The four-channel architecture in teleoperation with force feedback has been studied in various existing literature. However, most of them focused on Lawrence architecture and did not research other cases. This paper proposes two other four-channel architectures: passive four-channel architecture and passive four-channel architecture with operator force. Furthermore, two types of multilateral shared control architecture based on passive four-channel architecture, which exists in space teleoperation, are put forward. One is dual-master multilateral shared control architecture, and the other is dual-slave multilateral shared control architecture. Simulations show that these four architectures can maintain stability in the presence of large time delay.
Top1. Introduction
Teleoperation has significant application in space (Sheridan, 1993; Nohmi, 2003), undersea (Jordán & Bustamante, 2007), toxic waste cleanup (Manocha, Pernalete, & Dubey, 2001), and telesurgery (Tavakoli, Patel, & Moallem, 2006; Tobergte, Konietschke, & Hirzinger, 2009) projects. A human operator conducts a remote task via the master and slave manipulators, and the contact force information is reflected to the human operator in bilateral teleoperation. The force-reflecting bilateral control can distinctly improve the performance in teleoperation, which enables the operator to feel present at the remote location even though not really there. It is well known that the time delay is an intractable problem in bilateral teleoperation, especially in the space operations, the delay time sometimes will reach several seconds; this can easily destabilize the bilateral control system.
Up to now, the stability of a bilateral teleoperation system had been studied in various existing literature, and the most important method among them was passivity theory, which was introduced to the bilateral control to stabilize the system under random delay times. At first, Hannaford (1989) proposed a 2-port network mode, and took a hybrid matrix as bilateral representation, and then Anderson and Spong (1989) presented a control law for teleoperators by using passivity and scattering theory, which overcame the instability caused by time delay. Thereafter, Niemeyer and Slotine (1991) proposed a notion of wave variable to characterize time delay systems and got a new configuration for force-reflecting teleoperation. Kawashima et al. (2009) presented a controller for bilateral teleoperation based on the modified wave variable control method which provided superior position and force tracking performance compared to the traditional wave-variable-based method. Recently, Lee and Spong (2006) proposed a control framework for bilateral teleoperation of a pair of multi-degree-of-freedom nonlinear robotic systems under constant communication delays. Chopra et al. (2006) then improved the traditional passivity-based configuration by using additional position control on both the master and slave robots, to solve the steady-state position and force-tracking problem. Furthermore, Shahdi and Sirouspour (2009) proposed a systematic design procedure for improving teleoperation fidelity while maintaining its stability in the presence of dynamic uncertainty and a constant time delay.
With the development of the bilateral control strategy, Lawrence (1993) defined the performance of transparence and presented a general four-channel bilateral control mode. Since he proposed the four-channel control architecture, many researchers had amended and improved his works. For example, Hashtrudi-Zaad and Salcudean (2002) improved the Lawrence architecture and used master and slave local force feedback to improve the system performance and stability. Sumiyoshi and Ohnishi (2004) introduced the transformation of the Hashtrudi-Zaad architecture to clarify the meaning of the local force feedback. Guiatni et al. (2005) then utilized sliding-mode control in four-channel architecture. Naerum and Hannaford (2009) furthermore stated and proved necessary and sufficient conditions for transparency of the Lawrence architecture.