Models And Algorithms For Complex Supply Chain Coordination And Optimization

Since 1990's, supply chain management is widespread concerned as a strategy idea that can enhances firm's competitive ability. The academia and practitioners do a lot of studies and discussions for the supply chain management around the management strategy, models, the cooperation relationships of supply chain entities, the operation coordination and performance evaluation, etc.. With the development of economic globalization, the number of echelons and members of the supply chain is increasing. It makes the supply chain more and more complex. However, the current coordination and optimization theory of complex supply chain is not enough and complete. It can not provide more comprehensive approach introduction and theoretical foundation for supply chain management of real entities. After the transportation regulation is relaxed, transportation industry develops widely and diversifies. There are many transportation modes and charging way. Current models of supply chain coordination and optimization can not adapt the new development situation of transportation industry quite well. In view of the above actual situation, this paper studies the coordination and optimization problem of the complex supply chain with dynamic demand from the perspective of cost minimization with the help of some theories and methods such as game theory, systems engineering, optimization, algorithm design and computational complexity. The goal is to design models and algorithms of supply chain coordination and optimization which can adapt to the current situation.This paper analyzes the coordination problem of two-echelon supply chain with general item and dynamic demand, and establishes mathematic models containing transportation decision. Discover the coordination intends of each entity from their decision-making before coordination. Establish a game model for the two-echelon supply chain under information sharing, in which the number of dispatch can be different from period to period. Show the change of each entity's decision before and after coordination. Design improved genetic algorithm to solve the game model. Via numerical test, this paper proves that no matter which entity dominates in the supply chain, the cost of each entity decreases after information sharing. Then, this paper studies the coordination of two-echelon supply chain with discrete dynamic demand. Let retailer's cost and supplier's profit be the utility functions, this paper establishes the game model under quantity discount coordination, and proposes the existence requirement when the strategies after coordination are the Nash equilibrium policy. Exposit the relationship among decision variables, cost and discount rate under both all-unit and incremental quantity discount coordinations. The effect under two kind of quantity discount coordination is proposed by comparing the results of numerical test.Take the transportation mode, complexity of supply chain and demand type into account, this paper provides a class of complex two-echelon supply chain system optimization problem for general item. In the problem, the demand is discrete, and transportation modes and cost are considered. For centralized supply chain, this paper establishes system optimization models by combining the production cost and different dispatch cost, analyzes the structure properties and designs algorithms to solve them. For the decentralized supply chain, let general increasing functions represent the production (include transportation) and inventory cost functions, establish the mathematic model. A Fully Polynomial Time Approximation Scheme (FPTAS) is designed with the character of increasing function. For the mixed supply chain, the system optimization models under different transportation modes are established which include optimal production, transportation, procurement and inventory decisions. The optimal properties are analyzed and optimal algorithms are designed to solve these models.Then, this paper studies the coordination problem of two-echelon supply chain for perishable item with transportation decision, in which the demand rate is the general continuously differentiable function. A game model is founded for supplier's production and retailer's procurement under incomplete information. The incomplete information is reflected in the demand forecast based on orders and market investigation, respectively. A mathematic model with transportation decision is established for the supply chain system after information sharing. In the model, the cycles of retailer's order and supplier's production, the quantity and the number of dispatch can be different. Take the interaction of supplier and retailer's decision, this paper admits different production cycle and procurement cycle, and establishes the optimization model of production, procurement and transportation. It is proved that the optimal solution exists and is unique.At last, this paper studies the system optimization problem of three-echelon supply chain with discrete dynamic demand, in which the transportation decision is added into the system decisions. According to different transportation modes used in procurement and dispatch, this paper establishes different system cost multi-variable optimization models for the three-echelon supply chain consisting of multi-supplier, one distribution center and multi-retailer. The optimal algorithms are designed to solve the multi-variable optimization models which contain procurement time and quantity, dispatch time and quantity, the transportation modes of procurement and dispatch.