Research And Application Of Peanut Yield Prediction Based On Combinatorial Model

China is a big country in peanut cultivation.From 2011 to 2021,the peanut output increased from 16.2 million tons to 18.2 million tons,with a growth rate of12%.However,with the increase of peanut output and planting area,diseases,insect pests,weather and other factors have a greater impact on the peanut industry,making it difficult to grasp the trend of peanut output.It is very important to analyze the characteristics of peanut yield fluctuation and forecast trend because of the difficulty of prediction and low accuracy caused by the strong fluctuation of peanut yield.This thesis takes the annual output of peanut of Luanxian Baixin Peanut Planting Professional Cooperative as the research object,and the main work is as follows.Firstly,the rule of peanut yield and the influencing factors of peanut yield were studied.Based on the analysis of the growth environment and fluctuations in groundnut production factors,on the basis of selecting the maximum temperature,average temperature,minimum temperature,relative humidity,light intensity,rainfall,the planting area,the effective irrigation area,soil fertilizer,applying pesticide content,machinery total power for correlation analysis,from the perspective of quantitative analysis of the affecting factor on the changes of peanut production,Combined with the correlation analysis results,it was concluded that environmental factors and operational factors were significantly related to peanut yield.Secondly,based on the analysis of 11 influencing factors of peanut environment and production operation,multiple linear regression prediction model,BP neural network prediction model,random forest prediction model and their combination prediction model were constructed respectively,and then the characteristics and practical application advantages of each model were compared and analyzed according to the experimental results.The results show that the combined model has the characteristics of high stability,high accuracy and small probability error of the other three models,and improves the prediction accuracy.Finally,through field research and demand analysis for cooperative peanut production,peanut production design visualization service system’s overall architecture and relevant function module,the model algorithm and data integration is to small peanut yield prediction program,the existing data as support,application We Chat small programs,and systems development and visualization technology,peanut yield visualization service system,More intuitive and accurate display of peanut yield over the years change rules and trends,for practitioners to provide convenient and real-time services.Model based on combination of peanut yield prediction research use combined model prediction algorithm provide a scientific theory basis for peanut yield prediction,small programs meet the demand of cooperative application systems development and management,improve the enthusiasm of cooperative production efficiency and the growers in the realization of the peanut yield accurate forecasts and promote the development of peanut industry is of great significance.