Adaline Neural Network and Real-Imaginary Instantaneous Powers Method for Harmonic Identification

Numéro de la revue: 36
Auteurs: Larbi Hamiche1, Salah Saad1, Leila Merabet1 & Fares Zaamouche2

 1LSELM Badji-Mokhtar University, BP 12, Annaba 23000, Algeria,
2Department of Mining Engineering, Faculty of science and technology,
Larbi Tebessi University, 12000, Tebessa, Algeria

Télécharger l’articlep2


In this paper, a shunt active power filter (SAPF) is used for reactive power compensation and harmonic current suppression. Algorithm based on (ADALINE) Neural Network and Real and Imaginary Instantaneous Powers (ANN-RIIP) is developed in order to estimate and identify the harmonic terms individually and online. This algorithm is simulated in MATLAB/Simulink environment and implemented experimental in a dSPACE 1104 controller. The results obtained by simulation are in agreement with the experimental results confirming the effectiveness of the proposed ANN-RIIP algorithm. The results have also demonstrated that the speed and accuracy of ADALINE can improve the performances of active power filters.

Keywords: Active Power Filter – Harmonics Currents -ADALINE Neural Network -“pq” theory – Fourier series.