Multi-objective optimization of passive filters in DG-integrated distribution networks utilizing dynamic EMT-based co-simulation

Abstract

The increasing penetration of distributed generation (DG) introduces variability in network impedance and harmonic emissions that can degrade the performance of passive harmonic filters designed for nominal operating conditions. This paper develops a reliability-oriented C-type filter synthesis framework that couples the Secretary Bird Optimization Algorithm (SBOA) with MATLAB–PSCAD electromagnetic transient (EMT) co-simulation. Candidate filter parameters are evaluated directly from simulated point-of-common-coupling (PCC) voltage waveforms under Monte Carlo realizations of grid impedance, background harmonic distortion, DG harmonic injection, and component tolerances. Probabilistic constraints are imposed on total and individual voltage harmonic distortion, while a conditional value-at-risk metric and lifecycle cost are considered as optimization objectives. The resulting design is compared with conventional and optimization-based benchmark filters under a common EMT-based evaluation framework. For the investigated medium-voltage distribution system, the optimized filter reduces PCC total voltage harmonic distortion (TVHD) to approximately 2.5%, with no limit violations observed in the considered Monte Carlo samples, while also improving power factor and reducing system and filter losses. The results indicate that incorporating EMT-based scenario evaluation into filter synthesis can improve harmonic-compliance margins under the uncertainties considered in this study.