A bi-objective model for bus transit network and lane reservation integrated optimization

Abstract

Bus transit network design is one of the most important strategic planning problems in urban public transportation, which is to define the number of bus lines and their operating paths such that a given origin-destination demand can be covered. Although rapid and reliable bus transit service has increasingly become one of the most key and important factors attracting passengers to travel through buses, especially for large cities in China, most of the studies in the literature ignore such issue along with bus transit network design. This study investigates a new bi-objective bus transit network design problem taking into account bus lane reservation aiming to provide rapid and reliable transit service. Moreover, both the total travel time of all passengers including penalties for the transfers and the total negative impact caused by bus lanes are minimized. For the problem, a bi-objective integer non-linear programming model is first developed. Then, it is equivalently reformulated as an integer linear program, which is further tightened by adding explored valid inequalities. The problem is shown to be NP-hard. Finally, an exact $\varepsilon$-constraint method is proposed to yield the Pareto front. Extensive numerical experiments on benchmark and randomly generated instances confirm the effectiveness and efficiency of the proposed approach.