High-frequency modelling of variable-frequency drive systems based on physics-informed neural network

• Main Author: Raymond, Sammuel
• Other Authors: See Kye Yak
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering
• Online Access: https://hdl.handle.net/10356/176429
• Institution: Nanyang Technological University

Variable-frequency drive (VFD) systems are found in many operations around the world because they offer an outstanding efficiency when it comes to power conversion. Despite that, these systems may cause high-frequency phenomena such as overvoltage ringing within the motor which may lead to accelerated aging and even breakdowns. In order to mitigate or eradicate the effects, high frequency modeling of VFD systems have been used to alleviate the design process. This project aims to develop a physics-informed neural network (PINN) model that can precisely gain equivalent circuit parameters of VFD systems to accurately predict high-frequency phenomena in VFD systems by training the prediction model using the measured impedances of the system.