Optimization of grid bus transit systems with elastic demand

Abstract

Current analytic models for optimizing urban bus transit systems tend to sacrifice geographic realism and detail in order to obtain their solutions. The models presented here shows how an optimization approach can be successful without oversimplifying spatial characteristics and demand patterns of urban areas and how a grid bus transit system in a heterogeneous urban environment with elastic demand is optimized. The demand distribution overthe service region is discrete, which can realistically represent geographic variation. Optimal network characteristics (route and station spacings), operating headways and fare are found, which maximize the total operator profit and social welfare. Irregular service regions, many-to-many demand patterns, and vehicle capacity constraints are considered in a sequential optimization process. The numerical results show that at the optima the operator profit and social welfare functions are rather flat with respect to route spacing and headway, thus facilitating the tailoring of design variables to the actual street network and particular operating schedule without a substantial decrease in profit. The sensitivities of the design variables to some important exogenous factors are also presented.

Planning, operation, and control of bus transport systems: A literature review

“…To avoid oversimplifying urban areas, the authors propose an analytical model which assumes a rectangular grid and zone dependent passenger demand. Chien and Spasovic (2002) consider elastic demand, i.e. demand dependent on the level of service to determine the line and stop spacing, headways, and fares.…”

Optimization of demand-responsive transit systems using zonal strategy

“…Li and Quadrifoglio (2011) developed an analytical model to determine the optimal number of service zones in a residential service area. Chien and Schonfeld (1997) proposed a modelling approach to optimize the design and operation of bus transit systems in a realistic geographic environment, while Chien and Spasovic (2002) developed an optimization approach to deal with practical geographic variability.…”

A multiobjective planning model for intercity train seat allocation

“…As such, the model can be used for railway systems of multiple lines or networks (e.g. Assad, 1980;Tong and Wong, 1999;Chien and Spasovic, 2002), where the train service plan and travel demand are given. As the model can work on a single train run (independent of other scheduled runs) for a given service plan and travel demand, an optimal seat allocation plan can be generated in a short amount of computer time, normally within minutes.…”