Naslov
Neural Network and Fuzzy Logic Based Control of Induction Machines

Autori
Vukadinović, Dinko

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, pp prezentacija, ostalo

Izvornik
Mathematical Applications in Science and Mechanics - Proceedings of the 4th European Conference for the Applied Mathematics and Informatics (AMATHI '13) / Trisovic, Natasa ; Rasteiro, Deolinda ; Tenorio, Angel F. ; Simian, Dana ; Minea, Alina A. ; Roushdy, Mohamed ; Salem, Abdel-Badeeh M. ; - Dubrovnik : WSEAS Press, 2013

ISBN
978-960-474-305-6

Skup
4th European Conference for the Applied Mathematics and Informatics (AMATHI '13)

Mjesto i datum
Dubrovnik, Hrvatska, 25.06.2013. - 27.06.2013

Vrsta sudjelovanja
Plenarno

Vrsta recenzije
Nije recenziran

Ključne riječi
Artificial Neural Networks; Fuzzy Logic; Induction Machine Control; Parameter Estimation

Sažetak
Induction motors have been used as the workhorse in industry for a long time due to their being easy to build, highly robust, and having generally satisfactory efficiency. In addition, induction generators play an important role in renewable energy systems such as energy systems with variable-speed wind turbines. The induction machine is a nonlinear multivariable dynamic system with parameters that vary with temperature, frequency and magnetic saturation. Considering that neural networks are capable of handling time varying nonlinearities due to their own nonlinear nature, they are suitable for application in induction machine systems. This lecture presents a brief review of applications of artificial neural networks and fuzzy logic for induction machines. Most applications of neuro and/or fuzzy theory in induction machine control systems focus on advanced controllers for speed, position or voltage, where the conventional PI controller is replaced by a neuro and/or fuzzy controller. Few other applications will also be shown in this lecture, such as: neural network-based speed estimator, neural network-based inverter control, applications of neural networks in waveform processing and delayless filtering, identification of machine parameters based on fuzzy/neural concepts and approaches for the efficiency improvement in induction machine systems. Some of the presented simulation and experimental results are obtained at the Research laboratory for Power Electronics of the Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture in Split, Croatia.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika

POVEZANOST RADA