Intelligent Material Synthesis and Device Engineering for Next-Generation Power Electronics

초록

This paper investigates precise material synthesis control and optimised semiconductor device design for advancing power electronics. A soft sensor model combining slow feature analysis (SFA) with least squares support vector machine (LSSVM) is proposed for predicting cobalt oxalate average particle size, demonstrating improved accuracy over SFA-Neural Network methods. Additionally, technology computer-aided design (TCAD) simulations of β-Ga<inf>2</inf>O<inf>3</inf> Schottky Barrier Diodes (SBDs) analyse the impact of drift layer doping and thickness on performance characteristics. The effectiveness of a field plate (FP) termination is demonstrated, and its length is optimised considering the trade-off between breakdown voltage enhancement and high-frequency performance limitations due to parasitic capacitance. These studies highlight methodologies for enhancing material quality control and device performance in power electronic applications. © The Institution of Engineering & Technology 2025.

키워드