A fast and robust integral observer algorithm for fundamental and harmonics detection

Abstract

In recent decades, power quality is considered as an important factor since more electronic devices are connected to the power system. The electronic devices not only require high power quality but also generate a lot of pollution such as harmonic current back to the power system. Therefore, active power filters are designed to solve the problem. The active power filter requires a lot of control and the most important part is the reference current setpoint which is determined by the harmonic current detection algorithm. This thesis presents the integral modulation observer algorithm that offers a fast solution for the extraction of the fundamental component of a periodic distorted waveform. It is used as a very fast method to obtain the total harmonic current up to infinite order by subtracting the fundamental component from the distorted waveform. The method can be applied to both single and three-phase systems and requires the current signal for computation. The integral observer algorithm can be associated with a modulation function (MF). Previously, a polynomial function was used. In this new study, it is discovered that the robustness of the algorithm against noise can be significantly by using an exponential modulation. Comparative studies based on the use of several time-domain and frequency-domain methods have been conducted. It is found that the proposed integral observer algorithm associated with the exponential modulation function offers the best performance. This thesis provides the detailed analysis and performance of this new approach. The results have confirmed by both mathematical analysis and real-time experiments that the algorithm is one of the best solutions for the active power filter.