AI Satellite-based Pyranometer

PHOTON is also equipped with a virtual pyranometer, designed with the state-of-the-art neural networks technology.  It can provide a reference irradiance, from the global satellite images and metadata, for PV plants requiring no (physical) pyranometers or further calibrating existing pyranometers. Moreover, the virtual pyranometer is able to trace data back years ago for any locations in the world, perfectly for evaluating solar plant performance.

 
螢幕快照 2019-05-31 下午4.39.40.png

Plant IRR Evaluation

Most of the time, there is no historical irradiance data. Take Taiwan as an example, only 30 weather stations have pyranometers. How can I get historical reference data for specific location?


 
螢幕快照 2019-05-31 下午4.40.03.png

Maintenance Required

It’s common to have unstable PR over 100% due to pyranometer soiling which make the most important benchmark useless. Regular maintenance work are required including cleaning and calibration.


 
螢幕快照 2019-05-31 下午4.39.52.png

Multiple Plant Assessment

The model and installation of pyranometers for different solar plants are usually different. It doesn’t mean the plant with PR 93% is better than 85%. If the calculated PR is not reliable for evaluate the performance of solar plants, what benchmark can we leverage?