KOM 512 - System Identification

Course Objectives

• To understand the important of modeling concept, types of models , mathematical modeling and system identification for control science and engineering
• To teach nonparametric methods in system identification; transient, frequency, correlation and spectral analysis
• To teach parametric methods in system identification; linear regression, the least squares estimation, the best linear unbiased estimate
• To understand the properties of input signals which are used for a system identification experiment and the concept of persistent excitation
• To teach model structure and model estimation
• To teach prediction error methods, instrumental variable methods, recursive system identification methods, system identification in closed loop
• To teach model validation and model structure determination

Course Description

Modeling concept, types of models and examples. Comparison of mathematical modeling and system identification. Flowchart of system identification, basic definitions and concepts. Nonparametric methods in system identification; transient, frequency, correlation and spectral analysis. Parametric methods in system identification; linear regression, the least squares estimation, the best linear unbiased estimate. Input signals and persistent excitation. Model estimation. Prediction error methods, instrumental variable methods. Recursive system identification methods, system identification in closed loop. Model validation and model structure determination