Command sequence replanning using discrete-event-based task model in tele-robotic part mating

Abstract

A sequence replanning/control method is proposed to overcome geometric uncertainties of the slave environment in our a priori knowledge of a discrete-event-based tele-task model. A class of Controlled Petri Net (CPN) is adopted as a modeling tool of tele-robotic part-mating process and the discrete-event control policy of the process is formulated on the basis of the CPN model. In the case when an unexpected state transition occurs due to the uncertainty of the slave environment, recovery paths and their costs are extracted from the a priori knowledge and one of them to minimize the given cost is chosen as the replanned sequence in order for the slave process to escape from the unexpected state and resume the nominal plan.