. To test this hypothesis, fast backwards perturbations of various magnitudes were applied to 12 healthy young subjects (3 male, 9 female) aged 20 to 32 years. The subjects were instructed to stand quietly on a hydraulic servo-controlled force platform with their arms crossed over their chests, then to recover from a perturbation by returning to their upright position, without stepping or lifting their heels off the ground, if possible. Initially, the subjects were either standing upright or leaning forward. The force platform was movable in the translational direction and programmed to move backward with various ramp displacements ranging from 1.2 to 15 cm, all with the duration of 275 msec. For each trial, the kinematics and ground reaction force data were recorded, then used to compute the net joint torques, employing a least squares inverse dynamics method. Optimization methods were used to identify a set of equivalent feedback control gains for each trial so that the biomechanical model incorporating this feedback control would reproduce the empirical response. The results showed that the kinematics, joint torque, and feedback gains gradually scaled as a function of the perturbation magnitude before they reached the biomechanical constraint, and the scaling became more severe with an initial forward lean. For example, the model suggested that the magnitude of the ankle joint angle feedback to ankle torque was smaller in the leaning trials than in the initially upright trials, as if the subjects experienced a larger postural perturbation in the leaning trials. These results imply that the central nervous system restrained the postural responses to accommodate the additional biomechanical constraint imposed by the forward posture, thereby suggesting that the central nervous system is aware of body dynamics and biomechanical constraints. The scaling of the postural feedback gains with the perturbation magnitude and initial lean indicates that the postural control can be interpreted as a feedback scheme with scalable gains.


Sign in / Sign up

Export Citation Format

Share Document