Using ChatGPT for the Knowledge and Design of Power Electronics Converters

The paper reveals the potential applications of the ChatGPT in power electronics. By analyzing academic articles and other relevant materials, ChatGPT can efficiently summarize key concepts and then apply knowledge in power electronics design. To highlight its uses in the autonomous design application, two customized generative pre-trained transformers (GPTs) are developed with a retrieval-augmented generation (RAG) framework to select the optimal topology and design the system's parameters with the predefined specifications. Furthermore, the coding ability is assessed to generate the executable codes within a Simulink function block, realizing the maximum power point tracking (MPPT) algorithm for a boost converter in a photovoltaic application scenario. The effectiveness of the designed function is verified through the simulation waveforms. Finally, the paper discusses ChatGPT's advanced ability for textual analyses. After giving a published paper, some challenges with corresponding solutions are proposed by the chatbot. Through these demonstrations, this paper offers new sights to revolutionize design methodologies.