A Test-Bed for Visual Servo Control of Artificial Muscle Micro-Robot with Parallel Architecture

Abstract

Artificial muscles, or soft smart materials, are being increasingly used to build micro-manipulators. Electroactive polymer (EAP) actuators, including electronic EAP and ionic EAP, are attracting interest from research community because of their potential to achieve high strains and therefore high displacements. However, their application is limited by their ability to generate large forces. A parallel architecture, i.e. handling objects using multiple polymer actuators, can greatly enhance their load carrying ability. A significant technical challenge presented by micro-manipulators based on artificial muscle actuators is the nonlinearity of actuator dynamics. Currently, there is no satisfactory model that can used in the control design. Visual servo (VS) control of the displacements of a robot in closed-loop using the data provided by one or multiple cameras is presented as a possible approach to this problem. A parallel architecture micro-gripper is designed as a prototype of a micro-gripper. A vision system is developed to identify dynamic behaviors of IPMC arms and design a control system for the parallel micro-gripper. Preliminary identification and control results are given. Visual servoing motion control is introduced