Optimal design, modeling and control of a long stroke 3-PRR compliant parallel manipulator with variable thickness flexure pivots. (December 2019)

• Record Type: Journal Article
• Authors: Miao, Yang; Zhijiang, Du; Lining, Sun; Wei, Dong
• Is Part Of: Robotics and computer-integrated manufacturing. Volume 60(2019)
• Page Start: 23
• Page End: 33
• Publication Date: 2019-12
• Subjects: Flexure pivot -- Compliant parallel manipulator -- Inverse kinematics -- Trajectory tracking control; Robots, Industrial -- Periodicals; Computer integrated manufacturing systems -- Periodicals; Robotics -- Periodicals; Robots industriels -- Périodiques; Productique -- Périodiques; Robotique -- Périodiques; 670.285
• DOI: 10.1016/j.rcim.2019.05.014
• Languages: English
• ISSNs: 0736-5845

Abstract

Highlights: A long stroke 3-PRR compliant parallel manipulator with variable thickness flexure pivots is proposed. The geometric parameters of the flexure pivot and the manipulator are optimized to take comprehensive consideration of their performance. An accurate inverse kinematic model which consider the parasitic rotational center shift of the flexure pivots is established. A close loop control scheme which combines an on-line learning RBFNN and a disturbance observer is designed. The 3-PRR compliant parallel manipulator can achieve micron scale translational tracking accuracy and micro-degree rotational tracking accuracy. Abstract: The variable thickness flexure pivot (VTFP) is a promising flexure hinge to construct long stroke compliant mechanisms with high precision, since it combines both the advantages of the classical flexure pivot and the notched flexure hinge. In this paper, a 3-PRR compliant parallel manipulator (CPM) is proposed by employing the VTFPs to serve as the passive rotational joints. Geometric parameters of the VTFP and the 3-PRR manipulator are optimized by genetic algorithm to obtain the desired motion performance of the CPM. A prototype 3-PRR CPM is fabricated using the optimized results. In order to consider the parasitic rotational center shift of the VTFPs, an accurate inverse kinematic model (AIKM) is established, numerical results show the superiority of the AIKM compared with the rigid inverse kinematic model (RIKM). Moreover, an on-line learning radical basis function neural network (RBFNN) is established to compensate the unmodeled factors of the system, and a disturbance observer (DOB) is designed by utilizing the proposed AIKM and the RBFNN compensator. The observed external disturbances of the system is compensated via a feed-forward compensation strategy. Experiments show that the 3-PRR CPM can achieve micron scale trajectory tracking accuracy over centimeter's motion range and micro-degree rotational tracking accuracy by the proposed control scheme.