时间序列分析模型在江苏省河蟹总产量预测中的应用

根据Box-Jenkins建模原理,采用ARI MAR(2,0,0)模型对江苏省河蟹总产量进行了拟合及预测研究。结果显示,所建模型的残差序列为白噪声,AIC为48.015 7,模型拟合值与实际值的相关系数为0.995 1,拟合度为0.990 1,表明组建的ARI MAR(2,0,0)模型结构合理,拟合效果好;用2006年江苏省河蟹总产量进行验证,相对误差为3.8%,表明ARI MAR(2,0,0)模型对江苏省河蟹总产量时间序列的变化趋势进行拟合和预测是可行的,是一种短期预测精度较高的预测模型。

Application of Time Series Analysis Model on Total Yield of Chinese Mitten-handed Crab in Jiangsu Province

According to Box-Jenkins theory and ARIMA(2,0,0) model,the time series of total yield characters of Chinese mitten-handed crab in Jiangsu province was identified,simulated and predicted.The results showed that after model identification,the residual error of ARIMA(2,0,0) model were all the white noise series,the AIC was 48.0157,the correlation coefficient was 0.9951,and the goodness of fittest was 0.9901.The total yield of Chinese mitten-handed crab in Jiangsu in 2006 was tested with this model and the relative error was 3.8%.This case shows that ARIMAR(2,0,0) model is a feasible and accurate method to predict the total yield of Chinese mitten-handed crab.