Replenishment method and device

Abstract

The invention relates to a replenishment method and device. The replenishment method comprises the steps that inventory data, historical sales volume and replenishment period of an SKU are collected;giving different weights to the historical sales volume according to the time sequence, and calculating a sales volume cumulative density function of the SKU in the next replenishment period; according to the sales volume cumulative density function of the SKU in the next replenishment period, calculating the corresponding SKU goods holding capacity when the expected inventory cost value is minimum based on a preset constraint condition; and comparing the inventory data of the SKU with the stock capacity to obtain the replenishment capacity, so that the inventory data of the SKU is not less than the stock capacity, the stockout cost and the stock cost of sku are comprehensively considered, the stock capacity calculated by the model is optimal on the overall cost of the warehouse, and inventory management can be scientifically and systematically carried out on the warehouse.

Description

Technical field [0001] The invention relates to the field of inventory management, in particular to a replenishment method and device. Background technique [0002] With the rise of online shopping, the need to improve logistics and supply chain mechanisms and technologies is becoming more and more urgent. While warehouse management is already at a critical position in the supply chain, its inventory management directly affects the fulfillment rate of merchandise orders and the cost of warehouse inventory management. [0003] For the determination of the replenishment point R in the existing technical solution, [0004] A simple static inventory (holding quantity R) specified based on the historical sales average and variance of the SKU. Because the sales volume of each sku fluctuates greatly, this solution cannot adapt to the changing trend of sales, and it is easy to cause holdings R is inaccurate; [0005] The second is to use machine learning methods to predict the sales of sku ...

AI Technical Summary

Problems solved by technology

[0004] A simple static inventory (holding volume R) specified based on the average and variance of the historical sales volume of the SKU, because the sales volume of each SKU fluctuates greatly, this solution cannot adapt to the changing trend of sales volume, and it is easy to cause the inventory volume to change. R is not allowed to be obtained;

[0005] The second is to use machine learning methods to predict the sales of SKUs, so as to determine the lower limit of replenishment (stock R). Since the sales of SKUs are affected by many complex factors, the prediction error will be large, and the lower limit of replenishment for this plan The value only considers the out-of-stock cost (the temporary replenishment cost due to the out-of-stock behavior, and the temporary replenishment cost is generally greater than the regular replenishment cost), and does not consider the inventory cost

[0008] 1. The lower limit of replenishment cannot capture the changes in SKU sales in real time, resulting in a low order fulfillment rate on certain days;

[0009] 2. The lower limit of the replenishment model fails to comprehensively consider inventory costs and out-of-stock costs;

[0010] 3. The stock-out cost in the warehouse that needs to be considered in the replenishment model is difficult to estimate. When the cost of stock-out cannot be directly measured by the market price or cost of the sku, the labor cost of temporary replenishment can be used instead. However, there are two ways to estimate this value. One difficulty: first, there are very few skus with temporary replenishment, and it is impossible to directly calculate the temporary replenishment manpower for all skus in the warehouse; second, it is difficult to convert the temporary replenishment manpower into a specific amount of cost

Images