A set-valued filtering approach

In this chapter, the problem of event-based state estimation is analyzed using an approach called “set-valued filtering”. With this approach, the design of an event-based estimator becomes a much simpler task. The main benefit of using the set-valued filtering approach is that the properties of the event-based estimates designed for deterministic event-triggering conditions can be analyzed without relying on any assumptions or approximations of the probability distributions. The properties of the set-valued event-based estimates also lead to a new way of designing event-triggering conditions to simultaneously fulfill requirements on the estimation performance and the sensor-to-estimator communication rate.