黄河流域棉花品种产量及其组分的长期预测及技术研究

Study on the Long Term Prediction and Technique of Yield and Its Components of Cotton Varieties in the Yellow River Valley

黄河流域棉花品种区域试验参试品种的皮棉产量、单株铃数、单铃重及衣分等资料为原始数据 ,应用谐波分析技术建立模型进行了拟合及预测研究。经拟合及精度检验 ,拟合指数均在 0 .90以上 ,历史符合率达 1 0 0 % ,模型为一级精度。可预测第 2 0轮区试参试品种的皮棉产量为 1 2 0 9kg· hm- 2 ,单株铃数为 1 5.1 6个 ,单铃重为 5.1 5g,衣分为 38.55%。

Based on lint yield, bolls/plant, single boll weight, lint percent of the varieties that joined the Yellow River Valley Regional Variety Test, according to the technique of harmonic analysis,the author has built a long term prediction model to predict the developing trend of the new varieties. As a result of agreement and accuracy examine,the agreement index is >0.90,the historical coincidence of the model reached 100%,and its accuracy reached first grade(good),respectively. Forecast the varieties that joined in the 20 round regional test, the lint yield, bolls/plant, single boll weight and lint percent reached 1209kg·hm 2 , 15.16 boll/plant, 5.15g and 38.55%,respectively. This study supplied a new method for long term forecasting of the developing trend of cotton varieties.