Research Of Forecasting Model In Soybean Pest Yield Loss Based On Chaos-genetic Neural Network

China has a long history of soybean planting and is a big soy-bean growing country.Soybean is not only the main planting type crop in northeast China,but also plays a significant role in the grain production nationwide,with considerable economic and ecological benefits.As one of the main reasons for the decline of soybean yield,the increasing harm rate has posed a great threat to soybean production and quality.Therefore,it is necessary to predict and control the production loss caused by agricultural pests and diseases in advance,and it is of great practical significance to carry out in-depth research and analysis.In view of the above problems,this paper studies the prediction model of soybean pest yield loss based on the chaotic-genetic neural network(CGA-RBFNN).Firstly,the genetic algorithm is optimized by chaotic search,and the radial basis neural network(RBFNN)is optimized based on the chaotic genetic algorithm(CGA),and then the out-put loss prediction model is constructed to predict the loss of soybean production caused by pests.Finally,the prediction model is embedded into the intelligent decision system,and the intelligent management decision system for soybean pests is designed and implemented.The main research aspects are as follows:(1)Data acquisition and pre-processingVarious kinds of sensors are used to monitor and extract data from soybean field.The obtained data is pre-processed by normalization and other methods,making it a valid data that can be fully used in simulation experiments.(2)Research on chaos-genetic algorithmIn order to find the optimal solution more suitable for optimization,the chaos search algorithm is used to optimize the genetic algorithm.It is clear that the chaos genetic algorithm can avoid falling into the local optimum to a great extent.The method is to add the chaotic variable to the optimization process of genetic algorithm,then generate a series of small fluctuation solutions around the generated optimal solution to obtain the optimal solution.(3)Research and application of chaos-genetic neural network fusion algorithmA fusion algorithm based on chaos-genetic algorithm to optimize RBF neuralnetwork is proposed.The algorithm uses chaos genetic algorithm to optimize the weights,thresholds and hidden data centers from hidden layer to output layer of RBF neural network.Based on the optimized algorithm,a yield loss prediction model based on chaos-genetic neural network fusion algorithm is constructed,which can predict the soybean yield loss caused by insect pests.(4)Development of intelligent management decision system for soybean pestsUsing the above research method,the intelligent management decision system of soybean insect pests is developed on the platform of.NET,using C# language.The system includes the functions of user management,insect pest query,diagnosis and prevention,yield loss prediction,expert consultation and help.This paper is based on the “precise technology integration and demonstration of corn operation base on Internet of Things” project that belongs to Science and Technology Department of Jilin Province,combined with data mining technology,artificial intelligence technology,intelligent decision expert system and the prediction model of soybean pest yield loss is constructed.In the same time,the writer also develops intelligent management decision system for soybean pests.The system uses a variety of information technology to provide decision analysis and advice for the whole planting period of soybean,so that soybean growers can accurately diagnose and prevent the various pests of soybean,and it also can reduce the output loss caused by pests,increase soybean production,and provide technical support for soybean production and operation management.