ENGINEERINGNET.EU -- Mobile machines become increasingly complex and have to execute robust services under changing impacts.
Under the interdisciplinary “Organic Computing in Off-highway Machines” (OCOM) project, researchers at the Karlsruhe Institute of Technology develop a flexible overall machine management system tailored to these requirements.
It is based on the so-called observer/controller architecture that monitors the tractor system and interferes whenever a given target function, such as fuel consumption or pollutant emission, can be improved.
For this purpose, the observer acquires characteristic data about the current state of the system and assigns them to a cluster. This cluster is transmitted to the controller.
A mapping function allocates optimum tractor adjustments to every cluster, as a result of which the system behavior is influenced according to the target function given by the user.
A history and online evaluation module evaluates the consequences of the adjustments in practice.
An adaptation module contains a model-based learning method that determines the potentials of new tractor adjustments in a given system state.
This organic overall machine management is planned to be integrated in a real tractor.
(picture: KIT, Chair for Mobile Machines)
