Optimal Design of Electric Motors for Automotive Applications

The paper describes how, exploiting AI, it is possible to design electric motors for automotive applications. Both traction motors and motors for auxiliary functions are dealt with. Given the requested performance of the motor (objective functions) and the constraints, the design variables defining the motor are derived by means of a multi-objective programming approach. Usually tenth of either objective functions or design variables are considered. Aspects related both to electromagnetic and mechanical performance are taken into account, in a multi-physics framework. The issues referring to thermal, structural and noise-vibration-harshness are considered for defining the Pareto-optimal sets both in the design variable domain and in the objective function domain. Such domains can be found by either supervised learning or reinforced learning, two well-known AI algorithms. Basic constraints related to manufacturing are included in the optimization process. A couple of examples are produced to show how electric motors can be optimally designed.