基于相空间重构和支持向量机的小麦条锈病预测方法

The Method of Wheat Stripe Rust Predicting Base on Phase Space Reconstruction and Support Vector Machine

摘要:

小麦条锈病预测对小麦灾害的预防和控制具有十分重要的意义.本文使用相空间重构和最小二乘支持向量机相结合的方法对成都市小麦条锈病发病率进行预测.实验中使用成都地区的气象数据作为模型的输入参数，发病率作为模型的输出参数，对20例发病病例样本进行预测，得到了比较好的预测效果.实验中还与简单最小二乘支持向量机预测结果作了对比，结果表明，相空间重构和最小二乘支持向量机结合的模型预测准确率明显高于简单最小二乘支持向量机，因此此模型对小麦条锈病预测是十分有效的.

Abstract:

Wheat stripe rust prediction has very vital significance for the prevention and control of wheat disaster.This paper forecast the wheat stripe rust incidence in Chengdu using PSR and LSSVM.The Meteorological data was as input parameters, and the incidence rate was as the output parameter in the test. To predict the 20 cases samples ,this paper got better predictive effect.The method in this paper is much better than SVM, through the comparative analysis.So,the model Is very effective to For wheat stripe rust prediction.