The medium-term forecast time scale of wind power is several weeks or months, and the fluctuations of wind power during this time scale are related to the plan maintenance of the wind farm or the electricity network.
1. Wind power forecast classification
There are many classification methods for wind power forecasting. Generally speaking, there are the following classification methods:
1. According to the predicted physical quantities, they can be divided into: wind speed predicted output power and direct predicted output power.
2. According to mathematical models, it can be divided into: continuous forecasting, time series model forecasting, Kalman filter method and intelligent neural network method forecasting.
3. According to the input data, it can be divided into: do not use the method ofnumerical weather prediction and use the numerical weather prediction method.
4. According to the temporal dimension, it can be divided into: very short-term forecasting, short-term forecasting and medium and long-term forecasting. Among them, the classification by temporal magnitude is generally recognized by all and is the most used.
Among them, the very short term forecast, the short term forecast and the medium and long term forecast:
1. The very short-term time scale of wind energy. the power forecast is 0 to 4 hours, a rolling forecast is 15 minutes, and the time resolution is 15 minutes. Mainly used for real-time dispatching to solve power grid frequency regulation problems.
2. The short-term wind power forecast time scale is 0-72 hours, and the temporal resolution is 15 minutes. It is mainly used forreasonably organize the power generation plan of conventional units and solve the problem of. peak regulation of the electricity network.
3. The medium- and long-term wind power forecasting time scale is weeks or months. Fluctuations in wind energy over this time scale are linked to the maintenance plan of the wind farm or power grid.
2. Wind power forecasting methods
Wind power forecasting methods can be divided into: one method is to use physical methods to calculate the power output of the wind farm based on conditions digital weather. forecast data; the other is to calculate the power output of the wind farm using physical methods; a method is a statistical forecasting method based on the relationship between numerical weather prediction, energy production of the eo parklink and measured data online.
The global method refers to a method that uses both physical methods and statistical methods.