Optimization of Stock Basing on Improved Grey Prediction Model: A Case Study on Garment Supply Chain

Abstract

Grey prediction model is the core of grey system theory. It has been widely used because of the lack of strict requirements and restrictions on data. However, when the general grey prediction model is be applied to predict the random oscillation sequence, the accuracy is not ideal. For this reason, this paper selects a small sample with a large amplitude of oscillation characteristics of the initial product demand on the market for research. The grey prediction method is be improved innovatively, and the envelope curve of grey oscillation interval can be used to predict the value interval of demand, then to determine the partition function of demand. Basing on the actual operating data from garments enterprise, the author applied the improved grey prediction method and used the Monte Carlo method to simulate the demanding amount for next three cycles. It makes the setting and optimization of safety stock more accurate and effective data support, reduces the cost redundancy caused by information uncertainty, and improves the service level.