Date of Graduation
Container transport operations have been extending inland, providing more comprehensive service across the shipping network. Accordingly, container transport operators are making extensive capital investments in deploying inland container depot (ICD) networks. Optimizing the location of such facilities financially is vital for both capital and operating efficiencies. Currently, there are no models at the regional network level to guide container operators in locating ICDs on their networks. This research studies the ICD location problem and develops a comprehensive ICD location model. Based on comprehensive analysis of the container transport industry, focusing on ICD operations, this thesis developed a useful formulation of the ICD location problem. It recognizes and emphasizes the need to embody the endogenous demand and market competitiveness in the container transport business. The formulation combines the multinomial logit model of discrete choice analysis to quantitatively describe the shipper's behaviors and preferences, addressing both the endogenous demand and market competitiveness. Fixed charge facility location problems are considered or proven to be NP complete. The use of multinomial logit model inevitably further complicates the formulation in both the objective function and constraints, which justifies the use of heuristics to solve the formulation. Based on the mathematical properties of the formulation, a problem specific tabu search heuristic algorithm was designed for its solution. Tabu search parameters, tabu tenure, slack parameters, and cutoff parameters, are tested to decide their contributions and sensitivity to the computational time and optimality. Statistical inference was used to evaluate the performance of tabu search algorithm. The result of the evaluation shows that the solutions of tabu search algorithm are within the confidence interval of optimal solution of this formulation. This proves that the designed tabu algorithm is efficient and effective in solving the problem.
Xu, Yuanquan, "A discrete choice based facility location model for inland container depots." (1999). Graduate Theses, Dissertations, and Problem Reports. 10064.