Design a Rate-Hysteresis Reduction Task-space Control on a Stewart Robotic Platform

Abstract

A task-space control scheme is established by a derived dynamic model on the end effector associated with a few sensors employed on the specific location mounted on top of the piezo-driven 6DOF Stewart robotic platform. As piezoelectric actuators are used on the legs to manipulate the platform following an assigned trajectory, various manipulating speeds in operating frequencies may appear, making it more challenging to deal with the rate-dependent hysteresis effect. Despite the development of an Inverse Preisach model, building a large amount of information to identify the hysteresis loop at the different operating frequencies is complex and time-consuming. Instead, a run-to-run (RtR) adaptation procedure. The use of an internal model control incorporated with the exponentially weighted moving average (EWMA) method is proposed with a discrete-time formation that automatically adjusts the variant relation between the input driving voltage and the output displacement of the leg actuators. In an experiment, free-space manipulation and machining of various work parts with different stiffness are carried out. The proposed feedforward controller, which consists of the Inverse Preisach model and the RtR_EWMA, is demonstrated to be effective by the findings. The task-space computed torque control approach works noticeably better when combined with the feedforward controller than when used alone.

DOI: 10.61416/ceai.v25i4.8636