Electromagnetic transient (EMT) simulations play a vital role in the analysis and design of modern power systems and applications involving power-electronic converters, electrified transportation systems, and shipboard power systems. In such studies, synchronous machines are commonly represented using general-purpose lumped-parameter qd0 models. However, conventional machine models often rely on magnetically linear representations that neglect magnetic saturation and cross-saturation. Under heavily loaded operating conditions, these simplifications can lead to inaccurate predictions of machine characteristics, system impedance, and power-quality indices such as harmonic distortion. As a result, improving the accuracy and computational efficiency of synchronous machine models remains an important research topic. This thesis advances the modelling of synchronous machines by developing computationally efficient approaches to represent magnetic saturation and investigating their impact on system-level harmonic behaviour. Two saturation modelling approaches are introduced: a two-dimensional lookup-table-based explicit flux-correction model and a memory-compact single-polynomial saturation model capable of representing d-axis, q-axis, and cross-saturation effects. The proposed models are implemented and evaluated through extensive computer studies. The influence of saturation modelling fidelity on the harmonic performance of a cycloconverter fed shipboard propulsion power system is examined using detailed simulation models. Furthermore, harmonic mitigation strategies based on passive, active, and hybrid filtering approaches are investigated under realistic operating conditions of a representative Canadian Coast Guard Ship. The results demonstrate that an accurate representation of synchronous machine saturation improves the prediction of harmonic distortion and provides a more reliable basis for the harmonic-mitigation solutions in modern terrestrial power systems and electrified ships.
Shadman Saqlain Rahman (Thu,) studied this question.