Optimization of the Distribution of Stocks in Multi-Level Supply Chains with a Discrete Process of Supply

One of the main reasons for the high logistics costs in Russia is inventory control. The need to study the theory and the methodology of the organization of warehouse logistics and inventory management is considered the most relevant. The supply chain affects the company's profits and costs by managing illiquid stocks, shortages, increased stocks, and seasonal demand, choosing the optimal strategy for moving goods to warehouses. The majority of companies point out the importance of order-based inventory management (Reorder Point) strategies in conjunction with statistical modeling, but cost management does not take place in the most optimal way. First, the problem of choosing strategies in multi-level supply chains is extremely rare in theory. Secondly, specialists in inventory management, mainly, make management decisions based only on statistics, which is not always correct. In the final work, a three-level supply chain model was built (based on the AnyLogic application package), taking into account such features as a static state of the stock, for example, in the “acceptance zone” in which the material flow does not immediately enter the implementation phase and expense, and some time is registered. This principle is a novelty of the thesis, since there are few methods for optimizing multi-level supply chains and no publications of such models have been previously observed. All statistical processing was analyzed in the package R and STATISTICA. Also in the master's thesis, the method of weight coefficients, based on a neural network, is used to determine the need for the supplied products, depending on the population of cities.