# Deep Neural Network-Based Design of Planar Coils for Proximity Sensing Applications

## TL;DR

This paper introduces a deep learning method to design planar coils that produce specific magnetic fields, improving efficiency and performance for various applications.

## Contribution

A novel deep learning approach for designing planar coils based on desired magnetic field maps is introduced.

## Key findings

- The method generates coil designs that closely match target magnetic fields with high accuracy.
- It enables simpler coil structures and improved performance for proximity sensing and wireless power transfer.
- The approach is applicable to inductive sensors and electromagnetic compatibility systems.

## Abstract

This study develops a deep learning procedure able to identify a planar coil geometry, given the desired magnetic field map. This approach demonstrates its capability to discover suitable coil designs that produce desired field characteristics with high accuracy and efficiency. The generated coils show strong agreement with target magnetic fields, enabling manufacturers to achieve simpler structures and improved performance. This method is suitable for inductive proximity sensors, wireless power transfer systems, and electromagnetic compatibility applications, offering a powerful and flexible tool for advanced planar coil design.