LS Electric said on the 25th that it developed technology to forecast renewable energy generation based on artificial intelligence (AI) and recently completed a demonstration. The move is seen as a strategy to accelerate its power brokerage business.
The "AI-based wind power generation forecasting model" was applied to a Jeju wind farm late last year and recorded 92% forecasting accuracy even in November, when weather volatility is high. LS Electric said this reduced the typical 10% forecasting error rate to 8% and corresponds to a level of accuracy at which power producers can receive incentives under the Korea Power Exchange's "renewable energy generation forecasting program."
LS Electric succeeded in precisely forecasting changes in wind power output through a "dual-forecast structure" that integrates and analyzes, using Machine Learning and Deep Learning algorithms, extensive weather information; regional characteristics such as terrain, altitude, and slope; and turbine-by-turbine performance deviations resulting from the fine characteristics of individual turbines.
Based on this technology, the company will step up its power brokerage business, including a "virtual power plant (VPP)." A VPP is a system that, using digital technology, integrates and operates diverse, physically dispersed renewable energy sources as if they were a single power plant. AI forecasts the output of distributed resources and optimizes supply strategies to offset the variability of renewable energy.
Renewable energy producers earn additional profit the lower their forecasting error rate is. When a renewable energy producer participates in a VPP, it forecasts solar and wind generation a day in advance, and if the next day's error rate versus actual generation meets the threshold, it receives a settlement payment. This is why technology that uses AI to improve forecasting accuracy is considered essential in the power brokerage business.
An LS Electric official said, "We expect a triple-win effect by resolving the intermittency of renewables with more advanced wind generation forecasting technology, contributing to the stabilization and efficiency of Korea's power grid, and maximizing renewable power producers' revenue," adding, "We will accelerate our VPP business by providing renewable energy producers with highly accurate forecasting technology and innovative solutions that are absolutely necessary."