(R,S) Envanter Politikası İçin Statik Fiyatlandırma Kararları
Özet
Along with the process of globalization, the competition between companies has become quite intense. In
this competitive environment, companies aim to achieve the operational efficiency and to increase their
market share. In line with this purpose, their capability of matching demand with supply has become crucial.
Inventory and pricing decisions directly influence the supply and demand processes. In recent years, many companies concentrate their effort on taking these decisions together and, therefore, minimizing the risk of potential gap between demand and supply. In this context, while demand process is influenced by price regulations, supply process is controlled by inventory policies. These decisions jointly serve the purpose of increase in business profitability. Therefore, companies need to make optimal inventory and pricing decision in order to attain maximum profit. This necessity gives birth to important mathematical models, developed by researches, for guiding companies while they develop their pricing and inventory policies.
In this thesis, a firm which sells a single product over a finite planning horizon is considered. Inventory is
controlled based on (R,S) policy and price dependent stochastic demand function is assumed to be in an additive form in each period. The stochastic component of price dependent demand function is a random variable which follows normal distribution. Total cost is comprised of fixed ordering cost, unit ordering cost, holding and shortage costs. The proposed inventory-pricing policy to control the system described is named as (R,S,p) policy. The essence of (R,S,p) policy lies in determining replenishment intervals, orderup-to levels and pricing decisions at the beginning of the planning horizon. In this context, two quadratic mixed integer programming formulations are proposed for finding the near optimal parameters of (R,S,p) policy.
Furthermore, in this study, the proposed mathematical models are extended by employing the assumptions that companies often face in real life applications. Consequently, the computational efficiency of proposed models are compared by means of many test instances.