Performance improvement of harmonic detection algorithm in three phase three wire shunt active power filter under balance voltage condition

Abstract

Effective harmonic current identification is critical for shunt active power filters (SAPF) to provide accurate and sufficient compensation. This study proposes a modified synchronous reference frame fundamental (MSRFF) method for harmonic extraction in three-phase, three-wire systems. A band pass filter (BPF) was designed by combining low-pass and high-pass filters in the direct-quadrature (d-q) reference frame to improve filtering performance. Unlike traditional methods using phase-locked loops (PLL), this approach employs unit vector templates for synchronization and relies on direct current measurements from load currents. The band pass filter, with low cutoff frequencies, effectively isolates harmonic components in heavily contaminated systems, outperforming other filtering methods. System performance was evaluated using matrix laboratory (MATLAB) simulations, where total harmonic distortion (THD) values were reduced to 2.19% with a low pass filter, 0.99% with a conventional band pass filter, and 0.98% with the combined filter approach. The results demonstrate that the proposed strategy can accurately track and estimate harmonic signals, offering a robust solution for shunt active power filter applications.