Research On Theory And Algorithm Of Twice-order Newsvendor Model

Newsvendor problem is also called single period inventory problem. It is astochastic programming model making optimal order decisions for short cycle productsto maximize profit or minimize cost. Since Edgeworth, a famous economist applied thecenter-limit theorem to solve the optimal bank cash reserves problems in1888; thenewsvendor model had progressively gained extensive attention in academics.Researchers at home and abroad have widely used it in production, service, logistics,management, finance and other fields.Due to several uncertain factors in the model, deciding the optimal order quantityhas been researchers’ main task. In order to determine the optimal ordering and makemore profit for deciders, it need to balance between ordering too much and too little. Theintroduction of the second order can effectively solve this contradiction, which is themain idea of this paper, i.e., twice-order newsvendor model.To solve the twice-order model, two-stage method will be applied to model atwice-order newsvendor model with fixed recourse. The equivalence between recoursemodel and discussed model will be verified, and then L-shaped algorithm for the modelwill be present. Considering decision makers may face demand with unknowndistribution function, part one of chapter four researches distribution-free twice-ordernewsvendor model and inferred the optimal order quantity and corresponding best profit.When compared with classical newsvendor model, it can be observed that in twice-ordermodel more expect profit can be generated. In part two of chapter four, twice-orderdistribution-free newsvendor model under random and fuzzy environment is established,some properties of optimal solutions will be discussed and numerical example verifiesthat more uncertainty causes larger optimal order quantity and less except profit.