Open AccessArtificial neural network‐based discrete‐fuzzy logic controlled active power filter
Author(s) -
Saribulut Lütfü,
Teke Ahmet,
Tümay Mehmet
Publication year - 2014
Publication title -
iet power electronics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.637
H-Index - 77
eISSN - 1755-4543
pISSN - 1755-4535
DOI - 10.1049/iet-pel.2013.0522
Subject(s) - rectifier (neural networks) , artificial neural network , harmonics , control theory (sociology) , computer science , fuzzy logic , harmonic , neuro fuzzy , signal (programming language) , electric power system , thyristor , filter (signal processing) , fuzzy control system , power (physics) , electronic engineering , engineering , artificial intelligence , time delay neural network , voltage , types of artificial neural networks , control (management) , physics , programming language , quantum mechanics , electrical engineering , computer vision
Artificial neural network (ANN) is a computational algorithm based on the structure and functions of biological neural networks. It is used for modelling of the non‐linear systems that cannot be mathematically expressed by the formula and extraction of the system dynamics, expressed by using the complex mathematical equations, such as harmonics. To show the effective usage of ANNs in the power system, the fundamental harmonic of a load with six‐pulse thyristor controlled rectifier is extracted with ANN by using the system variables that are difficult to express with each other. Then, a new approach is proposed to generate the reference signal for compensating the harmonics of the current by using discrete fuzzy logic in this study. In addition, a simple and useful method to determine the circuit parameters of the active power filter (APF) is proposed to reduce the rating of the required filter and the capacitor values without affecting its efficiency. Case studies are performed to test the performance of the proposed control algorithm for APF.