Analysis and comparison between different methods of current reference generation for active filters control

Abstract

In modern distribution systems the proliferation of non-linear loads results in a deterioration of the quality of voltage waveforms at the point of common coupling (PCC) of various consumers. Therefore, power-conditioning equipment is becoming more important for electric utilities and their customers. With the rapid development of semiconductor devices in power and control circuits, a new generation of equipment for power quality, the active power filters, has been developed. Their advantages over conventional means are more flexibility and very fast control response. The control of an active filter comprises two major parts: the reference current computation and the current control. There are two fundamental methods of generating the reference current: (i) frequency–domain methods, based on the Fourier analysis and (ii) time-domain analyze, based on the theory of instantaneous imaginary power in the three-phase circuits, often called p-q theory. The paper begins by presenting the principle of active filtering and the basic instantaneous imaginary power theory. In the hypothesis of a distorted and/or unsymmetrical load voltage system, the p-q theory has proven limitations; consequently the paper reviews and evaluates other two reference current calculation methods. Finally a comparative analyze of the three methods features is carried out.