Abstract |
The intensive use of power converters and
other non-linear loads in industry and by consumers in
general, has produced an increasing deterioration of the
power systems voltage waveforms. The presence of harmonic
distortion in power lines results in greater power losses in the
distribution systems, interference problem in communication
systems and, sometimes, in operation failures of electronics
equipments. Owing to these problems and to preserve the
power quality, the active filters are used for harmonic
elimination. In this filters, a perfect harmonic identification
is fundamental, therefore, nowadays many identification
algorithms are available, such as, DFT, Fourier Series, ect.
However, in this work a more robust algorithm based on
Kalman filtering is presented. Also, several grid voltage
models in state variable are shown, since the Kalman Filter
needs a model of the evaluated system (in this case, of the
grid). Moreover, the advantages of this method with respect
to the most used technique, FFT, are commented. |