Computer-Aided Comprehensive Operation Analysis of DC-DC Converters Considering Uncertain States of Diodes

From Graph Theory to Graph Neural Networks (GNNs): The Opportunities of GNNs in Power Electronics

“…(ii) Converter-level: Leverage the explicit expression power, graph-theory-based modeling and analysis has got some considerable attention, including equivalent circuit identification [3], multilevel converter topologies with parallel connectivity [5], open-circuit fault diagnosis for Ttype inverter [15], operation analysis of DC-DC converters considering uncertain states of diodes [33], sneak circuit analysis of semi-DAB [34], fault diagnosis of modular power supply [35], etc. In terms of converter relationship research, isomorphic converter derivation and operation design is proposed in [21], revealing some unique yet broadly existing interconnections between voltage-source and current-source converters.…”

“…Serving as a common language among various disciplines, graph theory can be leveraged as a powerful tool to enable systematic modelling and analysis [3][4][5][6][7], design new converters [8][9][10][11][12], control their operations [13,14], estimate and identify potential issues [15][16][17], optimize the systems [18][19][20], or even facilitate the understanding of the interconnections and interactions between components in power-electronics-based systems [21][22][23]. Especially in recent years, innovative research keeps emerging, covering component-level, converter-level and system-level of power electronics [24][25][26][27][28][29][30][31][32][33][34][35][36][37]. Leveraging the power of graph, these research works not only feature systematic understanding but also open more possibilities, e.g., integrated with automation and AI.…”