Application of Prediction-based Particle Filters for Teleoperations over the Internet

J.-y. Lee, S. Payandeh, and L. Trajković (Canada)


Teleoperation, Internet, Particle filter, Sequential MonteCarlo method, Bayesian prediction


In this paper, we present the prediction-based particle filter approach for processing motion and force data in teleoperation over the Internet. We first introduce the prediction-based particle filter algorithm, one of the Sequential Monte Carlo methods based on the recursive Bayesian prediction. The prediction algorithm is applied to dynamic models of the motion and force data flows in the state-space formulation. It is applied to the motion data transmitted to the slave controller and to the reflecting force data received at the master controller. Experiments are performed using the haptic device within a virtual 3D graphical environment. In each experiment, the motion and reflecting force data extracted from the haptic device are used to verify the prediction performance of the proposed method.

Important Links:

Go Back