Air quality index prediction method based on combined model

• Main Authors: ZHANG HENGSHENG, GAO SONG, HE ZHUOJUN, LIU ZIYUE
• Format: Patent
• Language: Chinese; English
• Published: 14.03.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention discloses an air quality index prediction method based on a combined model, and provides a difference fusion seasonal prediction model DF-SPM based on RF and CLA, in the prediction process, an optimal threshold interval search algorithm is used in different seasons, the optimal threshold intervals of four seasons are learned and searched, and on the basis, the optimal threshold intervals of the four seasons are obtained. And a final prediction result is selected from the RF prediction value and the CLA prediction value. According to the model, seasonal periodic characteristics of the AQI are fully considered, a data set of each year is divided by taking four seasons as a time scale to search an optimal threshold interval, and threshold intervals of different time periods can be finely obtained so as to obtain higher prediction precision. The model uses a determination strategy of an optimal threshold interval to be superior to a single threshold strategy, the optimal solution search difficulty i