Rapid and accurate Type 2 charging ports detection for electric vehicles using YOLOv8
The efficiency and convenience of electric vehicles (EVs) are of the utmost importance as they become more prevalent. The vision system, responsible for precisely detecting charging ports to guarantee proper connector alignment, is a critical element of automatic charging stations. This research foc...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Proceeding Paper |
| Language: | English English |
| Published: |
IEEE
2024
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/115851/7/115851_Rapid%20and%20accurate%20Type%202_Scopus.pdf http://irep.iium.edu.my/115851/13/115851_%20Rapid%20and%20accurate%20Type%202.pdf http://irep.iium.edu.my/115851/ https://ieeexplore.ieee.org/document/10675522 https://doi.org/10.1109/ICSIMA62563.2024.10675522 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The efficiency and convenience of electric vehicles (EVs) are of the utmost importance as they become more prevalent. The vision system, responsible for precisely detecting charging ports to guarantee proper connector alignment, is a critical element of automatic charging stations. This research focuses on the necessity of a port detection model that is particularly effective for Type 2 charging ports, which are frequently employed in electric vehicles. The objective of the investigation is to improve the precision and efficiency of port detection, a critical step in the development of fully automated charging stations. We annotated the data on the Roboflow website and trained a model with the YOLOv8 algorithm on Google Colab using a dataset of 223 images. The research showed that the optimal performance was achieved with a train-test ratio of 70:30, resulting in a mean Average Precision (mAP) of 99.5%, precision of 99.5%, and recall of 100%. The model can detect Type 2 ports in real-time using a webcam with a less than one second detection speed. The results of this study suggest that the system can be effectively implemented in robotic arms for automated EV charging stations, thereby significantly improving the user experience by reducing the necessity for manual intervention. This research establishes the foundation for future automated electric vehicle charging technology developments. |
|---|