Search

Rail transportation has experienced a rebirth in the last few decades, and a very large investment will certainly be made in new railway lines in the years to come—especially in high-speed rail lines. The success of such investment is heavily dependent on rail ridership, which in turn is dependent on the location of railway stations. In this paper, we present a mixed-integer optimization model that determines the optimal location (and number) of stations along a railway line that will be introduced over an existing transportation network. The stations are chosen within a set of possible locations defined a priori according to the objective of maximizing travel cost savings. The model takes into account the sensitivity of rail ridership to time losses because of stops at intermediate stations, as well as (static) competition from other modes. The practical usefulness of the model is illustrated with a case study involving a high-speed rail line expected to be built in Portugal in the future: the Lisbon-Porto line.

This paper considers the problem of predicting on-time performance on a partially double-tracked rail line with scheduled traffic. In particular, the model presented herein explicitly incorporates the uncertainty in train departure times along with more complex line structures than have been considered previously. Extensive Monte Carlo simulations are used to validate the predictions made by the proposed model under an optimal meet/pass planning process.

The length of a rail line that will minimize total transportation (user and operator) costs, and the threshold demand necessary to ensure that the resulting length of the rail line is nonzero are investigated for an urban transportation corridor. A density of demand for travel to the central business district (CBD) represented by a general function P(x) passengers per unit length of the corridor where x is the distance from the CBD is considered. The line-cost as a function of x, the costs of the rail fleet, rail and bus operating costs and passenger time costs are also considered in the formulation. The fleet size is formulated considering the peaking of demand relative to time. When the line cost is nonuniform there could be several line lengths at which the total transportation cost is minimized or even maximized locally. When the line cost per unit length is uniform, a minimum transport cost rail line of nonzero length exists only if the net gain in travel time and operating cost of transporting the total ridership a unit distance by rail, when compared to bus, exceeds the marginal line and fleet costs per unit length. In either case, the minimum transport cost rail line length can be determined easily. The effects on the line length of shifts in demand are investigated. Closed-form solutions for the line length are obtained for the cases of sectorial and rectangular corridor-sheds with uniformly distributed demand per unit area.

The optimum zone structure for a planned urban commuter rail line with one to many or many to one type demand during peak periods is analyzed. The objective is to minimize the passenger time costs as well as the relevant system operating and capital costs. The analysis is based on the interaction of a range of variables such as the number of zones, zone boundaries, uniform station spacings, fleet size and train headway. The approximately optimum values of the decision variables are determined mostly in closed form. A numerical application and sensitivity analyses are presented.