As an emerging shared mobility option, bike share has the potential to improve transportation sustainability. Understanding the mobility pattern, environmental benefits, and impact of system changes helps improve bike share systems (BSSs). However, existing literature about has the following gaps: (1) there is a lack in detailed understanding about bike share’s travel patterns (2) few studies have considered the heterogeneous travel mode choices to quantify BSSs’ environmental impacts; (3) the station interactions in the system expansion process are not well studied; and (4) very few studies have quantitatively compared the user experience and operational challenges in different types of BSSs.
This dissertation aims to address these gaps to provide better understanding of the travel patterns, benefits and impact of system changes in BSSs to assist the policy making and development of BSSs. To achieve this objective, various modeling frameworks and methods were developed. (1) The statistical property of bike sharing trip distance and duration are first analyzed to provide fundamental basis for the modelling of BSSs. (2) A Bike Share Emission Reduction Estimation Model (BS-EREM) is proposed to quantify the environmental benefits from BSSs. The BS-EREM estimates the transportation modes substituted by bike share trips, with the consideration of heterogeneous travel mode choices. The GHG emission reductions contributed by eight case study BSSs were then evaluated using BS-EREM. (3) For system expansion, the competition/complement interactions between stations are revealed using a segmented regression model. The study also shows that incorporating features about such interactions significantly improves the demand predictions for system expansion. A Spatial Eccentricity Quantile based Ensemble Model (SEQEM) is proposed to identify a spatial range that the station interactions take effects. (4) A comprehensive stochastic simulation framework is proposed to evaluate the user experience and system operations in different types of BSSs, which estimates actual origins-destinations of travel demands and integrates the user behavior model and rebalance optimization model. The case-study results reveal that the user rerouting behaviors can indirectly affect system performances. Overall, systems with high usage intensity can benefit from transitioning their station-based systems into hybrid systems.
In summary, this dissertation provides a holistic understanding of the mobility patterns, environmental benefits, and the impacts of system expansion and system types, which assists the policy making and the development of BSSs. The proposed models are transferable to different cities to support the development of sustainable micro-mobility systems.