Asif Rana's Home Page

Updated: 23-Apr-2017

Research Profile

PhD Dissertation Abstract

It is known that time delay in bilateral teleoperation can drive a system to instability. Time Domain Passivity Control (TDPC) deals with the stabilization of haptic interfaces in teleoperation using the notion of passivity directly in the continuous time variables like force and velocity. In this work, first it is shown that TDPC can be extended to stabilize the time-delayed teleoperation by considering the communication channel as an active component and then, to design passivity controllers for it on master side using a Kalman filter based recursive prediction of slave side energy. However, such a scheme is prone to large corrective impulses generated by passivity controllers as the scheme only comes into effect when the net energy goes negative, while on other time instants it stays out of the control loop in order to provide maximally transparent teleoperation. These impulses degrade the performance of teleoperator. It is thus further proposed, that the derivative of net energy should also be computed in real-time, and as soon as this term becomes negative, indicating a decline in the net energy, the passivity controllers should immediately compensate this active behavior. This forces the system to always dissipate energy and thus stop the occasional accumulation of a large amount of negative energy. In addition to that, parabolic power integration is employed to provide non-linear estimation of net energy in the communication channel.

The approach thus developed is then used to stabilize time-varying delays. In order to provide a time-base for the predictor, a first order one-step ahead prediction of RTT (Round Trip Time) is used. Beta distributed and TrueTime network simulator based delays are used to evaluate the system performance. Simulation results are given showing the efficacy of the proposed approach.

Keywords: Time Domain Passivity Control, Stabilization, Teleoperation, Time Delay, Telerobotics

MS Thesis Abstract

Telerobotics is a scheme that allows humans to extend their manipulative skills over a network by combining human's cognitive skills and robot's specialized working abilities. Efficient control of the robot over LAN in the presence of time delays and data loss is a dynamic research field. The purpose of this work is to implement a reliable teleoperation of PUMA 560 robot over a LAN. In order to pursue this goal, a completely distributed telerobotic framework is developed using .NET Remoting and SOAP(Simple Object Access Protocol) technologies to provide multistream, multithreaded environment for real-time interaction between client and server side components. Computer vision and force feedback techniques are implemented and their performance is evaluated in order to enhance the maneuverability of the operator telemanipulating the robot.

Keywords: Computer Vision, Telerobotics, Augmented Reality, Multistream, Multithreaded, Force Feedback