Integrated Inventory And Transportation Predictive Optimization Models For Multi-echelon Supply Chains Based On Characteristics Of Varying Demands

In recent years,with the rapid development of market economy and information network technologies,the relationship between the production and operation activities of enterprises and their upstream and downstream enterprises in the supply chain has become more and more closely.Supply chain management directly affects economic benefits of enterprises.With the increasingly complex domestic and international environment,supply chain management of enterprises is not limited to vertical operation and large-scale production but gradually changed to be driven by customer demands.In supply chain management,if an enterprise would like to improve its competitiveness in the market economy,it is necessary to establish a close cooperative relationship with its upstream and downstream enterprises,obtain timely market feedback information,and optimize operating costs while satisfying customer demands.Then,the economic efficiency of the enterprise can be improved.In contrast to the single structure of traditional supply chains,modern supply chains have a variety of characteristics of wide range,large scale,and high complexity,which have gradually evolved into huge and complex multi-echelon systems.The difficulties and risks of their management are also gradually increasing.Potential challenges and problems in multi-echelon supply chain management may arise,such as inventory redundancy,inefficient transportation,and supply disruptions.Therefore,designing a fast and efficient optimization model is of great significance for scientific management of multi-echelon supply chains.This thesis considers two types of demand varying characteristics,and takes integrated inventory and transportation management of multi-echelon supply chains as the research object.System theory,dynamic management,quick response,and economic model predictive control are used as theoretical foundations,and research methods such as multi-objective optimization and mathematical programming are applied to analyze the optimization process of integrated inventory and transportation management of multi-echelon supply chains.From the perspective of integrated management of inventory and transportation for multi-echelon supply chains,this thesis is dedicated to investigating the following aspects:First of all,based on economic model predictive control,an integrated inventory and transportation optimization model for multi-echelon supply chains is proposed to improve problems raised by only inventory or transportation management,such as low transportation efficiency,product redundancy,increased operating costs,etc.Under the condition of known and deterministic forecasting demands,by analyzing the dynamics of inventory level and backorder of enterprise at each node,operating costs of inventory holding,backorder and transportation are optimized and meanwhile,the constraints of inventory,two different transportation modes,full-truckload(TL)and less-than-truckload(LTL),can be satisfied.The proposed predictive optimization model can effectively manage multiple products and handle varying customer demands.Without manual intervention,the proposed predictive optimization model can provide enterprises with real-time optimal decisions such as orders,shipments,and the number of vehicles dispatched.Under the condition of deterministic demand forecasts,all the customer demands can be satisfied in time and the total operating cost is minimized,which reflects the advantages of automated multi-echelon supply chain management.Secondly,an integrated inventory and transportation predictive optimization model based on stochastic demands is proposed for multi-echelon supply chains.By using a scenario-based approach,it can effectively deal with a large amount of uncertainties in real customer demands.The proposed predictive optimization model can minimize the total operating costs of multi-echelon supply chains while satisfying varying customer demands by punishing retailer backorders.Since real customer demands are unknown in advance,the optimal decisions of inventory and transportation management obtained with stochastic demands may lead to a shipping risk for retailers.By analyzing this risk,a condition for selecting the number of scenarios is designed to guarantee that this risk can be maintained within a certain range.With stochastic demands,the scenario-based integrated inventory and transportation predictive optimization model can effectively reduce operating costs with a guaranteed risk level,and provide real-time optimal operations for automated inventory and transportation management.Last but not least,a case study of the Y-company multi-echelon supply chain is carried out.According to the actual operating cost and transportation constraints,optimal decisions for integrated inventory and transportation management are made for the Y-company multi-echelon supply chain.From three aspects of inventory,backorder,and transportation capability,the effectiveness of this predictive optimization model in complex supply chain of multi-echelon,multinode and multi-product is verified.Through a comparison with the cycle counting strategy currently adopted by this supply chain,the outstanding performance of the predictive optimization model is demonstrated.Additionally,the theoretical contributions and managerial insights are summarized,which provides supply chain managers with management concepts such as management models and path optimization,and also provides theoretical support for future supply chain design in terms of storage settings and location selection.In summary,this thesis has proposed two integrated inventory and transportation predictive optimization models for multi-echelon supply chains under two types of demand varying characteristics.These models can respectively address the problems of product redundancy and low transportation efficiency in a single inventory or transportation management,as well as the problems of supply-demand imbalance and shipment risks caused by uncertain customer demands.Based on the proposed management models,the predictive optimization models have been applied into the Y-company multi-echelon supply chain,which can fill the gap in the application of complex supply chains with multiple levels,nodes,and product flows,and can also provide theoretical guidance and management insights for multi-echelon supply chain management in the retail industry.