Self-Tuning Fuzzy Task Space Controller for Puma 560 Robot
- PDF / 1,844,944 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 76 Downloads / 221 Views
ORIGINAL ARTICLE
Self‑Tuning Fuzzy Task Space Controller for Puma 560 Robot Azita Azarfar1 · Babak Azarfar2 · Mojtaba Vahedi1 Received: 8 April 2019 / Revised: 4 December 2019 / Accepted: 5 October 2020 © The Korean Institute of Electrical Engineers 2020
Abstract Since in most robot applications the desired paths are determined in task space or Cartesian space, it is important to control the robot arm in task space. In this paper, a fuzzy controller with modifiable scaling factors is proposed to control the robot end-effector in task space. The controller is a fuzzy system with a mechanism to change the scaling factors when the error is bounded under a predetermined value. The controller is designed in joint space and is developed to work space by using inverse Jacobian strategy. The simulations results on Puma 560 robot manipulator illustrate the high performance of the presented control method. Keywords Final robot task space · Self-tuning fuzzy system · Puma 560 · Inverse Jacobian method · Robot tracking
1 Introduction In most robot applications, a desired trajectory for robot end effecter is designed in task space or Cartesian space. However, the joint coordinates contain the control actions. To design a joint space controller for task objectives, it is necessary to carry out the inverse kinematics transformation to obtain the joint space desired path. To avoid solving the inverse kinematic problem, some researchers used inverse or transpose Jacobian method to control a robot end-effector to a set point or to a trajectory in task space [1–3]. In Ref. [4], an adaptive Jacobian controller is proposed with uncertainty in kinematics. Several research works on approximate Jacobian matrix in robot control have also been presented [5–7]. It should be noticed that an inexact Jacobian matrix results in imprecise transformation from task-space to joint-space, and it creates some difficulties in control design.
* Azita Azarfar [email protected] Babak Azarfar [email protected] Mojtaba Vahedi [email protected] 1
Department of Electrical and Computer Engineering, Shahrood Branch, Islamic Azad University, Shahrood, Iran
Department of Mining and Metallurgical Engineering, University of Nevada, Reno, Reno, NV, USA
2
The performance of control strategies has been improved by using intelligent control systems [8, 9]. They have the ability to overcome the limits of classical and adaptive methods. The model dependency is usually decreased by using these methods. Some of intelligent methods such as fuzzy systems have in-built adaption and decision making capabilities to overcome the uncertainty [10]. As robotics problems are nonlinear, many research works developed intelligent methods for robot control. For example, Ref. [11] used the indirect adaptive fuzzy method to control the contact force. Fuzzy sliding mode control is used to control an unknown robot [12]. The authors in Refs. [13, 14] proposed a sliding mode control system in which the coefficients of sliding surface are optimized by optimiz
Data Loading...
