用逐步回归分析模型预测水稻稻瘟病流行趋势

Epidemic trend prediction of rice blast by stepwise regression analysis model

水稻稻瘟病是我国水稻栽培中的重要病害,对该病害防治的关键是要做好预测工作。应用逐步回归分析,对影响水稻稻瘟病流行趋势的多种预报因子进行优选,择出优势因子并组建水稻稻瘟病流行趋势的逐步回归分析模型。验证结果表明:历史符合率达97.4%,3年预测结果的平均符合率也达到85.3%,符合率较高,为该病害的预测及其防治提供了依据。

Rice blast can cause the yield loss of rice in a large scale in China, the key to control it is forecasting this disease. By means of stepwise regression analysis, the authors selectead the optimum factors out of the weather factors influencing the epidemic trend of the rice blast. Then a stepwise regression analysis model to forecast the epidemic trend of rice blast was set up. Result showed that the accurate rate was 97.4%, and the accuracy of testing-forecast was also 85.3%. The accurate rate much high, the model can provide a foundation to forecast and control rice blast.