| Abstract |
The component-model-creation process for multi-objective optimization tools in power electronic applications has been rarely investigated in literature. In this work, a new approach to model creation is shown, which includes public internet information and a web scraping technique. Web scraping is known from data science. It is a possibility to extract data from websites in real time and to save information into a desired data format. This publication first introduces the basics of web scraping and shows how it can be applied to the generation of power electronic component models. The "PFC choke" is used as an example in this paper. The component cost models, the size and efficiency models for two different choke series are generated using web scraping. It is discussed whether it makes sense to create a model for both series or whether one model should be created for each series. Finally, all models are parameterized so that they could be integrated directly into a virtual prototyping routine. The approach can be used universally for every power electronic component. |
