| 摘要: | The rapid growth of urban populations and private vehicle ownership has exacerbated
many cities' traffic congestion and environmental degradation. Shared micromobility
services, including bike-sharing (BSS) and electric moped-sharing systems (EMSS), have
emerged as viable solutions to these issues by providing sustainable and flexible
transportation options. However, these systems face significant operational challenges,
including spatial and temporal imbalances in supply and demand, as well as suboptimal
infrastructure placement. While existing research has explored the influence of built
environment factors on shared micromobility usage, most studies rely on global models
that assume spatially uniform relationships, often overlooking the impact of spatial
heterogeneity. In addition, existing spatial optimization models for locating shared
micromobility stations typically evaluate demand coverage based on individual stations,
without considering the dual-node nature of user trips. This simplification may result in
substantial overestimation of system performance. Moreover, for EMSS, existing research
has primarily focused on placing battery swapping stations, with limited attention given to
the integration of parking and battery-swapping functions. Such integration is crucial for
enhancing both user experience and operational efficiency, yet it remains underexplored in
existing research. This dissertation addresses these gaps through three empirical studies on
BSS and EMSS in Taipei. The first study examines the impact of built environments on
shared micromobility usage using the Multiscale Geographically Weighted Regression
model to capture spatial heterogeneity. The second study develops the Flow Termini
Coverage Model (FTCM), incorporating origin-destination flow dynamics to improve the
siting of bike-sharing stations. By accounting for both trip ends, the FTCM offers a more accurate representation of travel behavior than conventional single-node coverage models.
The third study introduces the concept of EMSS hubs, multifunctional facilities that
integrate moped rental, return, and battery-swapping, and proposes the EMSS Hub
Location Problem (EHLP) model to support strategic hub placement. The EHLP provides
a comprehensive framework for EMSS infrastructure planning by incorporating three types
of demands. Collectively, these studies advance the shared micromobility field by
developing novel spatial analytical methods and optimization models. The findings offer
important implications for policy and practice, demonstrating how advanced spatial
methods can support more effective and sustainable shared micromobility planning. |
