Comparative study of SARIMA and seasonal index model in predicting non-occupational carbon monoxide poisoning
- VernacularTitle:基于SARIMA和季节指数模型的非职业性一氧化碳中毒事件预测研究
- Author:
Wantong HAN
1
;
Yongqiang ZHANG
1
;
Shichang DU
1
;
Wei WANG
1
;
Kai QU
1
;
Xin HE
1
;
Cixian XU
1
;
Xiumei SUN
1
;
Qiran SUN
1
;
Jinyao ZHANG
1
;
Fan BU
1
;
Xingui SUN
1
Author Information
- Publication Type:Journal Article
- Keywords: SARIMA model; Seasonal index model; Non-occupational CO poisoning; Prediction
- From: Journal of Public Health and Preventive Medicine 2025;36(6):12-16
- CountryChina
- Language:Chinese
- Abstract: Objective To establish a prediction model for the occurrence of non-occupational carbon monoxide poisoning events in Beijing, and to provide scientific basis and theoretical support for the prevention and warning of poisoning events. Methods Based on the monitoring data of non-occupational carbon monoxide poisoning events in Beijing from 2016 to 2024, the seasonal ARIMA model and seasonal index model were established to analyze the data and predict the occurrence of events. Results Between 2016 and 2024, a total of 436 cases of non-occupational carbon monoxide poisoning were reported in Beijing, showing a downward trend. The established SARIMA model and seasonal index model were SARIMA (1,0,0) (1,1,0) 12, Yt = (-0.0339t+5.8863) × St, and the average relative errors were 65.42% and 29.19%, respectively. In terms of months, the SARIMA model had better predictive performance during April and summer (June to August), while the seasonal index model was superior in other months. By combining the two models, the predicted number of events in 2025 was as follows: 3, 2, 2, 3, 1, 5, 2, 7, 1, 1, 1, and 2. Conclusion The seasonal index model has the best prediction effect on the non-occupational carbon monoxide poisoning events in Beijing throughout the year, and the number of summer events predicted by SARIMA model is closer to the actual values. The two models can be combined to predict the trend of non-occupational carbon monoxide poisoning, which provides a scientific basis for the prevention and control of carbon monoxide poisoning in the future.
