Deep Learning-Based Prediction of Key Performance Indicators for Electrical Machines

The design of an electrical machine can be quantified and evaluated by Key Performance Indicators (KPIs) such as maximum torque, critical field strength, costs active parts, sound power, etc. Generally, cross-domain tool-chains are used to optimize all the KPIs from different domains (multi-objective optimization) varying given input parameters in largest possible space. This optimization process involves magneto-static finite element simulation obtain these decisive KPIs. It makes whole a vehemently time-consuming computational task that counts on availability resources with involvement high cost. In this paper, data-aided, deep learning-based meta-model is employed predict quickly accuracy accelerate full reduce its costs. focus analyzing various forms data serve geometry representation machine. Namely, cross-section image allows very general description relating topologies classical way scalar parametrization geometry. impact resolution studied detail. results show prediction proof validity minimize time. also indicate quality image-based approach made comparable based parameters.