Time-periodic state estimation with event-based measurement updates
bookPart
To reduce the amount of data transfers in networked systems, measurements can be taken at an event on the sensor value rather than periodically in time. Yet, this could lead to a divergence of estimation results when only the received measurement values are exploited in a state estimation procedure. A solution to this issue has been found by developing estimators that perform a state update at both the event instants as well as periodically in time: when an event occurs the estimated state is updated using the measurement received, while at periodic instants the update is based on knowledge that the sensor value lies within a bounded subset of themeasurement space. Several solutions for event-based state estimation will be presented in this chapter, either based on stochastic representations of random vectors, on deterministic representations of random vectors or on a mixture of the two. All solutions aim to limit the required computational resources by deriving explicit solutions for computing estimation results. Yet, the main achievement for each estimation solution is that stability of the estimation results are (not directly) dependent on the employed event sampling strategy. As such, changing the event sampling strategy does not imply to change the event-based estimator as well. This aspect is also illustrated in a case study of tracking the distribution of a chemical compound effected by wind via a wireless sensor network. © 2016 by Taylor & Francis Group, LLC.
Topics
TNO Identifier
954677
ISBN
9781482256
Publisher
CRC Press
Source title
Event-Based Control and Signal Processing
Editor(s)
Miskowicz, M.
Place of publication
Boca Raton
Pages
261-279
Files
To receive the publication files, please send an e-mail request to TNO Repository.