Incorporating renewable energy in a well-balanced manner in addition to thermal power generation and nuclear power generation will enable stable power supply and demand while reducing the burden on the environment and costs.However, among renewable energies, photovoltaic power generation, which has been attracting particular attention in recent years, has the problem that the power generation efficiency varies greatly depending on the amount of clouds on that day.If the ratio of solar power generation is increased in the future, how to make up for this fluctuation will become a big issue.

　Therefore, it is important to predict the amount of solar power generation in advance and increase the amount of other power generation accordingly.The group used a technique called confidence interval prediction to enable such predictions.Prediction with confidence can be used to predict the expected amount of solar radiation and its reliability over time.In other words, it turns out that the amount of solar power generation can only generate this much at worst, and it is possible to predict in advance the amount that will need to be supplied from other power sources.

　Prediction interval with reliability is a method that is becoming the mainstream in the world in forecasting power generation of renewable energy.In the future, in addition to these predicted values, we aim to build a system that derives the optimum operation method for the entire power supply system while taking into account the cost of starting and stopping thermal power generation.