Soft Computing Control Techniques (ILV)

Back
Specialization AreaRemote Systems
Course numberM2.05284.31.691
Course code
Curriculum2016
Semester of degree program Semester 3
Mode of delivery Presencecourse
SPPW2,0
ECTS credits3,0
Language of instruction English

The students will understand soft computing techniques and their applicability in the control design.
The students will be able to design, program and apply fuzzy logic system to solve engineering control problems where only expert linguistic knowledge is available.
The students will be able to design, program and apply feedforward artificial neural network for approximation of system's dynamics and integrate it into different controllers.
The students will be able to apply evolutionary algorithms for optimization of the parameters and structure of other controllers.
Students will learn how to use available software tools.
Students will be able to evaluate and compare soft computing approaches and conventional control approaches for a given problem.
The students will get familiar with latest research and applications of soft computing control approaches.

Introduction to soft computing methods.
Fuzzy logic and fuzzy logic control systems. Linguistic variable, membership functions, fuzzy operators, fuzzy rule and rule base, compositional rule of inference, fuzzy logic system, fuzzy identification systems. Fuzzy logic controllers. Fuzzy control of inverted pendulum.
Evolutionary algorithms. Optimization problems in control. Genetic algorithm: encoding technique, ininitialization procedure, genetic operators, parameter setting. Applications in control theory. Optimization of fuzzy logic system by using evolutionary algorithms. Multi-depot multi-vehicle problem.
Artificial neural networks (ANNs). Artificial neuron, feedforward artificial neural networks, structure of ANNS, back-propagation training algorithm, estimation learning, over and under-fitting, momentum and learning rate adaptation. Recurrent neural networks. ANN based controllers. Computed torque neural network controller, sliding mode neural network motion controller. Adaptive neuro-fuzzy systems (ANFIS), ANFIS in robot control.
State of the art and conclusions. Overview of latest control applications of soft-computing control applications based on selected journal papers. Concluding notes.

Devendra K. Chaturvedi, Soft Computing: Techniques and its Applications in Electrical Engineering, Springer, 2008.
S. G. Tzafestas (Ed.); Soft Computing in Systems and Control Technology, World Scientific, 1999.
Timothy Ross, Fuzzy Logic with Engineering Applications, John Wiley & Sons Inc., 2010.
Sandhya Samarasingh, Neural Networks for Applied Sciences and Engineering: From Fundamentals to Complex Pattern Recognition, Auerbach Publications, 2006.
Selected scientific papers.