Design And Implementation Of Logistics Delivery System Based On Traffic Flow Forecasting

With the rapid development of logistics delivery business and the increasing demand for door-to-door delivery of goods,better logistics delivery service has become one of the important demands for a better life.In the whole process of logistics delivery,terminal logistics delivery is one of the most important links which directly handover with customers.Whether goods can be delivered to customers in time will directly affect the customer's user experience.In order to improve the efficiency of urban terminal logistics delivery,it is particularly important to accurately and objectively evaluate the cost of the delivery process(such as transportation cost,time cost,etc.)in the process of optimizing the route selection.Firstly,this paper establishes a cost estimation model of logistics dispatch at the end of the city to calculate the dispatch cost objectively,and then designs a logistics dispatch system based on this model to improve the efficiency of logistics dispatch personnel.The main work of this subject is as follows:(1)In order to improve the delivery efficiency of logistics personnel,this paper develops the logistics delivery system based on JavaEE and MySQL technology,and develops the corresponding app client based on Android platform and Baidu Map Open Platform.The main functions of the system include positioning function,route navigation function,dispatcher account management function,dispatcher dispatch Express function and road congestion prediction and evaluation function.(2)In order to accurately and objectively evaluate the cost of dispatching process(such as transportation cost,time cost,etc.),this paper puts forward a logistics dispatch at the end of a city by studying and analyzing the determinants of the cost of dispatching at the end of a city(driving mileage,unit mileage cost,congestion coefficient,etc.)and considering the intrinsic relationship among driving mileage,unit mileage cost,congestion coefficient and so on.The delivery cost estimation model provides a basis for the selection of logistics delivery routes.(3)In order to choose a better delivery path,this paper firstly obtains different road congestion states through the optimized K-means clustering algorithm,and then uses BP neural network combined with statistical theory to obtain the optimized model to predict road congestion coefficient.The experimental results show that theprediction accuracy of the classical BP neural network model is 94.3%,while the prediction accuracy of the final prediction model is 97.5% in the prediction stage combined with statistical methods.Combined with the cost estimation model proposed in this paper,the cost of each dispatch path under the condition of congestion coefficient limitation can be calculated,and the optimal dispatch path can be obtained.(4)Through functional testing of the completed logistics delivery system(positioning function,route navigation function,account management function of distributors,express delivery function of distributors,prediction and evaluation function of short-term road congestion status),the results show that the comprehensive ability of the system has reached expectations and can run normally on-line.