State and Parameter Estimation
State and Parameter Estimation

Zur Modellierung (mathematische Beschreibung) eines dynamischen Systems werden vollständige Informationen über die Modellstruktur, die Zustandsgrößen und die Modellparameter benötigt. In dieser Vorlesung werden Methoden
• zur Zustandsschätzung
• zur Parameteridentifikation
• zur Systemidentifikation
behandelt. Ferner werden Methoden zur direkten Identifikation von Reglern und Beobachtern vorgestellt.

Die Studierenden sollen verschiedene Methoden zur Zustandsschätzung und Parameteridentifikation kennenlernen und diese in Form von Algorithmen umsetzen können.

[1] S. X. Ding, Vorlesungsskript "State and parameter estimation” (wird jährlich aktualisiert, per Download verfügbar, will be updated and available for download)
[2] T. Kailath and A. Sayed and B. Hassisi, Linear estimation, Prentice Hall, 1999.
[3] R. Isermann and M. Münchhof, Identification of Dynamic Systems Springer-Verlag, 2011
[4] B. Huang and R. Kadali, Dynamic Modeling, Predictive Control and Performance Monitoring - A Data-driven Subspace Approach. Springer-Verlag, London 2008
[5] S. X. Ding, Data-driven design of fault diagnosis and fault-tolerant control systems, Springer-Verlag, 2014.

Zustands- und Parameterschätzung

A dynamic system is well described by its model structure, state variables and model parameters. In practice, they are often unknown and should be identified or estimated. In this course, basic methods for the identification and estimation of state variables and system parameters are introduced.
The course consists of four thematic blocks.
In Block I, State estimation - Kalman filter and observer schemes, different types of Kalman filters and observer schemes are introduced on the assumption that the system model and parameters are available, including
• state estimation in static processes
• State estimation in (linear) dynamic processes
• H2 optimal observer.
In Block II, Parameter identification -
Least squares parameter estimation schemes, parameter identification is dealt on the assumption of a given system structure. Topics like parameter estimation in static processes, parameter estimation in dynamic processes and recursive algorithms are addressed.
In case that the system is a block box, system identification is needed. In Block III, System identification -
Subspace identification methods (SIM), the basic ideas and procedure of SIM are first introduced. It is followed by some standard SIMs. Block IV, SIM-added identification of kernel and image representations and data-driven design of feedback controllers and observers, is dedicated to the introduction of some data-driven design methods for controllers and observers.

The students should learn basic state estimation and parameter identification methods and be able to implement them in form of algorithms.

[1] S. X. Ding, Vorlesungsskript "State and parameter estimation” (wird jährlich aktualisiert, per Download verfügbar, will be updated and available for download)
[2] T. Kailath and A. Sayed and B. Hassisi, Linear estimation, Prentice Hall, 1999.
[3] R. Isermann and M. Münchhof, Identification of Dynamic Systems Springer-Verlag, 2011
[4] B. Huang and R. Kadali, Dynamic Modeling, Predictive Control and Performance Monitoring - A Data-driven Subspace Approach. Springer-Verlag, London 2008
[5] S. X. Ding, Data-driven design of fault diagnosis and fault-tolerant control systems, Springer-Verlag, 2014.