A new switching table based neural network for direct power control of three-phase PWM-rectifier. (27th May 2020)

• Record Type: Journal Article
• Authors: Fekik, Arezki; Denoun, Hakim; Zaouia, Mustapha; Hamida, Mohamed Lamine; Vaidyanathan, Sundarapandian
• Is Part Of: International journal of modelling, identification and control. Volume 33:Number 4(2020)
• Page Start: 358
• Page End: 368
• Publication Date: 2020-05-27
• Subjects: ANN -- artificial neural network -- DPC -- direct power control -- instantaneous active and reactive power -- PWM -- pulse width modulation -- switching table -- UPF -- unity power factor; Engineering -- Methodology -- Periodicals; Science -- Methodology -- Periodicals; 001.42
• Languages: English
• ISSNs: 1746-6172

Abstract

Direct power control (DPC) is one of the newest techniques to control the PWM converter without network voltage sensors. This control technique is built on the idea of direct torque control (DTC) for an induction motor, which is applied to eliminate the harmonic of the line current and to compensate the reactive power. The principle of this control is based on instant active and reactive power loops. This article proposes an intelligent control approach to improve this control technique, such as artificial neural network (ANN), applied to the switching table. The comparison with conventional DPC shows that the use of DPC-ANN ensures smooth control of active and reactive power in all sectors and reduces current ripple. Finally, the developed DPC was tested by simulation. The results proved the excellent performance of the proposed DPC scheme in comparison with the conventional DPC.