Study On Application Of Corn Yield Grade Prediction Technology Based On Combination Algorithm

Thanks to the rapid development of computer technology,more and more production models have become more modern.They were not only widely applied to military and civil affairs,but also achieved great success in agriculture.Modern agriculture already changed the traditional agricultural production model,in addition it has become an effective way to increase the national economy and it also has unlimited potential for development.The application of computer technology to agricultural production and development is also increasing.Maize is one of the large-scale food crops grown in China,the yield of corn is largely determined by soil nutrients.In traditional agricultural production,,it is the farming experience that functions on the improvement of the soil nutrients and maize yield increase.Because many people do not understand the effects of soil nutrients on corn yield,the wrong fertilization resulted in a reduction in corn yield.The application of predictive modelling techniques to the effect of soil nutrients on corn yield grades not only allows fertilization and top dressing to be applied as early as possible,but also saves a lot of manpower and material resources.This paper uses a predictive model to analyze the effect of soil nutrients on corn yield grades,designs and implements a prediction system for corn yield grades,provides services through the Wechat Official Account.The main research content is as follows:1)Based on the Internet of Things technology,a monitoring station for farmland environment and crop growth was established in Helong Town and Kai’an Town,Nong’an County,Jilin Province.Simultaneously a system for collecting,storing,analyzing,and processing corn growth soils was established to create a data platform for soil nutrient data storage,accumulating material for later research.2)the interpolation method in data mining techniques are used to supplement the missing values and establish a soil nutrient database based on the results.3)According to soil nutrient,a combination forecasting model is established,such as: Bagging,Boosting,Stacking,and traditional models,such as: deep learning,logistic regression model,to predict the corn yield level.We copare each model and select the best model.4)Wechat Official Account is used as the carrier to communicate with customers,use Tencent Cloud to build our own dedicated servers,combine data mining technology and MySQL database,relying on the server platform to design and build a corn production level forecasting system.After storing,sorting,and processing the collected data,we constructed a corn production level prediction system through data mining,database,and system construction techniques.With the use of interpolation method to increase the integrity of the data,and use the AUC value as a criterion for discriminating the model to compare multiple prediction models,we select the Stacking model as the final model,finally build the server and Wechat Official Account as a communication carrier to provide customers with corn Grade prediction service.The aim is to predict the corn yield level through analysis of soil nutrients,and to enhance the real-time and accuracy of soil nutrient prediction on corn yield level,and to provide technical support for intelligent decision-making of corn production.