Load forecasting is the key link of power system operation and dispatching, and correctly forecasting load is of great significance to ensure the safe and stable operation, optimal dispatching and economic operation of power grid. Commonly used load forecasting methods include statistics, artificial neural network and time series.

Statistical methods mainly rely on the statistical characteristics of historical data to predict the future load. By analyzing the periodicity, trend and seasonality of historical load data, the corresponding mathematical model is established to forecast.

Artificial neural network method is to train historical data through artificial neural network, so as to learn the internal relationship between data and use the learned knowledge to predict load.

Time series method is based on the principle of time series analysis, which regards load data as a time series data, models it with time series model, and forecasts the future load with the model.

In practical application, various methods are usually used for load forecasting to improve the accuracy and reliability of forecasting.