MULTI-CRITERIA OPTIMIZATION AND COMPARATIVE EVALUATION OF PASSIVE HARMONIC FILTERS USING ADVANCED META-HEURISTIC ALGORITHMS

Abstract

Harmonics, which negatively affect power quality, are among the primary problems caused by non-linear loads. These harmonics lead to distortions in voltage and current waveforms, excessive current and voltage surges, insulation failures, and malfunctions in power electronics-based equipment in industrial power systems. Therefore, the development of optimization-based harmonic filter design methods suitable for industrial applications is of great importance for improving power quality and ensuring system reliability. In this study, five different passive harmonic filter topologies (Single-Tuned Harmonic Filter, FirstOrder High-Pass Filter, Second-Order High-Pass Filter, Third-Order High-Pass Filter, and C-Type Filter) were optimized under a multi-criteria framework using four advanced meta-heuristic algorithms (Genetic Algorithm – GA, Particle Swarm Optimization – PSO, Differential Evolution – DE, and Grey Wolf Optimization – GWO). The optimization process simultaneously minimizes three fundamental performance criteria: average total harmonic distortion (THDI_mean), the IEEE-519 standard violation penalty, and reactive power consumption (QC). Results show that Single-Tuned filters optimized with PSO and DE algorithms achieved the lowest harmonic distortion (THDI_mean ≈ 0.577), while C-Type filters stood out with low reactive power requirement (~1.9×10⁴ VAr) and superiority in suppressing third harmonics. High-pass filters (1st–3rd order) performed worse across all algorithms. Multi-criteria optimization, simultaneously considering harmonic reduction, compliance with standards, and reactive power balance, offers more balanced and industrially applicable solutions than single-criterion approaches.