Electrical Infrastructure Monitoring: Case of NTDCL’s 500kV Network Insulator Detection With YoloV8
High voltage electrical infrastructure inspection requires condition monitoring of transmission line assets to avoid any possible failures or emergency. Detection of insulators in strings is linked with electrical infrastructure monitoring pertaining to the insulator fault classification. The datase...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Open Access Journal of Power and Energy |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11096616/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | High voltage electrical infrastructure inspection requires condition monitoring of transmission line assets to avoid any possible failures or emergency. Detection of insulators in strings is linked with electrical infrastructure monitoring pertaining to the insulator fault classification. The dataset widely available for insulator monitoring are either synthetic, lab created or publicly not available. In this paper, an indigenous dataset is created using Autonomous Aerial Vehicles (AAV) technology, capturing images in diverse topographical ambience across different transmission lines/circuits managed by National transmission and dispatch company ltd. in Pakistan. For detection of insulators in string, object detector model You Only Look Once-version 8 (YOLOv8n) is trained on created dataset of 3618 images, 603 being original and other augmented, after preprocessing and augmentation techniques were applied. The model’s performance is up to the mark with accuracy of 92%. The precision and recall being 0.95 and 0.90 respectively, whereas F1 score of the model peaked at 0.95 at confidence level of 0.652. |
|---|---|
| ISSN: | 2687-7910 |