Prediction of Lost Data in Virtual Control System with Networked Sensors

In majority of the control applications, individual sensors, or dedicated ´loop controllers´, are hooked to the central control room directly through some kind of a bus or data highway. Virtual control system (VCS) paradigm assumes a virtual controller (control room), existing in a virtual network space, but with real endpoints. The real endpoints are realized with network connected sensors and actuators. One of the main issues in such an environment is the loss of control during communication failures, both by the way of delayed data, as well as due to communication media failure. . This paper proposes the concept of such a VCS, and suggests the implementation of a predictor, which extrapolates previous data, to fill-in missing data during short communication failures. Improvement in control, if any, has been investigated in a simulated environment.