Long-term soil temperature and humidity high-precision prediction method based on autoregressive moving average model

The invention discloses a long-term soil temperature and humidity high-precision prediction method based on an autoregressive moving average model. Firstly, data preprocessing is carried out on historical soil humidity and temperature data; the order of an ARMA model is determined through an AIC criterion; and fitting is conducted on the ARMA model by using historical data of soil humidity and temperature to obtain a determined ARMA model, the future soil humidity and temperature are predicted through the ARMA model, and the irrigation water amount of a future long-term intelligent irrigationsystem is determined in combination with a predicted value of the weather bureau for the future weather. The ARMA model for predicting the soil humidity and temperature is also continuously adjusted and optimized, so that the effectiveness of the ARMA model is ensured, and the ARMA model has an actual guidance value for the soil humidity and temperature which are predicted for a long time; meanwhile, the reliability of th