An experimental study on the vision-based control and identification of planar cable-driven parallel robots

详细信息    查看全文

• 作者：Hassan Bayani ^{Hassan_bayani@ut.ac.ir" class="auth_mail" title="E-mail the corresponding author}Author Vitae ; Mehdi Tale Masouleh ; ^{m.t.masouleh@ut.ac.ir" class="auth_mail" title="E-mail the corresponding author}Author Vitae ; Ahmad Kalhor ^{akalhor@ut.ac.ir" class="auth_mail" title="E-mail the corresponding author}Author Vitae
• 关键词：Cable-driven parallel robots ; Pole placement controller ; Adaptive sliding mode controller ; Vision ; Control and system identification
• 刊名：Robotics and Autonomous Systems
• 出版年：2016
• 出版时间：January 2016
• 年：2016
• 卷：75
• 期：part_PB
• 页码：187-202
• 全文大小：1957 K

文摘

This paper presents the implementation of a vision-based position control for planar cable-driven parallel robots. The main contribution of this paper contains three objectives. First, a method is used toward kinematic modeling of the robot using four-bar linkage kinematic concept, which could be used in online control approaches for real-time purposes. In order to ensure positive tension in cables, as a basic essential property, static and dynamic equations of robot are also obtained. Second, in order to track the position of End-Effector, an online image processing procedure is developed and implemented. Finally, as the third contribution, different control approaches are applied in order to validate the model with plant and obtain the most promising controller. As classic controller, pole placement approach is suggested and results reveal weaknesses in modeling the uncertainties. Due to the latter incapability, sliding mode controller is utilized and experimental tests represent effectiveness of this method. However, the chattering phenomenon in the beginning of the robot operation is the main insufficiency of this controller. Hence, in order to present a more accurate controller, adaptive sliding mode controller working alongside with an identification method on the model is applied. The provided identification procedure is based on Recursive Least Square approach, which rebates the effects of uncertainties in the parameters of the model. Moreover, results of these controllers confirm accommodation between the model and robot. The proposed procedure could be well applied for any kind of planar cable-driven parallel robot.