Optimizing fare structure and service frequency for an intercity transit system

This study presents an approach to jointly optimize service headway and differentiated fare for an intercity transit system with an objective of total profit maximization and with consideration given to the economic and social sustainability of the system. Service capacity and fleet size constraints are considered. The optimization problem is structured into four scenarios which are comprised of the combinations of whether the Ranges of Travel Distance (RTD) is fixed or variable and if the time period is for a single period or for multiple periods. A successive substitution method (specifically, a modified Gauss Southwell method) is applied to solve for the optimal solutions when the RTD is considered fixed, while a heuristic solution algorithm (specifically, a Genetic Algorithm) is developed to find the optimal solutions when the RTD is considered to be optimized. The methodology discussed in this dissertation contributes to the field of transportation network modeling because it establishes how to solve the fare and headway design problem for an intercity transit system. Intercity transit agencies are faced with the challenge of determining fares for a very complicated setting in which demand elasticity, realistic geographic conditions, and facility locations of the transit system all must be taken into account. A real world case study - Taiwan High Speed Rail is used to demonstrate the applicability of the developed methodology. Numerical results of optimal solutions and sensitivity analyses are presented for each scenario. The sensitivity analyses enable transit planners to quantify the impact of fare policies and address social equity issues, which can be a major hurdle of implementing optimal fare policy to achieve maximum profit operation. According to the sensitivity analysis, the total profit surfaces for various headways, fares, and RTD are relatively flat near the optimum. This indicates that the transit operator has flexibility in shifting the solution marginally away from the optimum without significantly reducing the maximum profit. By varying the elasticity parameters of fare and demand one can observe how these variables affect the optimized RTD. The results indicate that as the elasticity parameters of fare increase or demand decreases, the optimal number of RTD increase while the boundaries of RTD are concentrated in the range of shorter travel distances