三化螟时间序列预测模型研究

Studies on the Time Series of Rice Stem Borer

本文用时间序列分析方法对三化螟灯下蛾量的变动以及各代发生程度的变化规律进行研究,建立了3个预测模型,并比较了世代差分、季节差分和距平移动平均三种数据处理方法对模型的影响.结果表明,所建模型的回报验证符合率分别为72%,70%和65%;世代差分和季节差分两种数据处理方法较适合于本文的材料,而距平移动平均方法则拟合效果不太理想。

Three forecasting models of rice stem borer were developed with the analysis method of time series.The generation difference, seasonal difference and distance-moving average methods were used in the preparation of data.It is shown that the accuracy of these models is 65-72% and the results suggest that the generation dif- ference and seasonal difference are better than the distance-moving average in developing time series models of rice stem borer.