| Creators: |
Stillig, Javier and Parspour, Nejila and Ewert, Daniel and Jung, Thomas J. |
| Title: |
Feasibility Study on Machine Learning-based Method for Determining Self and Mutual Inductance |
| Item Type: |
Conference or Workshop Item |
| Event Title: |
(Proceedings of the) 2022 International Seminar on Intelligent Technology and Its Applications (ISITIA) |
| Event Location: |
Surabaya, Indonesia |
| Event Dates: |
July, 20-21, 2022 |
| Date: |
21 July 2022 |
| Divisions: |
Informationsmanagement |
| Abstract (ENG): |
This paper presents a method for inductance calculation of coils based on a machine learning algorithm. To show the feasibility of the approach, we generate a set of artificial training data describing a configuration of two planar spiral coils in varying dimensions and positions to each other in order to calculate their self- and mutual inductance. Afterwards, the data is used to train and evaluate three different machine learning models. Our evaluation shows that multiple linear regression with polynomial features reaches almost the same precision as the reference FASTHENRY2, but is orders of magnitude faster. With this novel machine learning based algorithm we enable new applications, where real-time prediction of inductances or coupling factors is advantageous |
| Forthcoming: |
No |
| Language: |
English |
| Uncontrolled Keywords: |
Machine Learning ; Inductance Calculation ; Wireless Power Transfer |
| Citation: |
Stillig, Javier and Parspour, Nejila and Ewert, Daniel and Jung, Thomas J.
(2022)
Feasibility Study on Machine Learning-based Method for Determining Self and Mutual Inductance.
In: (Proceedings of the) 2022 International Seminar on Intelligent Technology and Its Applications (ISITIA), July, 20-21, 2022, Surabaya, Indonesia.
ISBN 9781665460811
|
 |
View Item in edit mode (academic staff only) |
