Bi-level programming for modelling inventory sharing in decentralized supply chains

This paper deals with inventory sharing in the context of a decentralized supply chain. The chain consists manufacturer who distributes set products through network independents Points Sale (POS). primary decision each player has to make is optimize its service level while maintaining minimum total cost. POS share their allowing transshipment enhance ability face demand and avoid shortages. problem modelled as 1-leader, n-followers Stackelberg game. A mixed integer bi-level program developed considering that decides first on levels routes be constructed knowing follower\x92s (POS) response function arising from own minimization In this game, incurs vehicle routing cost for regular shipments. addition holding costs, it assumed are willing incur both part lost sales associated shortage transshipment. To address combinatorial complexity provide efficient solutions large size instances, hybrid Genetic Algorithm coupled deep reinforcement learning developed. Results show mechanism under certain conditions allows whole achieve savings improve level.