| KEYWORD |
Computational Intelligence for Electrical Machine design in automotive environment
keywords DATA MODELLING, E-MOBILITY, ELECTRICAL MACHINES, ELECTRIFICATION, MACHINE LEARNING
Reference persons MAURIZIO REPETTO, LUIGI SOLIMENE
Research Groups CADEMA
Description The electrification in the transport sector requires the design of electrical machines that must attain new requests for performance in terms of efficiency, high speed, torque and cost.
New machines design must cope with the interaction of many physical domains (e.g. electromagnetic, thermal, structural etc.). The research for new modelling solutions in this area is currently active and widespread.
While stakes are becoming higher, also the computational power available to this task is opportunely increasing. This allows innovative numerical methodologies, such as data-driven and machine learning approaches. The Innovative ways of thinking will widespread new design procedures.
Adoption of Machine Learning procedures can help the task of quick design and of sizing new machines and also to optimize their performance.
In addition, the reduction of the computational cost of the design procedure is a value as it allows to minimize the burden of computer analyses for the desired task.
In this respect, the thesis project aims at creating, by means of existing analysis procedures, a robust dataset of electrical machines performance and at the creation of data-driven techniques for its handling and exploitation in an optimization loop.
See also 2023_surrogate_compumag.pdf
Required skills basic programming
Deadline 20/12/2024
PROPONI LA TUA CANDIDATURA