System Identification

The area of system identification has been an active one in the Systems and Control Group for several decades. There is some overlap with the area signal processing, and both the senior researchers have previously also done significant amounts of research in signal processing. The current research is supported by funds from the Swedish Research Council.

Senior researchers

Alexander Medvedev, Kristiaan Pelckmans,Torsten Söderström, Torbjörn Wigren

PhD students

Liang Dai

Projects

• Networked identification of nonlinear systems
• Estimation of material functions using system identification techniques
• Identification of dynamic error-in-variables
• System identification using continuous-time models
• General aspects
• Convex Optimization for System Identifcation

Some further projects in system identification can be found here.

Past projects

• Non-linear identification and machine learning in systems biology
• Identification of selective catalytic reduction systems in automotive applications
• Black-box identification of non-linear systems using ordinary differential equation models
• Identification of non-linear Wiener systems
• Identification and control of non-linear solar collector plants with long and varying delays
• Recursive parameter estimation under lack of excitation
• Unbalance estimation of separators
• Identification of nonlinear systems and modelling of periodic signals
• Numerical algorithms in signals and systems
• System identification and model reduction of diffusion models
• Identification and estimation using quantized output signal measurements

Free book

The book System Identification by Torsten Söderström and Petre Stoica is out of print.
It is available for download here.

Software

Free software for identification of nonlinear systems are available here.

Some achievements

• Students who in recent years have completed a PhD thesis in the research area include Emad Abd-Elrady, Bharath Bhikkaji, Linda Brus, Magnus Evestedt, Mei Hong-Bjerstedt, Magnus Mossberg, Peter Nauclér, Agnes Rensfelt, Joakim Sorelius, Erik K. Larsson and Kaushik Mahata.

• One research monograph was authored:
  • T. Söderström: Errors-in-Variables Methods in System Identification. Springer, 2018.

• Three textbooks in the field have been co-authored, are used frequently, and were very well recieved:
  • L. Ljung and T. Söderström: Theory and Practice of Recursive Identification. MIT Press, Cambridge, Mass, 1983.
  • T. Söderström and P. Stoica: Instrumental Variable Methods in System Identification. Springer-Verlag, Berlin, 1983.
  • T. Söderström and P. Stoica: System Identification, Prentice-Hall International, Hemel Hempstead, UK, 1989. Available for download here.

• A large number of articles have been written. Details are given under each project.

• Torsten Söderström was appointed IEEE Fellow 1992, 'for fundamental contributions to system identification', and to IFAC Fellow 2006, 'for scientific contributions to system identification, for pedagogical explanations of identification to wide audiences and for inspiration to the scientific community'.

• Torsten Söderström was associate editor 1984-1991, and is editor 1992-present, for the IFAC journal Automatica, subject area System parameter estimation.