Institute of Transportation, MOTC

       The rising popularity of bicycle-sharing systems as an environmentally friendly and congestion-reducing means has raised awareness of the need for well-structured bikeway networks. A cycling network with isolated and continuous bikeways can increase cyclists’ willingness to cycle and thus cycling demand. So, designing a bicycle network to maximize cycling demand becomes essential under the designated budget. This thesis proposes a time-dependent mixed bicycle station and bikeway network design problem with demand elasticity, which aims to maximize the demand coverage of all periods. The problem is formulated in terms of multiple periods because in reality, a bikeway network needs to be completed in multiple years while the funding for constructing a bikeway system is often allocated periodically. The proposed demand elasticity expresses the willingness to cycle as a function of bicycle station access times to bike stations and the total travel time. A genetic algorithm with a station-based elimination heuristic and tailor-made mutation operators is proposed to solve the problem by determining the subset of bicycle stations and bikeways that covers the largest total demand served within a given budget in periods. Numerical studies show that the proposed demand elasticity function can be an effective tool to reflect the effects of travel time and budget level on served demand. In the multi-period design, for both two-period and three-period cases, an earlier high investment can achieve a higher served demand. Moreover, when budgets in later periods allow the opening of new stations, these stations reduce the total travel time of OD pairs and improve the number of served demand