Research And Application Of Wheat Head Blight Forecasting Based On Support Vector Machine

Wheat head blight is a typical meteorological disaster with strong explosive and destructive characteristics.It seriously affects the yield and quality of wheat.Wheat head blight can be prevented from being curable.Researchers have never stopped studying the control of Fusarium head blight.However,studies on the prediction models of wheat head blight often need a large number of parameters to support it.Compared with the actual data sample size,business demand and prediction results,this dissertation studies the prediction method of wheat head blight by using the excellent small sample optimization ability and generalization ability of support vector machine,which enriches the expression pattern of the prediction of wheat head blight.The main work of this dissertation is as follows:1.The extraction and pretreatment of wheat head blight inducing factor were completed.First of all,the characteristics and causes of wheat head blight were analyzed.The meteorological factors that may be pathogenic during the onset of wheat head blight were collected,and data mining technology was used to obtain the most important factors which contributed to wheat head blight,and proved by calculation.The daily mean temperature,daily average humidity and the number of days of continuous rainfall were extracted to form the eigenvalue matrix.2.A prediction model based on support vector machine(SVM)for wheat head blight was established.The support vector machine model was used to train the characteristic value matrix of the above meteorological samples and the rate of disease ear.The grid search method and cross validation were used to determine the kernel function and penalty factors in the support vector machine,and the experiment on the prediction of the severity of wheat head blight was completed.The results showed that the Wheat Fusarium head blight prediction model based on support vector machine had higher accuracy and could be used for the prediction of wheat head blight.3.A wheat head blight prediction system based on support vector machine(SVM)was developed.The system is based on the B/S architecture,mainly including three modules,the B/S browser client implements the parameter input and the interaction with the server side.The server side is based on Python and Anaconda to classify and predict the sample set of wheat head blight disease by support vector machines,and feedback the results tothe browser client;the database stores the back end system.The result of counting and classifying.