Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Event Based Control

Sampling is traditionally done periodically. For linear time-invariant sys- tems this leads to closed loop systems that are linear but periodic. The periodic nature can be avoided by considering the behavior of the sys- tems at times that are synchronized with the sampling instants (the stro- boscopic model). This approach leads to drastic simplifications because the systems can be described by time-invariant linear discrete time sys- tems. The theory which has emerged is elegant, and it has been used extensively in practical applications of computer controlled systems. Safe implementation can be accomplished by time triggered software archi- tectures. Tools for dealing with intersample behavior is also available. Control based on periodic sampling is thus a success both practically and theoretically. There are, however, deficiencies in the theory particularly for systems with multiple sampling rates and for systems with distributed computers. With multi-rate sampling the complexity of the system de- pends critically on the ratio of the sampling rates. Event based sampling is an alternative to periodic sampling. Signals are thus sampled only when there are significant events, for example when a measured signal passes certain limits. This type of sampling is natural when using many digital sensors, such as encoders, which are intrinsically event based. Using an analog with integration theory, periodic and event based sampling can be called Riemann- and Lebesgue-sampling. Event based control is nat- ural for networked control systems where delays and packet losses are common and for systems where sensing and actuation are expensive. The talk will present examples of systems with event based control. Analysis of some simple systems illustrate that event based control can indeed be beneficial. A general architecture for event based control which resembles structures in biological systems will also be presented.

Type of Seminar:
Public Seminar
Dr. Karl Astrom
Oct 08, 2009   11am

HG E 1.2
Contact Person:

Colin Jones
File Download:

Request a copy of this publication.
Biographical Sketch:
Karl Johan Åström was educated at The Royal Institute of Technology (KTH) in Stockholm. After working for IBM Research for f ive years he was appointed Pro- fessor of the Chair of Automatic Control at Lund Institute of Technology /Lund University in 1965 where he established a new department. Åström has broad interests in control and he is listed in ISAHighlyCited. He is a life Fellow of IEEE and he has Erdös number 3. Åström has received many honors among them, the 1987 Quazza Medal from the International Federation of Automtic Control and the 1993 IEEE Medal of Honor. He is a member of the Royal Swedish Academy of Science, the Royal Swedish Academy of Engineering Science and the US National Academy of Engineering.