Naslov
Extraction of Transformer Saturation Curve from Ferroresonance Measurements Based on Nelder Mead Optimization Method

Autori
Milardić, Viktor ; Tokić, Amir ; Uglešić, Ivo ; Xemard, Alain

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
IPST 2023 / - , 2023, 1-7

Skup
IPST 2023

Mjesto i datum
Solun, Grčka, 11.06.2023. - 15.06.2023

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Saturation curve, ferroresonance, Backward Differentiation Method, Ordinary Differential Equation System, Nelder-Mead optimization method, transformer modelling

Sažetak
An accurate representation of the transformer saturation curve is essential for calculations of low frequency transients (inrush current, ferroresonance, load rejection) and steady states (power quality problems, harmonics and subharmonics). This paper presents an original extraction methodology for determining the transformer saturation curve from the measured ferroresonant current and voltage waveforms. The proposed method is based on the formulation of a novel multiobjective function, using current and voltage obtained from the measurement. The proposed simulation-optimization method is based on the Backward Differentiation Method for solving the very stiff differential equation system that describes the ferroresonance as well as the Nelder-Mead optimization method used to minimize the proposed multiobjective function. Five functions are proposed for approximating the saturation curve. The best ferroresonance simulation results are obtained with the inverse extended Frolich function. The validity of the proposed methodology has been confirmed by a very good agreement between the measured and simulated results of the ferroresonant current and voltage under different ferroresonance scenarios (three different values of series capacitances).

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika

POVEZANOST RADA