Mixed-Reality-Based Teleoperation Grasping Control

Abstract

Abstract Traditionally, teleoperation means that the system sends a series of signal commands from the master while the slave manipulator receives and realizes the desired control operations. For the purpose of implementing more dexterous and complex tasks, we propose a novel framework with dual-hand master teleoperation systems under time-varied delays. In this paper, we emphasize studying the bilateral grasping teleoperation control, as the time delay causes a communication outage. Combining a wave-variable structure with a four-channel framework, an event-trigger-based bilateral sliding mode teleoperation control and an adaptive neural network are designed to effectively achieve master-slave trajectory tracking. In the virtual 3D environment, we created a mixed-reality interface based on dual-hand master teleoperation control that effectively responded to the two Omni manipulators' position transformation of the virtual manipulator. The time delay between the real slave force feedback and the virtual interface is addressed by designed event-trigger-based control in order to efficiently reduce the impact of time communication outage. The system's stability is analyzed and robot experiments are performed. From the experimental results, the telepresence platform innovatively applied virtual force feedback to reveal the soft target grasping and to accurately estimate the interactive force, enabling sensorless force feedback control.