Hybrid Fuzzy Logic-Artificial Neural Network Controller for Shunt Active Power Filter

Benyamina A., Moulahoum S., Colak I., BAYINDIR R.

5th IEEE International Conference on Renewable Energy Research and Applications (ICRERA), Birmingham, United Kingdom, 20 - 23 November 2016, pp.837-844 identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • City: Birmingham
  • Country: United Kingdom
  • Page Numbers: pp.837-844
  • Keywords: Hybrid fuzzy-neural controller (HFNC), Artificial neural network (ANN), Shunt active power filter (SAPF), Fuzzy logic controller (FLC), Direct current control (DCC), dSPACE 1103, SYSTEMS
  • Gazi University Affiliated: Yes


This paper presents a design and an experimental implementation of the three phase three wires shunt active power filter (SAPF) based on a hybrid fuzzy-neural controller for harmonics and reactive power correction. A direct line current control strategy is used to control the filter. Fuzzy logic controller has been modified and improved by artificial neural network system (ANN) to improve the DC voltage regulation loop of the capacitor and to get better performances. It hybridizes between the advantages of neural networks in noise rejection and the fast concept of fuzzy logic. First a fuzzy logic controller has been used to regulate the voltage of the capacitor and then an artificial neural network has been learned on the previous fuzzy controller and implemented in Vdc controller of the SAPF. Simulations have been performed to verify the robustness of the proposed HFNC controller. Finally, a real time manipulation is led via an experimental bench built on dSPACE 1103. The real time results prove that the proposed hybrid controller improved the performance of the filter under different disturbances and parameters variations.