Multi-level inventory quota formulating method based on data driving

Abstract

The invention discloses a multi-level inventory quota formulating method based on data driving, which comprises the following steps: inputting a material code, a time node, item information in which materials participate and corresponding demand data to obtain a collected data sample, and carrying out data preprocessing; based on the preprocessed data, obtaining a future demand plan predicted value of the materials according to a monthly period through a data-driven demand prediction strategy, and calculating and analyzing a storage mode suitable for the materials according to the material coverage rate and the material importance in the preprocessed data; calculating a mean value and a variance of inventory historical data of each warehouse and a correlation between each material and different warehouses according to the material inventory reporting period; and calculating the stock quota of the materials by using the future demand plan predicted value of the materials and the correlation, outputting the storage mode and the stock quota of the materials, and completing the formulation of the multi-level stock quota. According to the invention, the cost is reduced, and the supply chain efficiency is improved, so that the method has more advantages in competition.

Description

technical field [0001] The invention relates to the technical field of inventory quotas, in particular to a data-driven multi-level inventory quota formulation method. Background technique [0002] At present, with the continuous improvement of enterprises' requirements for supply chain management, the single consideration of inventory optimization in fixed areas will inevitably lead to a sharp increase in costs. It has become an urgent need to realize cross-regional multi-level warehousing. [0003] Aiming at the optimization problem of multi-level inventory, some research work has been done in recent years. After establishing the multi-level inventory model, the total inventory cost of the supply chain can be effectively reduced through the immune genetic algorithm and the particle swarm algorithm, but these methods only consider the cost factor and lack applicability; Zhao Baiqiang proposed a multi-level inventory of the supply chain based on time competition The model a...

AI Technical Summary

Problems solved by technology

After establishing the multi-level inventory model, the total inventory cost of the supply chain can be effectively reduced through the immune genetic algorithm and the particle swarm algorithm, but these methods only consider the cost factor and lack applicability; Zhao Baiqiang proposed a multi-level inventory of the supply chain based on time competition The model achieves the lowest total time cost through the multi-objective genetic algorithm, but the model lacks the optimization of the scheduling process; Yao Lanping improves the accuracy of demand forecasting, and uses the ABS-LMS matrix to classify and manage accessories, which solves the problem of high inventory levels problems, but inventory site occupancy has not decreased; Peter Ko ·· Chel et al. used a heuristic algorithm to analyze a system instance composed of five retailers to obtain the optimal ordering decision, and considered adding more sorting strategies in the future; WeiGeng considered six information access scenarios Inventory model, the ordering strategy is obtained through the dynamic programming method; HsinRau et al. prove that in the multi-level supply chain management, the decision-making effect from the overall situation is the best, and the inventory control is realized, but the model only considers a single member in each level of inventory , still needs to be further expanded

Although the above works have achieved multi-level inventory optimization to a certain extent, they have less consideration for safety stock requirements and have not optimized cross-regional inventory coordination. Therefore, theoretical modeling of multi-level safety stock optimization is still needed.

Images