en.Wedoany.com Reported - On July 16, Yalong River Company of China's SDIC Group officially released a smart operation large model for hydro-wind-solar integration in the Yalong River Basin, Sichuan. The system has been deployed at the Yalong River Basin Dispatch Center, integrating real-time data on hydrology, meteorology, wind power, and photovoltaic (PV) power, covering resource forecasting, power generation planning, production dispatch, equipment operation and maintenance, and electricity trading. It is China's first smart operation large model designed for large-scale hydro-wind-solar integrated energy bases.
This system establishes a multi-source data foundation covering the entire Yalong River Basin, integrating satellite remote sensing data, ground meteorological monitoring station data, basin hydrological station data, and sensor data from new energy stations into a unified intelligent forecasting system. Data from hydropower stations, wind farms, and PV power stations, which were previously relatively scattered, now form a unified data chain, providing a basis for joint forecasting across different regions, energy types, and time scales.
The large model streamlines the data flow between meteorological forecasting, hydrological forecasting, wind and solar resource forecasting, and power operation decision-making. Based on rainfall, water inflow, wind speed, solar irradiance, and real-time station operating status, the system can jointly assess hydropower inflow, wind power output, and PV output, and use the forecast results for power generation scheduling and cascade hydropower station dispatch in the basin. The large screen at the Yalong River Basin Dispatch Center can now simultaneously display forecast information on basin-wide hydrology, meteorology, wind power, and PV power.
In terms of water inflow forecasting, traditional models can typically provide accurate predictions for only about ten days. The newly released large model extends the basin water inflow forecast period to 60 days. Among these, the first 10 days can provide hourly-level forecasts with an accuracy improvement of approximately 5%; from day 11 to day 60, medium- to long-term daily-scale forecasts are available. The response time for meteorological forecasting has also been reduced from hours to minutes, allowing for longer preparation time for reservoir storage, unit power generation, and new energy consumption scheduling.
The Yalong River Basin features significant terrain variations, with some new energy stations located in high-altitude areas. Weather changes, local climate, and hydrological conditions significantly impact power generation capacity. Based on training with massive real-world data, the system has been specially optimized for plateau complex weather and hydro-wind-solar coordinated operation scenarios, focusing on solving issues such as the difficulty of unifying different energy data, inconsistent forecast periods, and the disconnect between forecast results and production dispatch.
This large model does not merely provide weather or power generation forecasts; it extends forecast results to the actual operational aspects of the energy base. The system covers the entire business chain of power generation, dispatch, operation and maintenance, and trading, providing data support for hydropower unit start/stop, wind and solar output scheduling, equipment inspection, electricity trading plans, and extreme weather response. This enables the regulation capacity of hydropower in the basin to be synergistically matched with the fluctuating characteristics of wind power and PV power generation.
Currently, the Yalong River Basin hydro-wind-solar integrated base has an installed clean energy capacity of nearly 23 GW in operation, with approximately 12 GW under construction. As hydropower, pumped storage, wind power, and PV projects within the base continue to increase, the smart operation large model will undertake larger-scale resource forecasting and coordinated dispatch tasks.
