Dissertation Defence: Numerically Efficient and Accurate Modeling Strategies of Multilevel Multimodule Solid-State Transformer for Accelerated Electromagnetic Transient Simulation

Hengyu Li, supervised by Dr. Liwei Wang, will defend their dissertation titled “Numerically Efficient and Accurate Modelling Strategies of Multilevel Multimodule Solid-State Transformer for Accelerated Electromagnetic Transient Simulation” in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Electrical Engineering.

An abstract for Hengyu Li’s dissertation is included below.

Examinations are open to all members of the campus community as well as the general public. Registration is not required for in-person exams.

---

Abstract

Multilevel multimodule voltage source converter (VSC) emerges as an innovative enabler, contributing to integrate renewable energy sources and distributed generators into the modern power systems. Multilevel multimodule solid-state transformer (SST), also known as power electronic transformer, has become an advanced technology in applications of smart distribution systems, EV charging stations and renewable energy systems. Various electromagnetic transient (EMT) simulation tools are widely adopted for fast control prototyping of SSTs. However, conventional EMT simulation of the SST detailed model (DM) requires tremendous computational efforts due to detailed representation of massive switches and discrete circuit components. Thus, the simulation time step has to be sufficiently small to guarantee simulation accuracy at expense of simulation efficiency. The prior-art works represent the submodules (SMs) by Resistive Switch Model (RSM) or L/C-Associated Discrete Circuit (L/C-ADC). However, these models either feature time-variant conductance (G)-matrices or incur fictitious numerical oscillations.

This thesis focuses on developing numerically accurate and efficient modelling strategies i.e., switching-function-based detailed equivalent model (SFB-DEM) and switching-function-based average value model (SFB-AVM) for EMT simulation of the SST. Compared to the state-of-the-art equivalent modeling approaches, the proposed models realize constant G-matrix, circuit decoupling through DC-link capacitors, significant node reduction, and flexible representation of SMs’ deblocking and blocking modes. Simulation studies have demonstrated that the proposed equivalent models can significantly enhance the numerical efficiency, compared to the DM and variable G-matrix DEM (VG-DEM). Taking a three-stage SST with 60 SMs for example, the proposed SFB-DEM achieves speedup by 211 and 6 folds while the SFB-AVM achieves speedup by 2127 and 56 folds, compared to the DM and VG-DEM respectively. Additionally, it has been proved in the case study that the use of combined implicit-explicit multi-step solvers contributes to improve numerical accuracy of the SFB-DEM without incurring numerical oscillations, compared to the prior-art DEMs employing Trapezoidal Rule + Forward Euler methods. Meanwhile, the switching interpolation technique has been implemented and integrated with the multi-step solvers for accurate representation of intra-time-step switching events. Simulation studies have verified that it can effectively improve modeling fidelity of the SFB-DEM in particular when carrying out large-time-step simulation.