Research And Realization Of Sugar Beet Prediction Based On BP Neural Network

As the main economically grown crops in Northeast China,sugar beet production and sugar content rate restrict the development of farmers' income and sugar industry,especially the harvest of sugar beet quality and output can increase farmers' income and promote rural economic development.Sugar beet yield and sugar content are the main problems to be solved,so this paper proposes an intelligent algorithm BP neural network to predict sugar beet sugar content research and implementation.This article mainly studies the prediction of sugar content in sugar beet from three aspects: theoretical analysis,algorithm simulation,and application realization.First,the principal component analysis method is used as an auxiliary method for screening parameters that affect the level of sugar content.This method can effectively improve the quality of input parameters and reduce the coupling between variables;Falling into the shortcomings of local minima,an improved particle swarm optimization algorithm is used to optimize,and a non-linear inertia weight improvement formula is proposed.The feasibility of this improvement is discussed and analyzed,and an IPSO-BP sugar beet sugar content prediction model is finally formed.A beet experiment area was used for model simulation application,and based on the simulation results,the sugar beet water demand irrigation control theory was proposed,and finally the functions of the sugar beet predictive control system were realized through software.Experiments show that the sugar beet sugar content prediction model proposed in this paper has certain practical significance and is also of reference value for guiding farmers to grow sugar beet.The prediction accuracy of IPSO-BP neural network is20%?30% higher than that of PSO-BP neural network and BP neural network,the prediction result is satisfactory.