Research On Optimization Models Of Production-Inventory Decision Making Under Uncertainty

Uncertainty in inventory problems is shown mainly in demands, costs and replenishments. Stochastic inventory problem has been thoroughly studied and achieved fruitful results. But in the real inventory decision-making problems, there are always other uncertain phenomena, such as fuzzy phenomenon, random fuzzy phenomenon, or fuzzy random phenomenon. Till now, the research in this area is still in its infancy stage. Therefore, examine how to establish the optimize inventory model under uncertain environments, would be of great value and theoretical significance.This dissertation considers the inventory decision-making problem in uncertain environments, and some optimization models according with the actual inventory control are proposed. Moreover, the corresponding algorithm is designed to solve these models. The contents are described as follows:In fuzzy environments, the economic order quantity expected value model without backorder and economic order quantity dependent chance programming model without backorder are constructed. And also, the economic production quantity expected value model with backorder and economic production quantity chance constrained programming model with backorder are constructed. Then fuzzy simulation and particle swarm optimization algorithm based on the fuzzy simulation are designed to solve the dependent chance programming model and chance constrained programming model, respectively. Whereafter, numerical examples are presented to illustrate the feasibility and validity of the designed algorithm.For the inventory problem in random and fuzzy coexisting environments, with assumptions that the percentages of nonconforming items are random variables and the cost parameters are fuzzy variables, the fuzzy random expected value model and fuzzy random dependent chance programming model are constructed, respectively. In order to solve these models, random simulation, fuzzy simulation, fuzzy random simulation and particle swarm optimization algorithm based on the fuzzy random simulation are designed. Then the validity of the designed algorithm is illustrated by numerical examples.For the inventory problem in the case that the percentages of imperfect items are random fuzzy variables, this dissertation also constructs the random fuzzy expected value model. In order to solve the presented random fuzzy expected value model, particle swarm optimization algorithm based on the random fuzzy simulation is designed. Furthermore, a numerical example is presented to show the high efficiency and good performance of the designed algorithm.For the continuous review inventory problem with backorders and lost sales for variable lead time in fuzzy random environments, with assumptions that the demand in leadtime follows the normal distribution and the free distribution, this dissertation constructs the fuzzy random expected value models, respectively. Then the propositions of the objective function are analysed. In addition, the iterative algorithms are designed to solve these models.