Public transit services are important assets to any major city. They improve mobility for all, reduce car dependence, and therefore the need for further highway expansion. It is ideal if the PTS are financially sustainable, with affordable fares and expedient quality. The key lies in the how the urban transit system are planned to ensure their financial sustainability. In transit system planning, two major levels of planning activities are involved: strategic and tactical levels. This thesis covers both of them by considering two important factors on each level: the interaction between the urban development and transit sketch plan in the strategic level and the transit scheduling and routing design taking into account the differential service types in the tactical level.
It is known that urban density is positively correlated with transit service usage. Previous studies, however, analyzed the problem descriptively based on statistical approaches. This thesis selects rail transit service as case study and seeks to derive some prescriptive results of the relationship between urban density and rail service planning. It considers an idealized metropolitan region with a central business district (CBD) at its center, whose population is distributed according to the density saturation gradient pattern. Trips generated from the region to the CBD are either served by the rail service supplemented with feeder buses, or by autos or taxis. We study the sensitivity of urban development density on the financial sustainability of the rail service by examining the supply and demand patterns. Through the analysis, the result sheds light on the threshold urban density required, below which the service cannot be sustained without subsidy. The results provide guidelines for planning urban developments with sustainable rail services.
When modeling the transit service demand pattern, this thesis considers the competition and trip makers' mode choices between transit service and autos (or taxi). The autos are first assumed to be free of traffic congestion and the travel speed for autos is uniform. Then, traffic congestion effect for autos is incorporated. The traffic demand assignment between the two modes reaches an equilibrium, which is captured by solving equivalent differential equations in this thesis.
Routing and scheduling designs for transit services are important aspects for the operational planning of the transit services. This thesis applies the ferry services in Hong Kong as a case study and investigates a multi-fleet ferry routing and scheduling problem that takes into account ferry services with different operation characteristics and passengers with different preferred arrival time windows. A logit model is used to represent passengers' service choices. The full problem is formulated as a mixed integer nonlinear programming problem and solved with a heuristic procedure that first fixes the demand and then decomposes the resultant model by ferry services. At each iteration of the algorithm, the demand is updated and the relaxed problem is re-solved. Numerical results for the case of ferry service network design in Hong Kong are provided to illustrate the properties of the model and the performance of the heuristic.
On the pursuit of the lower bound of the previous routing and scheduling problem, a method of converting a nonlinear programming into an equivalent mixed-integer linear program is developed. It successfully finds the application in getting the global optimum for continuous road network design problem (CNDP). CNDP, typically formulated as a bi-level program or a mathematical program with equilibrium constraints, is generally non-convex. The non-convexity stems from both the traffic assignment equilibrium conditions and the nonlinear travel time function. In this thesis, we first formulate the continuous network design problem as a single-level optimization problem with equilibrium constraints, and then we transform the equilibrium constraints into a set of mixed-integer constraints and linearize the travel time function. The final result is that we cast the continuous network design problem with equilibrium flows into a mixed-integer linear program, which possesses the desirable property of global optimality.