GA Optimized RLS–Based Shunt Active Power Filter for Power Quality Improvement under Nonlinear Load
Power quality degradation caused by nonlinear loads and extensive use of power electronic converters has become a critical concern in modern distribution systems, leading to excessive harmonic distortion, reactive power demand, and reduced system reliability. Shunt Active Power Filters (SAPFs) are widely recognized as effective solutions for mitigating these disturbances; however, their performance strongly depends on accurate harmonic estimation and optimal controller tuning under dynamic operating conditions. This paper proposes a Genetic Algorithm optimized Recursive Least Squares (GA-RLS) based SAPF for enhanced harmonic and reactive power compensation in a three-phase system. The RLS algorithm is employed for fast and precise extraction of fundamental and harmonic current components, while a Genetic Algorithm optimally tunes the RLS parameters to achieve faster convergence, reduced estimation error, and improved dynamic response. The proposed control scheme is implemented and evaluated using MATLAB/Simulink under nonlinear load conditions. Performance analysis is carried out through time-domain waveforms, FFT spectra, and Total Harmonic Distortion (THD) indices, and the results are compared with a conventional PQ-theory-based SAPF. Simulation results demonstrate significant improvement in power quality, with voltage THD reduced to 1.55% and current THD reduced to 6.49%, confirming the effectiveness and robustness of the proposed GA-RLS-SAPF strategy.