Abstract

If the models are the result of partial differential equations being discretized, they are often large-scale in terms of number of states and possibly also number of parameters. Estimating a large number of parameters from measurement data leads to problems of identifiability, and consequently to inaccurate identification results. The question whether a physical model structure is identifiable, is usually considered in a qualitative way, i.e. it is answered with a yes/no answer. However since also nearly unidentifiable model structures lead to poor parameter estimates, the questions is addressed how the model structure can be approximated so as to achieve local identifiability, while retaining the interpretation of the physical parameters. Appropriate attention is also given to the relevant scaling of parameters. The problem is addressed in a prediction error setting, showing the relation with gradient-type optimization algorithms as well as with Bayesian parameter estimation. The problem setting is motivated by a challenging large-scale modelling problem in petroleum reservoir engineering.

If time permits some attention will be given to the quantification of parameter uncertainty intervals on the basis of alternative test statistics in prediction error identification.

Short Biography

Paul M.J. Van den Hof received the Ph.D. degree from the Department of Electrical Engineering, Eindhoven University of Technology, The Netherlands in 1989. Since 1999 he is a full professor at Delft University of Technology, where since 2004 he is Director of the Delft Center for Systems and Contro. Since 2005 he also serves as scientific director of the national graduate school DISC (Dutch Institute of Systems and Control.

His research work has concentrated around system identification and identification for control, including aspects of closed-loop experiments, uncertainty modeling and parametrization issues. His research interests further cover signal processing and (robust) control design, while dealing with applications in industrial process control systems, physical measurement systems and motion control systems.

He has acted as General Chair for the 13th IFAC Symposium on System Identification, that was held in Rotterdam, the Netherlands in 2003. He has been a member of the editorial board of Automatica (1992-2005), the IFAC Council (1999-2005), and the Board of Governors of IEEE's Control System Society (2003-2005). He is an IFAC Fellow and a Fellow of IEEE.