DocumentCode :
2652451
Title :
Identifying a two linked robot with non-symmetrical modified friction and backlash-flexibility models
Author :
Bahrami, J. ; Keshmiri, Me ; Keshmiri, Mo ; Taghirad, H.D.
Author_Institution :
Mech. Eng. Dept., Isfahan Univ. of Technol., Isfahan, Iran
fYear :
2010
fDate :
14-18 Dec. 2010
Firstpage :
1154
Lastpage :
1159
Abstract :
Identifying a robotic system, especially in presence of some nonlinear phenomena, such as friction and backlash, is a troublesome problem. In this paper, initially single link parameter identification is discussed. Parameters such as actuators parameter, link inertia, friction and backlash are identified in this section. A combination of actuator´s model and a model of two flexible masses in attendance of friction and backlash is utilized for this purpose. Identifying two link robot parameters will be under debate afterwards and identification results for the second link are presented in three distinct forms. Furthermore, a method is introduced for selecting best estimated parameters set, in order to be exerted on systems controller. The effects of identified friction torques and backlash modeling on controller improvement are presented. Grey box structure model, available in Matlab´s system identification tool box is utilized in this research.
Keywords :
actuators; friction; multi-robot systems; parameter estimation; robots; torque control; Backlash flexibility model; Grey box structure model; Matlab system identification tool box; actuator parameter; backlash modeling; friction torque; link inertia; link robot parameter; nonsymmetrical modified friction; parameter estimation; single link parameter identification; DC motors; Friction; Mathematical model; Robots; Solid modeling; Solids; Torque;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Robotics and Biomimetics (ROBIO), 2010 IEEE International Conference on
Conference_Location :
Tianjin
Print_ISBN :
978-1-4244-9319-7
Type :
conf
DOI :
10.1109/ROBIO.2010.5723491
Filename :
5723491
Link To Document :
بازگشت