Using algorithms to improve the level of forecasting Xinjiang wind power forecasting accuracy is leading in China

In 2021, the staff of the State Grid Urumqi Power Supply Company conducted an electricity safety inspection on the wind farm in the Dabancheng area. Correspondent Wang Jinkui

Tianshan Network News (reporter Chen Rose reported) on February 14, the reporter learned from the State Grid Xinjiang Electric Power Co., Ltd. that more than 10 algorithms such as artificial intelligence (AI) in our region have applied more than 10 algorithms to predict new energy power, so that the ultra-short-term forecast accuracy of wind power has reached 96%, an increase of 4 percentage points year-on-year, exceeding the national standard and reaching the domestic leading level.

For a long time, the power generation of new energy has been unstable and difficult to predict, which has brought challenges to the construction of a new power system with new energy as the main body.

State Grid Xinjiang Electric Power Co., Ltd. continued to increase its scientific and technological innovation efforts, investing 2.2 million yuan to build a new generation of new energy power forecasting system. On September 17, 2021, Xinjiang's new generation of new energy power prediction system was officially launched, trying to apply AI and other technologies to dig deep into electrical operation data and wind and solar resource data. Through this system, you can grasp the output and power of new energy in the future period of time, which can effectively promote the consumption of new energy.

The new generation of new energy power prediction system can adopt different forecasting strategies for windy weather and strong wind weather according to the differences in Xinjiang's terrain, wind and solar resources, and power generation characteristics, and establish a more refined forecasting model to improve the accuracy of power prediction.

At present, 618 new energy field stations in Xinjiang have been connected to the new generation of new energy power prediction system, achieving full coverage of new energy field stations in Xinjiang. Nowadays, the ultra-short-term, short-term and medium-term forecast accuracy rate of wind power in Xinjiang has reached 96%, 92% and 81% respectively, and the forecast accuracy rate of photovoltaic power in each period has reached about 98%. Since the system was put into operation, the power grid has absorbed 860 million kWh of new energy electricity.

Li Guoqing, deputy director of the Hydropower and New Energy Division of the Power Dispatching and Control Center of the State Grid Xinjiang Electric Power Co., Ltd., said that the new generation of new energy power forecasting system that applies AI and other technologies is very powerful, which can not only provide accurate data support for power grid development planning and various power generation plans, but also scientifically optimize the maintenance plan of power transmission and transformation equipment and conventional power supply, expand the scale of electricity market-oriented transactions, and improve the level of new energy consumption in Xinjiang's power grid.