Research And System Implementation Of Crop Growth Model Based On Boosting Tree

With the rapid development of the global environment and industrialization and urbanization,land resources are generally faced with the problem of degradation,waste and defacement,which restricts the sustainable development of agriculture.Land remediation project is an important means to solve this problem,and the decision-making process of crop selection is especially important as the core part of the land improvement project.Crop selection process is to analyze the impact of various environmental factors on crop yields so that crops with higher yields can be planted according to specific land conditions.Therefore,crop yield prediction is the main problem in this thesis.The main content of this thesis is divided into two parts.The first part is the research of crop yield prediction model.Firstly,this thesis analyzes the existing problems of crop yield forecasting model.Then,the solution based on the integrated learning method is proposed,and the integrated learning method is optimized.Besides,on the basis of the above analysis,the limitation of the existence of the regression tree is analyzed,and the improved optimization scheme under the background of crop yield prediction is put forward.Finally,the proposed scheme is verified by the related experiment.The second part is the design and implementation of the crop decision analysis system.This thesis completes the definition of the system,the overall design and detailed design,and applies the model of crop yield prediction to the business system.In the research of crop yield prediction model,this thesis mainly completed the construction of crop yield prediction model based on lifting regression tree,realized the construction algorithm and gave the model parameter setting method.In the analysis of the regression tree model itself,the effects of outliers on the CART tree and the regression tree are analyzed.Then,the influence of the population is analyzed quantitatively.In the optimization of the model,this thesis proposes a construction method based on the data distribution factor model,discusses the feasibility of optimization,and implements the corresponding algorithm.In the design and implementation of crop decision analysis system,this thesis gives the definition of system requirements,and designs and implements the core modules of the overall system architecture.The system is divided into four parts:data acquisition,simulation,decision analysis and full-text retrieval.The results of the final experiment show that the improved scheme based on the data distribution factor in this thesis can effectively reduce the influence of outlier on crop yield prediction model,and the requirements of the number of outliers can meet the real situation of the application scenario,which can meet the needs of the system.