| The pharmaceutical retail enterprises in our country are in the stage of developing vigorously,which have faced fierce competition.In particular,to have a rapid development in the "post-epidemic era" and "14th Five-Year Plan" period,the enterprises have to deepen the innovation driving and promote the coordination of supply chain.To make the digital technology deeply integrate with the pharmaceutical retail industry,as well as solve the problems in store such as low replenishment efficiency and accuracy,long inventory turnover time,sluggish inventory funds and so on,the business data of a pharmaceutical chain limited company is taken as the basis,conducted field research on the company’s existing information system,analyzed demand,so a sales volume prediction and intelligent replenishment system for pharmaceutical retail industry is designed and implemented.The system mainly involves the three aspects as follows:(1)A replenishment planning model that satisfies distribution constraints is proposed.The phenomena of inventory accumulation or shortage will caused,since the sequential replenishment planning model adopts the same replenishment cycle for each commodity.Therefore,the indirect group strategy is adopted to build a replenishment planning model through the classification of commodities to enhance the turnover rate and improve the inventory utilization rate at the same time.(2)The proposal of an LSTM sales volume prediction model based on time series statistical features is made.In line with the calculation result in(1),it has obtained the replenishment cycle of different commodities.Based on the replenishment cycle,the prediction on the sales volume of medicines required in the future cycle is conducted.In this thesis,the sales prediction model,in which the special dates,promotions,weather features and so on are integrated.The total average index of the store is RMSE=40.85,MAE=25.1,MAPE=0.81.(3)The sales volume prediction and intelligent replenishment system for pharmaceutical retail industry is developed which is combined with the supply chain management optimization technology.According to(1)(2)model,the CPFR and S&OP plan are introduced,the process of store request for goods in the system is standardized,and the human decision-making cost is greatly reduced.According to the test and application results indicate that the system developed in this thesis also optimizes the store inventory structure while to improving the prediction accuracy and replenishment efficiency.In this way,the optimization of the digital business environment can be promoted and the modern management level of suppliers in the whole industrial chain can be enhanced. |
