Fang, Chun2017-08-022017-08-022017http://hdl.handle.net/1993/32339Ferromagnetic characteristics and magnetic coupling are two crucial aspects when investigating controlled switching transient phenomenon of three-phase five-limb power transformers. Classic Jiles-Atherton (J-A) hysteresis model is popular in engineering applications but model parameters identification remains a difficult undertaking. This thesis explored novel techniques to improve overall accuracy and stability of inverse J-A model. A comprehensive process using a constrained multi-objective particle swarm optimization (PSO) was developed to extract and optimize parameters from published measurements of M3 grade grain-oriented electrical steel. A new methodology was proposed to define a variable that substitutes constant J-A model parameter which substantially improved its precision on minor and major hysteresis loops and magnetization curve simultaneously. This modified model was incorporated into a duality based hybrid transformer model in PSCAD/EMTDC with a computational efficient solution algorithm. Finally, an extensive systematic analysis was conducted to evaluate feasibility and performance of controlled switching three-phase five-limb parallel-operated transformers.engJiles-Atherton, Hysteresis, Controlled switching, HVDC, PSOmaster thesis