A Study Of Inventory Management Optimization In Supply Chain Based On Genetic Algorithm

Inventory is temporarily idle resources for enterprises which are designed to meet market future needs.It not only release the contradiction between supply and demand in long time cycle,but also guarantee the continuity of enterprise production,thus effectively satisfy the random demand of the products.With the rapid development of world economy,enterprise management mode also constantly updated,which obviously have more strictly requirements for inventory management.With the distinct development and the successful application of supply chain management theory due to its own advantages,the supply chain inventory management has become the intensive concern for the industry and researchers.The new management solves the problem of inventory management supply chain from an overall perspective,regards the inventory involved within the supply chain as a single system,thus avoid the disadvantage of the traditional business inventories which are independent of each other.The core goal is: on the premise of guarantee the customer satisfaction target,reducing the whole inventory cost of the system as much as possible,in summary to achieve the overall optimal goal.This paper carried out the following aspects,first of all,the current supply chain inventory management characteristics and existing problems are analyzed,based on the modern inventory management strategy,decentralization model of inventory management system is established.Secondly,the mathematical optimization model for supply chain inventory management is built where the objective function is supply chain inventory cost,the storage space are set as the constraints,with the system of regional distribution center as an example.Thirdly,a hybrid method combing the basic genetic algorithm and gradient algorithm is presented for the proposed optimization problem.The advantage of genetic algorithm lies in its ability of finding global optimal,but the easy prematurity and the poor local search ability are existed.Therefore the basic genetic algorithm is hybrid with the gradient-based algorithm.By keeping the optimal chromosome in genetic algorithm as the starting point of gradient algorithm and importing step size determination algorithm to figure out the exact step size along the search direction,the convergence speed is increased by iterations.With the new method,the global optimal is reached and the computational cost is reduced,so it can ensure a good convergence to global optimal solution.The programming tools are applied to solve the well-established problems.The theory and method are used for the supply chain inventory management optimization of company A.Results are analyzed,and its applicability is verified.Application results show that the improved genetic algorithm optimization speed is faster than genetic algorithm optimization.Finally,the research work of is summarized,the existing problems are analyzed and the future work is described.