Output details

Algorithms for recursive/semi-recursive

bias-compensating least squares system identification within the errors-in-variables framework

This research was prompted by the need to handle industrial systems where the input to a subsystem is the output from another, i.e. interconnected subsystems. With interconnected subsystems a need arises to account for noise on both input and output, leading to an errors-in-variables formulation. Algorithms based on bias compensating least squares are available and the best approaches for each practical application may be selected via the developed simulation. The work was funded via a small feasibility project £30k by ISTA, GmbH, Essen, Germany, for dealing with interconnected heating systems. Contact: Bernd Boewer, Manager Concept & Technology Products, (e-mail) bernd.boewer@ista.com