Supply chain management in second half recent century extended 'so much and in current century this extension is continued. One of the scopes that to be interested is extension and adaptation of production and inventory control policy, under just in time philosophy, for material and production control in supply chain. Hence in this paper we will adapt pull-push CONWIP (constant work in process) policy in supply chain context and model this Strategy. Demand rate, production, setup and transportation time are uncertain. In optimization process, concerning supply chain complexity, we will use simulation optirilization procedure that has more flexibility for solving problem. Objective function is summation of total work in process, hack log and supply raw material that will be minimized. At last, model result will be studied and analyzed.

One of the main efforts in a pull system to meet the final goal of Just-In-Time (JIT) is to decrease Work-In-Process (WIP) without reducing throughput rate and service level. This paper proposes a new hybrid control policy using Kanban, Base Stock and CONWIP to form the final construction of production plan, and then the function of new model is shown by a 3 stage case study with probabilistic parameters. The results obtained using simulation and solving various problems indicate that the inventory level is reduced to an acceptable range. Also, Data Envelopment Analysis (DEA) is implemented for the evaluation and verification of the proposed policy with functional indicatives.

One of the efforts in a pull system to meet the final goal of Just-In- Time is reducing Work-In-Process (WIP) without reducing throughput rate an service level. This paper proposes a new hybrid control policy using extended Kanban, Generalized Kanban and CONWIP to form the final construction of production plan. Simulation of a 3-stage case study with probabilistic parameters showed the reduction of inventory level with acceptable service quality. Data Envelopment Analysis (DEA) is implemented for the evaluation and verification of the proposed model. The result obtained indicates that the proposed model is quite efficient and applicable to solve real world problems.