Designing electric traction motors involves simultaneously optimizing multiple objectives across electromagnetic, thermal, and mechanical domains. As the number of objectives grows, Pareto-based methods become less effective due to the increase in non-dominated solutions. This work investigates how the number and distribution of non-dominated points vary with the number of objectives, using a dataset of 16000 motor configurations. Results show that many-objective settings (more than four) pose challenges for Pareto ranking. Building on these insights, many-objective optimization techniques such as NSGA-III and game-theory-based aggregation will be investigated and compared, using surrogate models trained on this dataset to reduce the computational effort.

Many-objective optimization approaches for the multi-physics electric motor design problem / Lorenti, Gianmarco; Solimene, Luigi; Repetto, Maurizio. - (2025). (Intervento presentato al convegno 18th International Workshop on Optimization and Inverse Problems in Electromagnetism (OIPE)).

Many-objective optimization approaches for the multi-physics electric motor design problem

Lorenti, Gianmarco;Solimene, Luigi;Repetto, Maurizio
2025

Abstract

Designing electric traction motors involves simultaneously optimizing multiple objectives across electromagnetic, thermal, and mechanical domains. As the number of objectives grows, Pareto-based methods become less effective due to the increase in non-dominated solutions. This work investigates how the number and distribution of non-dominated points vary with the number of objectives, using a dataset of 16000 motor configurations. Results show that many-objective settings (more than four) pose challenges for Pareto ranking. Building on these insights, many-objective optimization techniques such as NSGA-III and game-theory-based aggregation will be investigated and compared, using surrogate models trained on this dataset to reduce the computational effort.
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/3003548
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo