Meidensha Corporation (Meiden) is pleased to announce that it has been chosen as a consignee in a Nagano prefectural government public selection procedure for a demonstration project to help hydropower plants optimally operate with the use of artificial intelligence (AI).

In the project, officially called the “Fiscal 2021 Demonstration Project on AI-based System to Support Hydropower Plant Operation Plans,” Meiden will establish a computer model to predict dam inflow volumes and a system to support hydropower plants operate with high efficiency. The project will take place from August 2021 to February 2022.

The Nagano prefectural government’s Bureau of Business Affairs created a center in April 2021 to promote the smartification of power plants and a shift to the centralized control of such facilities. In particular, hydropower plants are increasingly pressed to make operational plans based on accurate dam inflow volumes at a time when our planet is hit by increasingly severe and frequent water-related disasters as a result of climate change. AI-assisted operational plans will replace conventional plans made based on skilled workers’ judgements and experiences. As part of its efforts to promote smart industrial safety, the bureau decided to verify the feasibility of introducing such a system, which became possible due to the high accuracy of weather forecasting and AI-based dam inflow volume predictions which have improved for practical applications.

Project outline
1.Establishment of a dam inflow prediction model
Up to now, on-site workers adjusted the water level of a dam based on weather forecasts and their own experience. Under the dam inflow prediction model, on the other hand, the system will learn the dam inflow volume and past weather data and then predict the volume with high accuracy more than 72 hours in advance based on weather forecasts. The model can be used for many purposes as it is solely based on machine learning.

2.Establishment of a system to support power plant operational plans
The system will compile optimal power generation plans based on dam inflow volumes predicted by AI. The system automatically makes power generation plans by considering certain attributes involving power generation at dams and changes in the water level and volume in addition to conditions restraining the operation of generators and dams. Therefore, it can reduce the workloads of on-site staff. By formulizing these attributes and restrictive conditions, a power plant will be able to optimize its power volume it generates under its operational restrictions. Using the system is expected to improve its profitability, compared with that under the conventional power generation plans.




Meiden will contribute to realizing a sustainable society by promoting smart industrial safety through the provision of products and services which help power plants attain optimal power generation. 

Press contacts:

Corporate Communications & PR Promotion Division
e-mail: kouhou@mb.meidensha.co.jp