Envanter yönetiminde Markov karar süreci

In any industrial production, the need for strategic or contingency planning continues to attract more attention, especially with the growing competitiveness of today's markets. Managers of companies seek to adopt policies that in the long run minimize operational costs. In this thesis, a case study of an inventory control problem is studied and an optimal stationary policy is proposed. The problem has to do with optimal production scheduling of limestones, a firm located in Trakya, Turkey. A production schedule that minimizes expected discounted costs is sought. The mathematical modeling framework used for this study is infinite-horizon Markov decision process. The problem is formulated as a sequential decision problem and subjected to rigorous computational test. We proposed an initial policy which is obtained through policy iteration methodology and we check whether it is optimal. Based on this, our proposed policy is found to be optimal. Whiles this policy is optimal, we do not claim that it is the only optimal solution to the problem herein studied. Other forms of optimal policies to this same problem may exist.