State Estimation of Multichannel Networked Discrete Event Systems

Abstract

In a networked discrete event system, communication between a plant and a supervisor/agent is carried out via a wired/wireless communication network, in which communication delays are unavoidable. Hence, how to obtain state estimates under communication delays is an important problem. The existing methods assume that there is only one communication channel and, thus, the first-in, first-out (FIFO) property is satisfied. In complex systems distributed over several locations, several channels are often used to send the event occurrences to the (central) supervisor/agent, in which the FIFO property holds only for each channel and communication delays may be different for different channels. To address these new challenges, this paper investigates the state estimation problem in multichannel networked discrete event systems. Two methods are proposed. The first method is relatively simpler and computes only conservative estimates. The second method is more complex, but it computes the exact state estimates. Examples are given to illustrate the results.