Energy IT platform company Enlighten has secured a new patent for technology that improves the accuracy of solar power generation forecasting. Enlighten said on Sept. 16 it registered a patent for an electronic device and method for predicting solar output.

The technology is designed to improve precision by using two different types of weather forecasts together. It takes inputs from a localized forecast that closely analyses a small area and a global forecast that covers a wider region at the same time.

This allows it to use more weather information for analysis than conventional methods that relied on a single forecast. Enlighten also applied an AI-based cross-attention method rather than simply combining the two datasets. The approach lets the two models refer to each other's data and selectively extract the strengths of each forecast.

Enlighten said prediction accuracy improves compared with using a single forecast, achieving an error rate of around 5 percent. Even if one forecast shows a large error due to abnormal weather, the other can compensate, enabling stable predictions. Solar output forecasting accuracy is seen as a core competitiveness factor for new energy businesses such as power brokerage, power purchase agreements (PPA), demand response (DR) and virtual power plants (VPP). It stressed that the larger the forecasting error, the more likely it is to lead to settlement losses and reduced grid stability, and that securing forecasting technology is directly linked to company profitability.

Enlighten provides integrated energy solutions including power plant construction and operation and power trading through its energy IT platform, Baljeonwang. It aims to integrate distributed resources scattered nationwide into a virtual power plant like Octopus Energy's Kraken. Enlighten is preparing for an initial public offering based on these capabilities and recently received an A grade in a preliminary technology assessment.

Enlighten CEO Young-ho Lee (이영호) said, "This patent is a technology that simultaneously secures prediction accuracy and stability by using weather forecasts with different characteristics in a complementary way." He added, "We will continue to advance our technology and prove our competitiveness in the global renewable energy market beyond Korea."