Multiobjective optimization module (OP.CMO)

Description

Global, constrained, multi-objective and mixed variable optimization algorithm based on NSGA-II.

The algorithm can be used to optimize globally the electromagnetic and vibroacoustic behavior of an electrical machine.

Validations

Application to e-NVH

The multiphysic design of electrical machines requires to study different trade-offs such as torque per amp maximization, cost minimization, torque ripple minimization and acoustic noise minimization. These objectives are indeed often contradictory: low torque ripple does not mean low acoustic noise, and lower cost can mean thinner yoke and therefore higher vibration and noise.

Plot commands

See plot_OP commands such as:

How-to

The optimization can be applied to a MANATEE optimization project or to an external function to be optimized.

Read more