[1L3-GS-13] AI application: Social application (1)

[1L3-GS-13-04] A Study on Imbalance Discriminant Model to Select Locations of Wind Power Generators

〇Eriko Sato¹, Yaemi Teramoto¹, Toshihiro Kujirai¹ (1. Hitachi, Ltd., Research & Development Group)

Keywords:Machine learning, Wind power generator, Weather forecast, Prediction of power generation

Recently, the imbalance fee is imposed to a power utility as a penalty according to the amount of imbalance, which is the difference between the actual and the scheduled power generation based on the prediction of its, to keep the power grid stable. Especially, it is difficult for wind power utilities to predict the power generation depending on climate. So, in the near future, the revenue of wind power utilities will decrease when the high imbalance fee is imposed. Therefore, we focused on a method to select locations with low prediction errors to set up the wind power generator, and developed the models to evaluate operating revenue. These models are composed of a wind power estimation model at any points, a power conversion model, and an imbalance discriminant model whether high or low imbalance. We discussed the locations to set up the wind power generator using these models in the view point of operating revenue.