Browsing by Author "Tan-Phat Phan"

Now showing 1 - 2 of 2

A Novel 6-DOF Force/Torque Sensor for COBOTs and its Calibration Method
(2018-01-01)
Tan-Phat Phan
;
Chao, Paul C. -P.
;
Cai, Jia-Jun
;
Wang, Yu-Jen
;
Wang, Sheng-Ching
This paper intends to design a 6-DOF force/torque sensor along with its readout circuit for collaborative robots (Cobots) working environment. There are two major technologies to develop by this paper. The first technology is the capability of sensing contact forces and torques by a newly-designed six degrees-of-freedom (DOF) force/torque sensor at a fast speed to facilitate feedback control towards anti-collision for the Cobots with wide measurement range, precision, with a smaller size and lower costs. For the force/torque sensor structure, the finite element method (FEM) is established. The sensor is designed and optimized based on finite element analysis (FEA) while " maximum likelihood estimation (MLE)" is utilized to increase accuracy in estimation of force/torque by using appropriate statistical model. The second technology is the anti-collision control. If the collision occurs, the Cobot has an ability to response safely and instantaneously. The proposed control scheme and assisting 6-DOF sensor are capable of performing Cobots anti-collision in welding, grinding, polishing, assembling, etc. with high precision.
A SIX-DOF FORCE/TORQUE SENSOR FOR COLLABORATIVE ROBOT AND ITS CALIBRATION METHOD
(2017-01-01)
Trong-Hieu Tran
;
Tan-Phat Phan
;
Chao, Paul C. -P
;
Wang, Yu-Jen
;
Wang, Chun-Chieh
This study presents a new six-DOF force/torque sensor and its calibration method for a "collaborative robot or cobot". This new calibration method applies the so-called "maximum likelihood estimation (MLE)". MLE is used to determine and identify the parameters related to applied torques/forces and resulted deformations at varied locations of the sensor structure. Formulating the relations in a vector-matrix form, such parameters are captured as the coefficients in a matrix relating torques/forces to angular/linear deformations in different directions. In addition to applying MLE, finite element (FEM) modeling and analysis are conducted to generate realistic-like empirical data for the aforementioned calibration computation based on MLE. The matrix formed by these coefficients can then predict forces and torques with good accuracy via deformations detected by strain gauges. Seen in simulation results, the worst sum of error for three forces is less than 0.04%, while that for three torques is less than 0.00005%. Very small values of sum of error demonstrate a strong correlation of the model with the results obtained from MLE, indicating high precision in sensing forces and torques without decoupling matrix. The developed sensor is highly suitable for real-time sensing of multidimensional interactive forces/torques in industrial robots.