Research On Location-routing Model And Algorithm Of Distribution Center Considering Demand Forecasting

With the popularization of internet technology,the rapid development of e-commerce in China has driven the rise of the logistics industry.Although the overall development prospect of the logistics industry is broad,China’s logistics development started late compared with foreign countries,the industrial mechanism is not perfect,lack of efficient modern logistics system,there are still many shortcomings.Logistics system planning,as the basis of the logistics system,has a very important impact on the operation cost of logistics enterprises.As an important part of the logistics system network,the location of distribution center and vehicle routing planning directly affect the internal operation efficiency of enterprises and determine the core competitiveness of enterprises in the market.Traditional logistics service market separate distribution center location problems and vehicle routing problem,this not consistent with the current logistics development trend,integrating both formed location routing problem for research is also the trend of the development of the logistics,and also a focus of research,many domestic and foreign scholars have studied this.When the enterprise establishes the distribution center,often is the enterprise development initial stage or the development rapid rise stage.Over time,the change of the market,have got rapid development of e-commerce and logistics,customer demand is constantly changing,and the distribution center once established shall not change,so in building a distribution center,need to consider the future customer demand change,and change is usually unknown,so the forecast customer demand for the inventory of the enterprise,the cost of distribution center has very important influence,such as size and determine the construction scale of distribution center is an important basis,the forecast of the demand of customers can seize market opportunities,improve the utilization efficiency of distribution center,reduce the cost,is of great significance to the enterprise.Paper first select important index influencing the demand of customers,and uses the method of Pearson indicators affecting the demand for screening,screening indicators will affect the customer demand as the independent variable of input,then use the classical ARIMA time series and BP neural network to predict customer demand,respectively to calculate the average relative error,and using shapley model of two kinds of combination forecasting method,draw a new linear programming model for the prediction,finally using the forecast model to forecast the customer demand.Numerical examples show that the combined prediction model is more accurate than the single prediction model.Location-routing problem is an important problem in supply chain management and logistics system planning,which has a very important impact on the total cost.This paper studies the location-routing problem with time window considering the capacitate constraint of distribution center,establishes a multi-objective programming model with the goal of minimizing total cost and maximizing customer satisfaction,and proposes a two-stage algorithm to solve the problem.First,k-means clustering algorithm is used to determine the location of distribution center.Then,a customer division method based on time-space factors is proposed to determine the customers served by the distribution center.Finally,particle swarm optimization is used to plan the distribution path of each distribution center.Several numerical examples show that the proposed algorithm can effectively reduce the total logistics operation cost and total distribution path length compared with other existing algorithms,providing a new solution to the location-routing problem with volume constraint and time window.Finally,the sensitivity analysis of the scale of distribution center is carried out.From the analysis,it can be seen that the scale of distribution center has a very important impact on the distribution routing and total cost.Therefore,it is of great research significance to consider the location-routing problem of demand prediction.