Research On Yunda Express Business Volume Forecast Based On RBF Neural Network

In recent years,the rapid popularization of e-commerce has promoted the rapid development of express delivery terminal logistics.As China's third-party express logistics enterprises in the early development stage of "multiple,scattered,and chaotic",most express logistics enterprises have low informatization and terminal service standardization,and logistics services are relatively single,resulting in services low quality and poor service level.In view of the problem of express business volume prediction,only considering the traditional forecast indicators is not enough to meet the diversified express logistics needs of customers.Therefore,this paper includes the service index into the predictive index system.However what this research intends to solve the key issue is how to filter out the serviceable predictive indicators and use the serviceable indicators to predict the express delivery business volume to reduce the forecast error,and combine the prediction results to find the key influencing indicators to improve the express delivery business volume and the quality of its service.This paper takes Yunda Express as the research object,and uses SWOT analysis method,structural equation model and RBF neural network prediction model to predict the business volume of the express company,to reduce the error of Yunda Express business volume forecast,while considering the logistics service quality to predict the results,then,build a multi-factor Yunda express business volume forecasting index system for forecasting research.Firstly,find out the problems in Yunda Express' s businessvolume forecast through combing and analyzing the current status of Yunda Express' s logistics services and business volumeand the relationship between them;Secondly,aiming at the problems in the selection of Yunda Express' s indicators,initially constructed the direct serviceability and indirect serviceability prediction index system for express delivery business volume prediction.The index system includes two parts: qualitative and quantitative indicators.The qualitative index part adopts structural equation model combined with questionnaire survey method,through the design of the scale and the questionnaire,the direct serviceability prediction index which can predict the volume of China's express delivery can be selected,then screen out the indicators that can predict the business volume of Yunda Express and complete the construction of the Yunda Express business volume forecasting index system;then build an RBF neural network prediction model to predict the business volume of Yunda Express.At the same time,build GM(1,1)prediction model and unbiased RBF neural network prediction model for comparison and analysis;Finally,sensitivity analysis is used to identify key indicators of direct serviceability,and based on the results of sensitivity analysis and predictive index selection,countermeasures and suggestions for Yunda Express' s operations are proposed from two aspects: improving logistics service quality and increasing express delivery business volume.