Wind power forecasting based on time series ARMA model

Destruction of the ecological environment such as global warming and poor atmospheric pollution has forced carbon emission reduction and renewable energy utilization to accelerate. Under this circumstance, wind power generation as a clean renewable energy source has been valued by many countries and has increased globally, and domestic wind power generation is also being developed and promoted on a large scale. In order to further study the intermittent and unstable wind power, this paper uses the autoregressive moving average (ARMA) model based on time series algorithm to predict short-term wind power. The ARMA model is established based on the time series of wind power historical data, and the data is subjected to ADF test to prove the rationality of the selected data. The prediction model is analyzed and verified by using the measured wind power data of a wind farm from 0:00 am to 23:45 on the 1st of the month. In this paper, the prediction error is analyzed by MAE, RMSE and MdAPE. The analysis results show that the proposed model has less prediction error. The wind power prediction method based on ARMA model can effectively predict the wind power.