Multi-objective Optimization

Description

Manatee e-NVH software comes with a multi-objective optimization environment of e-machine magnetic circuit design variables. Default solver is NSGA-II genetic algorithm. Design variables include geometrical parameters from magnetic circuit templates (such as slot shape, notch shape, magnet pocket shape, skew rate), as well as winding design (coil pitch), rotor skew, or current harmonics.

Overall noise, vibration, force or torque levels, as well as specific harmonics can be used as objectives to be minimized or maximized. These variables can also be used to defined constraints or be simply tracked during the simulation.

Dedicated post-processings are automatically generated at the end of the optimization to extract the Pareto front, study the convergence rate of the algorithm or explore the final design space. It is also possible to extract a given design and run more detailed simulations on it.

Application

The optimization environment is designed to help Electrical and NVH Engineers improving the multi-physic performances of a given design, especially in terms of noise and vibrations and electromagnetics.

Demonstration