Networked Multi-Manipulator System and Its Teleoperation Using Adaptive Non-Singular Terminal Sliding Mode Control

Abstract

Teleoperation of multiple robot manipulators has been one of the most popular research areas in the robotics research community for the last couple of decades. Such complex systems can be decoupled into two subsystems, namely, multi-agent systems (MASs) and teleoperation systems.

In addition to the high nonlinearity of the networked multi-manipulator systems, deleterious effects, caused by network-induced constraints and the lack of exact robot modelling information, can make the control systems' desired performance and stability difficult to achieve. To meet these challenges, concepts from the non-singular terminal sliding mode (NTSM) control method are developed to achieve the exogenous disturbance rejection and the finite-time full-pose synchronization. Additionally, a new adaptive NTSM (ANTSM) scheme is designed for multi-manipulator systems where the models may be initially uncertain or slowly varying over time. To further improve the performance, a set of novel techniques are developed, including the use of novel mixed-type feedback, time-varying logistic-function-based control gain, and energy-index-based neighbour selection policy. The proposed ANTSM approach has also successfully been applied to the teleoperation control systems. In addition, the master manipulator uses a force predictor to estimate the real-time environmental force on the slave side so that the direct transmission of the force signals is avoided.

The proposed approaches for the MASs and teleoperation subsystems are integrated into a single-master-multiple-slave manipulator system. Simulation and experimental results validate the efficacy of the proposed schemes.