EAI Endorsed Transactions on Energy Web (Apr 2024)

Image Recognition of Photovoltaic Cell Occlusion Based on Subpixel Matching

  • Yuexin Jin,
  • Jinchi Yu,
  • Xiaoju Yin,
  • Yuxin Wang

DOI
https://doi.org/10.4108/ew.5751
Journal volume & issue
Vol. 11

Abstract

Read online

INTRODUCTION: During the operation of large photovoltaic power stations, they are often shielded by dust and bird droppings, which greatly reduce the power generation and even cause fires. Analysis of PV cell occlusion image recognition accuracy based on sub-pixel matching. OBJECTIVES: In order to find the location of the pv cells, we use the method of subpixel image matching. Improve recognition accuracy. METHODS: When the power plant is running normally, taken the original image for photovoltaic power station as the original sample, and then using the subpixel gradient matching algorithm, to match the original image and find out that the minimum matching values. RESULTS: If the calculation results is greater than a specified threshold, When the calculated result is greater than the specified threshold, the power station is considered abnormal. CONCLUSION: The experimental process shows that this method can better judge the operating status of photovoltaic power station, and can find out the location of mismatched photovoltaic cells more accurately, and the calculation accuracy reaches sub-pixel level.

Keywords