ARIMA(p,d,q)与GM(1,1)模型在洪涝灾害预测中之比较研究——以广东省做实证分析

Comparative Research of ARIMA(p,d,q) and GM(1,1) Model in the Flood Disaster Risk Prediction——Making Empirical Analysis Using Guangdong Province

运用经济时间序列ARIMA(p,d,q)模型与GM(1,1)模型对广东省的洪涝灾害进行了预测.研究结论显示,所建立的ARIMA(p,d,q)模型能够较好的对洪涝灾害进行预测,模型可以运用到各种主要气象灾害风险预测与管理中,但所建立的GM(1,1)模型虽然具有很好的精度,但预测效果并不好.这充分体现了模型选择在预测过程中的重要性.

The paper uses economy time series ARIMA model （p, d, q） and GM （1, 1） model predict flood disaster of Guangdong province. Research results show that the ARIMA（p, d, q） model can be good for risk prediction of flood disaster, the model can be applied to risk prediction and management of main Meteorological disasters ,but the GM（1,1）model has good precision, but forecast effect is not good, That shows importance of the selecting prediction model process.