Research On Application Of Distribution System Based On Neural Network Prediction Algorithm

In recent years,with the rapid development of e-commerce in China,the "Internet + agriculture" model has become a new trend,and it is particularly important for small and medium-sized enterprises to expand their online sales channels.Panjin crab e-commerce distribution system is a small program application developed by the author for hometown enterprises,predict future sales trends through actual orders,mainly use neural network technology to optimize the prediction algorithm,to solve the problems of regional distribution,online and physical inventory prediction,and brand awareness for enterprises.This article conducts algorithm learning and research on the demand forecasting of enterprises' orders.The main work is the optimization of order forecasting algorithms of dynamic neural networks,and a hybrid LM-LS forecasting algorithm is proposed.,innovations include the selection of thresholds in time series data denoising and the construction of a dynamic fuzzy WNN model,statistical data smoothing and accidental orders are fitted with Poisson distribution,after processing the order data in the distribution system,the model is established and solved,and the optimized high-frequency prediction is compared to obtain more accurate prediction results and accuracy.Finally,the optimization algorithm in the article is applied to the Panjin crab distribution system.The optimized dynamic algorithm can handle dynamic variables very well,any new order can be input in real time,real-time prediction,high-frequency prediction results can be accurate to the hour level.Based on the algorithm research,this article also summarizes the research on the significance and background of the development of e-commerce distribution systems,clarifies the development goals of the system,and then conducts research and analysis of the current mainstream distribution platforms,integrating the development status of domestic and foreign e-commerce platforms Contrast,research and develop suitable development methods for river crab and other fresh platforms.After the design of the system was completed,the system was officially put into operation,and the 1.0 version of Tongfu Crab E-commerce Distribution Platform was released in a small program.The main function pages of the front end and the background are shown in detail in the chapters in the text.The background order data provides the basis for algorithm research.Practice has shown that the prediction algorithm based on dynamic neural network has improved the prediction effect in the application.It not only provides data support for the sales of next year in the distribution system,but also effectively solves the problems of online and physical expected stocks and fresh logistics time limit requirements.At the same time,the e-commerce distribution system developed at the same time effectively solves the operational problems that occurred when the original enterprise joined the third-party platform,whilebroadening the sales channel of the crabs and increasing the brand awareness of the enterprise.