THE IMPACT OF AGGREGATION OF DOUBLY FED INDUCTION GENERATOR BASED WIND TURBINES ON TRANSIENT STABILITY

Abstract

Increasing wind power penetration challenges grid stability, yet simulating every turbine in a large wind farm individually is computationally prohibitive for transient stability studies. Existing aggregation techniques reduce this burden but are rarely evaluated alongside controller optimization, leaving open the question of whether aggregation can preserve the accuracy of transient stability while also improving computational efficiency. This study addresses that gap by developing an aggregated Doubly Fed Induction Generator (DFIG) wind farm model within the IEEE 39-bus benchmark system and tuning its grid-side, rotor-side, and DC-link PI controllers using Particle Swarm Optimization (PSO). The aggregated model's transient response was benchmarked against a fully independent (turbine-by-turbine) model under a range of symmetrical and unsymmetrical fault scenarios at Bus 6, using settling time, peak overshoot, and active/reactive power deviation as performance measures. PSO tuning improved settling time and peak overshoot by over 49% and 33%, respectively, relative to the unturned baseline. The aggregated model tracked the independent model's active and reactive power response to within 0.1% and 1%, respectively, while cutting simulation time by approximately 20%. These results show that coherency￾based aggregation, combined with PSO-tuned control, offers a computationally efficient alternative to full wind-farm simulation without sacrificing transient stability accuracy a finding with practical value for grid operators and researchers modelling large-scale wind integration