Research on Non-Stationary Temperature Change Prediction Model Based on Improved Autoregressive Theory
Temperature changes typically exhibit non stationarity on a time scale,which poses challenges to the re-search of temperature prediction models.In order to effectively predict future non-stationary temperatures,a pre-diction model that can handle non-stationary temperature changes is proposed based on the theory of autoregressive models.The accuracy of the model is tested by selecting the average temperature data of February,May,August,and November from Station 57662 in Changde City from 1951 to 2020,and predicting the changes in average temperature of February,May,August,and November at the station over the next 15 years.The research results show that the improved autoregressive model has better fitting accuracy than traditional autoregressive models,with a maximum fitting coefficient of 0.96,proving the feasibility of the model in dealing with non-stationary tempera-ture change prediction problems.
Autoregressive theoryNon-stationary temperature changeNon-stationary sequencesPrediction modelAverage temperature
万方数据
维普
