## TIEE3032 Syllabus - System Identification - 2022 Regulation Anna University

TIEE3032

SYSTEM IDENTIFICATION

L T P C

3003

COURSE OBJECTIVES:
• To elaborate the concept of estimating the state variables of a system using state estimation algorithms.
• To elaborate the concept of estimating the parameters of the Input-output models using parameter estimation algorithms.
• To make the student understand the various closed loop system identification techniques.
• To make the student understand the various closed loop system identification techniques.
• To provide the background on the practical aspects of conducting experiments for real time system identification.

UNIT I

NON PARAMETRIC METHODS

(7+2 SKILL)9

Nonparametric methods: Transient analysis - frequency analysis - Correlation analysis - Spectral analysis.

UNIT II

PARAMETRIC METHODS

(7+2 SKILL)9

Parametric model structures: ARX, ARMAX, OE, BJ models - The Least square estimate - Best linear unbiased estimation under linear constraints - Updating the Parameter estimates for linear regression models - Prediction error methods: Description of Prediction error methods - Optimal Prediction – Relationships between prediction error methods and other identification methods - theoretical analysis. Instrumental variable methods: Description of Instrumental variable methods - Theoretical analysis - covariance matrix of IV estimates - Comparison of optimal IV and prediction error methods.

UNIT III

RECURSIVE IDENTIFICATION METHODS

(7+2 SKILL)9

The recursive least squares method - Recursive Instrumental variable method-the recursive prediction error method-model validation and model structure determination. Identification of systems operating in closed loop: Identifiability considerations - Direct identification - Indirect identification - Joint input – Output identification.

UNIT IV

CLOSED- LOOP IDENTIFICATION

(7+2 SKILL)9

Identification of systems operating in closed loop: direct identification and indirect identification – Subspace Identification methods: classical and innovation forms – Relay feedback identification of stable processes.

UNIT V

NONLINEAR SYSTEM IDENTIFICATION

(7+2 SKILL)9

Modeling of nonlinear systems using ANN- NARX & NARMAX - Training Feed-forward and Recurrent Neural Networks – TSK model – Adaptive Neuro-Fuzzy Inference System (ANFIS) - Introduction to Support Vector Regression.

TOTAL: 45 PERIODS

OUTCOMES:
CO1 Ability to design and implement state estimation schemes. L5
CO2 Ability to develop various models (Linear & Nonlinear) from the experimental data. L5
CO3 Be able to choose a suitable model and parameter estimation algorithm for the identification of systems. L3
CO4 Be able to illustrate verification and validation of identified model. L3
CO5 Ability to develop the model for prediction and simulation purposes using suitable control schemes. L5

TEXT BOOKS:
1. Lennart Ljung, “System Identification: Theory for the user”, 2nd Edition, Prentice Hall, 1999.
2. Dan Simon, “Optimal State Estimation Kalman, H-infinity and Non-linear Approaches”, John Wiley and Sons, 2006,
3. Tangirala, A.K., “Principles of System Identification: Theory and Practice”, CRC Press, 2014, 1st Edition.

REFERENCES:
1. Cortes, C., and Vapnik, V., “Support-Vector Networks, Machine Learning”, 1995, 1st Edition.
2. Miller, W.T., Sutton, R.S., and Webrose, P.J., “Neural Networks for Control”, MIT Press, 1996, 1st Edition.
3. Van der Heijden, F., Duin, R.P.W., De Ridder, D., and Tax, D.M.J., “Classification, Parameter Estimation and State Estimation”, An Engineering Approach Using MATLAB, John Wiley & Sons Ltd., 2017, 2nd Edition.
4. Karel J. Keesman, “System Identification an Introduction”, Springer, 2011, 1st Edition.
5. Tao Liu and Furong Gao, “Industrial Process Identification and control design, Step-test and relay-experiment-based methods”, Springer- Verlag London Ltd., 2012, 1st Edition.