Research On The Methodology Of Optimization Modelling For The Multimodal Routing Problem Based On Transportation Scenarios

The sustainable growth of national economy and society requires a solid support from an efficient freight transportation system. With the constant improvement of the national integrated transportation system, multimodal transportation which is an advanced transportation organization means has been significantly promoted and applied in the transportation practice, and the multimodal service network therewith tends to be rapidly developed, which remarkably enhances the economic and social benefits derived from the transportation industry. The effective operation and management of the multimodal service network depends on a reasonable transportation planning. In the"Strategic-Tactical (Network Design)-Operational" 3-level planning architecture for the multimodal service network, the routing optimization that belongs to the operational level is oriented on the customers' multiple transportation demands, and directly reflects the economy, efficiency and reliability, etc. of the multimodal service network, and thus processes great value in both research and practical fields. Consequently, multimodal routing problem is one of the spotlights in the transportation planning field. However,the weaknesses in the existing multimodal routing literatures affect the feasibility of the model formulation and solution algorithm design in dealing with the real-world problems.Based on such situation, from the viewpoint of system engineering, this dissertation comprehensively considers six problem formulation characteristics including optimization object, transportation service pattern, network capacity, commodity flow integration, optimization criterion and solution method. Starting from the modelling of the fundamental multimodal routing problem, this dissertation first subdivides the problem according to the freight categories (conventional goods represented by the regular containers and unconventional goods represented by the hazardous materials),and further systematically explores the complex modelling and exact solution strategy design for the multimodal routing problem. The accomplishments achieved in this dissertation are presented as follows.(i) In the problem formulation, this dissertation analyzes the six mentioned formulation characteristics in a comprehensive manner. Specifically for the transportation service pattern setting, this dissertation emphasizes two categeries of service patterns: time-flexible service pattern represented by road services and schedule-based service pattern represented by rail services. Then this dissertation concludes and summarizes the railway schedule-based space-time constraints on the multimodal routing. As a result, the transportation service pattern setting in the modelling is extended from single pattern to multiple patterns, which is one of the important innovations in this dissertation.(ii) In the optimization modelling, this dissertation first addresses the fundamental multimodal routing problem considering single commodity flow, single transportation service pattern setting and single objective formulation, and constructs a 0-1 integer linear programming model. Based on the proposed concise model, sensitivity analysis on the multimodal routing with respect to the variation of the supply and demand is conducted in order to provide suppliers and demanders with quantitative transportation strategies on multimodal routs planning and multimodal transportation organization.Nextly, according to the node-arc-based modelling method in the fundamental problem,this dissertation further selects two most representative freight categories including conventional containers and hazardous materials as research targets, and integrates multicommodity flow, multiple service pattern setting and fuzzy rail service capacity into the multimodal routing problem for the transportation of the two categories of the goods. First a mixed chance-constrained integer nonlinear programming model is built to describe the complex multimodal routing problem for the conventional container transportation that aims at minimizing the generalized costs of the multiple commodity flows. Followed by analyzing and measuring the social and environmental risk in the multimodal service network with Gaussian Plume Model and Box Model, the environmental risk threshold constrained hazardous materials multimodal routing problem gets formulated by a bi-objective mixed chance-constrained integer nonlinear programming model that involves a multi-criteria optimization for the generalized costs of the multiple hazardous materials flows and the social risk along the planned multimodal routes.(iii) In the solution method design and empirical case optimization, this dissertation focuses on developing an exact solution strategy by identifying and linearizing the nonlinear components of the models, so that an equivalent linear reformulation can be gained to enable the problem to be effectively solved by any classical exact solution algorithms (e.g., Branch-and-Bound Algorithm) in the mathematical programming software(e.g., LINGO). As for the bi-objective optimization for the hazardous materials multimodal routing, the Pareto frontier for the problem can be generated by using the Normalized Weighted Sum Method together with the linearization technique. Finally,large-scale real-world cases are designed in this dissertation to demonstrate the feasibility of the proposed method. Simultaneously, sensitivity analysis and fuzzy simulation are adopted to discuss and summarize the relationship between the optimization results and the model parameters such as confidence value in the fuzzy chance constraint and environmental risk threshold value.