Module A: Intelligent systems in decision support

Course description

  • Introduction to intelligent systems. Intelligent systems in data-mining, classification and fault detection.
  • Basic methods of local nonlinear optimization used in intelligent systems and global nonlinear optimization methods for model identification.
  • Methods of global nonlinear optimization: simulated annealing, evolutionary algorithms, particle swarm optimization, genetic algorithms, branch and bound algorithms.
  • Unsupervised learning methods. Principle component analysis.  PCA in identification, data filtering, control and fault detection.
  • Data clustering. Methods of clustering: fuzzy c-means, Gustafon-Kessel fuzzy c-means, possibilistic c-means clustering, method of regression clustering.
  • Optimization of complex models. Verification and validation of models. Explicit and implicit optimization of model structure.
  • Static models. Model based on basis function formulation. Polynomial models.
  • Neural networks. Multilayer perceptron network. Radial basis function networks in function approximation.
  • Fuzzy and neuro-fuzzy models. Fuzzy logic. Types of fuzzy models. Estimation of fuzzy model parameter. Global and local estimation. Different structures of fuzzy controllers.
  • Nonlinear dynamical systems. Classical polynomial models in nonlinear modelling. Identification of dynamical fuzzy and neuro-fuzzy models.
  • Interval fuzzy model and families of functions.
  • Supervised hierarchical clustering in experiment design.
  • Control of nonlinear dynamical systems. Gain scheduling control algorithm.
  • Internal nonlinear model control algorithm. 2DOF control algorithm.
  • Nonlinear model based control. Predictive functional control (PFC) and fuzzy model based predictive functional control.
  • Predictive control based on dynamical matrix (DMC). Predictive control based on step response. Predictive control based on state-space model.
  • Predictive control based on nonlinear model and optimization.
  • Adaptive control and online adaptation. Robust adaptive laws. Model-reference adaptive systems. Fuzzy model-reference adaptive systems.
  • Monitoring, fault detection and isolation based on intelligent systems.

Course is carried out on study programme

Elektrotehnika 2. stopnja

Objectives and competences

To provide students with an understanding of the basic mathematical and computational principles of constructing artificial perception systems, which are an essential part of intelligent systems in automation and control.

Learning and teaching methods

  • lectures,
  • laboratory exercises and projects,
  • coursework.

Intended learning outcomes

Knowledge and understanding:

After completing this course the student will be able to demonstrate a knowledge and understanding of the:

  • construction of intelligent systems for decision support and control ,
  • identification of static and dynamic models based on intelligent methods,
  • construction of advanced control systems based on intelligent systems.

The use of knowledge:

The student will be able to use the acquired knowledge to construct technical systems for monitoring, forecasting, analysis, control and fault detection. The student will be able to critically evaluate the consistency between the acquired knowledge and the application in practice.

Transferable skills:

  • the use of literature and other resources in the fields of pattern recognition, machine learning and artificial intelligence;
  • the use of development tools and environments for computer programming (writing computer or using the Matlab development environment),

problem solving: problem analysis, algorithm design, implementation and testing of a program.

Reference nosilca

Izvirni znanstveni članek / Original scientific article

  1. ŠKRJANC, Igor. Evolving fuzzy-model-based design of experiments with supervised hierarchical clustering. IEEE transactions on fuzzy systems, ISSN 1063-6706. [Print ed.], 2014, vol. , no. , str. 1-12.
  2. ŠKRJANC, Igor. Fuzzy confidence interval for pH titration curve. Applied mathematical modelling, ISSN 0307-904X. [Print ed.], Aug. 2011, vol. 35, no. 8, str. 4083-4090.
  3. HARTMANN, Benjamin, BÄNFER, Oliver, NELLES, Oliver, SODJA, Anton, TESLIĆ, Luka, ŠKRJANC, Igor. Supervised hierarchical clustering in fuzzy model identification. IEEE transactions on fuzzy systems, ISSN 1063-6706. [Print ed.], Dec. 2011, vol. 19, no. 6, str. 1163-1176.
  4. BELIČ, Aleš, ŠKRJANC, Igor, ZUPANČIČ-BOŽIČ, Damjana, VREČER, Franc. Tableting process optimisation with the application of fuzzy models. International journal of pharmaceutics, ISSN 0378-5173. [Print ed.], Apr. 2010, vol. 389, no. 1/2, str. 86-93.
  5. ŠKRJANC, Igor. Confidence interval of fuzzy models : an example using a waste-water treatment plant. Chemometrics and Intelligent Laboratory Systems, ISSN 0169-7439. [Print ed.], Apr. 2009, vol. 96, no. 2, str. 182-187.

Study materials

I. Škrjanc: Inteligentni sistemi pri raziskovanju podatkov in odločanju, skripta, 2016.

Bodi na tekočem

Univerza v Ljubljani, Fakulteta za elektrotehniko, Tržaška cesta 25, 1000 Ljubljana

E: T:  01 4768 411