Slot Allocation Stochastic Models For Container Liner Shipping With Revenue Management

With the fast development of container shipping industry and the increasing intensity of competitive market, the optimization of container transportation flow system has become a hotspot that arouses great concern from the container transportation companies and interrelated organizations. One of the key problems is the container slot allocation under uncertainty. Traditionally the slot allocation aimed at lifting loading rate of the ships. But facing the current fierce competition and much lower freight rate, the liner shipping companies cannot raise profit by lifting loading rate separately. Dramatic changes are required in operational practices of cost reduction. Instead companies are supposed to use revenue management (RM) method and allocate slot to the cargos with higher marginal contribution based on market subdivision so as to avoid the fluctuant risk of freight rate and lift the profit RM can be defined as the integration management of inventory and price to maximize a company's profitability, which has been proved effective in the airline industry for many years. Compared with the airline industry, RM practice now has not been widely used in the container liner shipping industry, and the RM studies are relatively few.Similar to airline industry, as a typical service industry, container liner shipping industry also has the characteristics for the application of RM, such as relatively fixed and perishable capacity, advance booking, high fixed costs and lower marginal costs, stochastic but segmentable demand, fiercely competitive market and so on. Therefore RM also has considerable applicative promising in the liner shipping industry. However due to the difference between airline industry and container shipping industry, the RM methods successfully applied in the airline passenger industry can't be copied to cargo industry directly. Based on the analysis of the characteristics of container liner shipping RM, a liner shipping RM concept model is developed accordingly in the thesis. A series of mathematical models are established by stochastic and dynamic programming methods for the further analysis of the container slot allocation. The chance constrained-programming and robust optimization methods are introduced to solve the models to get satisfactory results in reasonable computation time. In addition, some of the parameters are evaluated and analyzed to get the right revenue management policy.Trade imbalances and the reusable characteristic of containers have made reposition of empty containers a special problem in container liner shipping RM. Therefore, two stochastic programming models are developed on capacity allocation, one with empty containers transportation involved and the other without. They aretransformed to linear integer programming models by the method of robust optimization. The thesis compares the results under two conditions by digital simulation, which shows the applicative value of these models and methods in solving the container imbalance problem faced by liner shipping companies under uncertainty.Then the thesis studies the group demand problem in liner shipping container slot allocation. There are at least two types of group demand in container liner shipping industry: one is batch demand which has large demand quantity but little bargaining power for discounts; the other is contract demand which has long-term contract with the carrier long before the sales season and strong bargaining power, so it can usually get discount freight rate. As to the first type of group demand, a dynamic and stochastic programming model is established. Based on the analysis of the characteristic of the model, the dynamic booking control policies are put forward based on threshold value without batch demand involved. When considering batch demand, the invalidate conditions of the threshold control policy are analyzed by digital simulations. As to the study on contact demand, firstly, based on the newsboy model simulation method, the optimization condition of the model and the optimization slot allocation policy is achieved under single-product single-leg state. Then the chance-constrained programming method is introduced to transform the multi-product multi-leg model into an integer programming model that can be solved directly. Finally, the optimization slot allocation policy is achieved by digital simulation, which showed the advantages for the company of using long-term contract.Different from passenger transportation demand, cargo demand generally is service-bound, and shippers have less concern about the container transportation route. Therefore, under circumstance of complex transport network and demand uncertainty, it is an important task for liner companies to allocate their limited container slots in multiple lines effectively, and select the right path for the transportation of the demand between port pairs. Based on the study of the pre-chapters, considering multiple sea routes and container transportation route choice, a stochastic programming model is proposed on the container slot allocation. The robust optimization method is introduced to solve the model. Finally, some of the parameters are analyzed through sensitivity analysis by case study, which shows the model and method effectively direct the liner shipping companies in the slot allocation decision and consequently lift the loading rate and the revenue level of the company.