WP4 – Optimization Modules

The work package is devoted to the creation of novel CAE technologies for the MOTOR Optimization Modules. A hybridized optimization strategy based on a hierarchy in the parameterization is developed that combines global and local optimization algorithms in order to efficiently tackle the curse of dimensionality. This makes it possible to combine the advanced computational tools from WP 2 Geometry Modules and WP 3 Simulation Modules into fully-automated design optimization workflows.

Objectives

  • To combine gradient-based detailed local searches based on a high number of degrees of freedom with gradient-free global searches on a reduced number of parameters in a hybridized optimization approach
  • To perform adjoint-based optimization on the CAD geometry
  • To determine the most significant parameters dynamically

Description of work

  • MM 4.1: Hybrid multi-level optimization method
  • MM 4.2: CAD-based adjoint optimization method
  • Task 4.3: Development of dynamic strategy of determining most significant parameters

Partners involved

The IVKDF is work package leader in WP4.
Contact info WP4 Leader: Herman Deconinck

The other partners actively involved in WP4 are: 

 

Print Friendly, PDF & Email