3D Design of Magnetic Components by Gaussian Kernel Regression -Forward Design and Inverse Design-

Yuki Sato, Magnetic Technology Group, Kilby Labs, Texas Instruments Japan Limited


This talk introduces a 3D modeling method of magnetic components with Gaussian kernel regression which is one of the machine leaning techniques. The modeling method is mainly subdivided by two methods: forward modeling and inverse modeling. In the forward modeling, the optimization techniques(GA and Newton method etc.) found optimal solution with changing the design parameters. On the other hand, design parameters are directly extracted from targeted electric parameters. The former method is often employed to design magnetic components using finite element method while it is difficult to directly apply the later method to finite element equation so a surrogated model is employed. In this talk, the both modeling methods using Gaussian kernel regression are introduced with 3D inductor model.

To read Proceedings, please sign-in.

Protected content here, for members only.
You need to sign in as a Regular JMAG Software User (paid user) or JMAG WEB MEMBER (free membership).

By registering as a JMAG WEB MEMBER, you can browse technical materials and other member-only contents for free.
If you are not registered, click the “Create an Account” button.

Create an Account Sign in 

Remember me
Sign In
Create an account (Free) About authentication ID for JMAG website

Search Filter

  • All Categories

Proceedings Archives