Skip to main content

JERA and JDSC develop system for predicting solar power generation

Published by , Deputy Editor
Energy Global,

JERA Co., Inc. and Japan Data Science Consortium Co. Ltd (JDSC) have jointly developed, and recently begun to operate, a highly accurate system for predicting solar power generation based on weather forecast data and past power generation performance.

Given the demand for energy decarbonisation as part of achieving a decarbonised society by 2050, solar power generation, which is technologically established and quick to construct, is seen as a promising short-term option for expanding renewable energy adoption in Japan. At the same time, since solar power generation performance varies with weather conditions, highly accurate predictions of the amount of electricity generated by solar power are needed in order to balance electricity supply and demand.

JERA and JDSC have jointly developed a highly accurate system for predicting the amount of electricity generated by solar power based on weather forecast data, and begun operating this system in conjunction with the start of operations at JERA's solar power generation facilities. The system is implemented in a serverless configuration and the automation of system operation, monitoring, model learning, and updating tasks significantly reduces its operational load. JERA and JDSC will further improve prediction accuracy by applying JDSC's expertise in advanced machine learning to the data gathered about the amount of electricity generated.

For more news and technical articles from the global renewable industry, read the latest issue of Energy Global magazine.

Energy Global's Autumn 2022 issue

The Autumn 2022 issue of Energy Global hosts an array of technical articles focusing on wave & tidal, waste-to-energy, energy storage, solar technology, and more. This issue also features a regional report outlining how green hydrogen is playing a key role in the renewable transition across Europe.

Read the article online at:

You might also like


Embed article link: (copy the HTML code below):