Control with Intermittent Sensor Measurements: A New Look at
Tomas Estrada Panos J. Antsaklis
Presented at the Workshop on Networked Distributed Systems for Intelligent Sensing and Control
Kalamata, Greece, Saturday, June 30, 2007
In many control systems, including networked control systems, feedback information is not necessarily contin-
uous or instantaneous, but intermittent, where the loop is closed for finite time intervals. Intermittent feedback is
not uncommon in applications, but it has not been adequately studied in control theory. The aim of this work is to
explore theoretically the advantages (and disadvantages) of intermittent feedback.
In this paper, we apply the concept of Intermittent Feedback to a class of networked control systems known
as Model-Based Networked Control Systems (MB-NCS). We first introduce the basic architecture for model-based
control with intermittent feedback, then address the cases with output feedback (through the use of a state observer)
and with delays in the network, providing a full description of the state response of the system, as well as a necessary
and sufficient condition for stability in each case. Extensions of our results to cases with nonlinear plants are also
presented. Finally, we propose future research directions.
In this paper, we deal with control systems where sensor measurements are available intermittently. We refer to
this concept as intermittent feedback. Intermittent feedback is displayed in nature and has been applied in a variety