Majlesi Journal of Energy Management Vol. 7, No. 2, June 2018 13 Sensorless Twelve Sectors Implementation of Neural DPC Controlled DFIG for Reactive and Active Powers Ripples Reduction Habib Benbouhenni 1* , Zinelaabidine Boudjema 2 , Abdelkader Belaidi 1 1- Laboratoire d’Automatique et d’Analyse des Systèmes (LAAS), Departement de Génie Électrique, Ecole Nationale Polytechnique d’Oran Maurice Audin, Oran, Algeria. E-mail: habib0264@gmail.com (Corresponding author), belaidiaek@gmail.com 2- Laboratoire Génie Électrique et Energies Renouvelables (LGEER), Electrical Engineering Department, Hassiba Benbouali University, Chlef, Algeria. Email: boudjema1983@yahoo.fr Received: January 2018 Revised: February 2018 Accepted: April 2018 ABSTRACT A direct power control (DPC) drive allows independent and direct command of reactive power linkage and stator active power by the selection of optimum inverter switching tables (STs) of an induction generator (IG). There is no need for any complex conversion of voltage or current. However, each vector selected from the ST cannot produce the required accurate voltage vector to provide the desired active and reactive powers. This results in the production of ripples in the reactive power as well as active power waveforms. In this study, we propose a technique to minimize active and reactive powers fluctuations. In this proposed strategies, the circular flux vector is divided into 12 sectors of 30 degrees and is compared to each other. Switching table is based on neural networks controller. The proposed strategies of 12 sectors DPC strategy are simulated and the comparison of their performances is presented. KEYWORDS: DFIG, DPC, Twelve Sectors, Neural Networks. 1. INTRODUCTION Nowadays, doubly fed induction generators (DFIGs) are widely used in the wind turbine due to their low maintenance need and simple form. Many different techniques have been tried to regulate reactive and active powers of the DFIGs. For years, the DFIG command market is dominated by the field oriented control techniques. However, the latest trend is the development of the DPC strategy because it is fast, simple and more advantageous [1]. The DPC method is sufficient to command with respect to changes in the DFIG parameters without using reverse current regulation in addition to providing fast dynamic reactive and active powers response [2]. The DPC control scheme does not require axes transformation and voltage decoupling blocks [3]. In this work, two strategies of the twelve sectors DPC strategy with the application of the neural networks has been considered. Two switching table is proposed for twelve sectors DPC in this paper and the application of the neural networks in the proposed strategies with DFIG-based wind power conversion systems (WPCSs). This work is divided into four sections. In Section 1, the introduction is presented. In Section 2, the description of the proposed strategies of twelve sectors DPC control scheme is presented. Section 3 deals with the description of the twelve sectors DPC control with the application of neural networks controller. Simulation studies are presented and discussed in Section 4. The paper is concluded with a short conclusion. 2. WIND TURBINE CHARACTERISTICS The wind energy conversion systems are recently getting a lot of attention, for being cost-variable, inexhaustible, safe renewable energy sources and environmentlly clean compared to the thermal and nuclear power generation systems. The wind turbine (WT) input power is usually [4, 5] : V S w P v 3 5 . 0   (1) Where, ρ is air density. Sw: Surface swept by the propeller (m 3 ). V: Wind speed (m/s). The mechanical power of WT is : V S w P v C p P m 3 5 . 0 .    (2) Where : Cp: The aerodynamic coefficient of power.