臺北市路邊停車研究: 以空間分析探討停車收費政策影響

Abstract

停車在都市規劃和政策決定過程中扮演著重要的角色，而一個較不完善的停車系統容易導致交通擁堵、空氣污染和其他問題，特別是在人口稠密的城市地區。過去數十年來學者也不斷討論停車行為會受各種因素影響。其中，調整停車費是最常見的實施措施，因此台北市政府於2015年推動全面收取路邊停車費的新政策，目的即希望能改變使用者停車行為，增加路邊停車周轉率，解決住宅區內車位被長期占用的問題。在本研究中，主要引用臺北市停車工程管理處公布的年度調查報告進行分析討論，其中涵蓋不同交通分區（TAZ）中的路面及路外停車需求、供給和違規停車數量。在分析停車模式時，透過應用空間自相關獲取停車費價格和停車比率的空間異質性。由於原本實施全市停車政策具其局限性，需要透過本地觀點進一步討論，利用地理加權回歸（GWR）可補充討論各停車分區間潛在停車因素與停車行為之間的關係。另外，優化加權函數的帶寬及考量混和參數的方式可改進估計模型，並解釋不同因子間的差異性。本論文主要採用了考慮主要變量、條件和特徵的方法，以便最終根據可解釋的單位調整停車政策，改善台北市的停車情況。最終，這種系統性分析也可望在其他城市實施，以從空間視角輔助政策審查分析。

Parking is extraordinary to be discussed during urban planning and policy-making, since a worse parking system might lead to traffic congestion, air pollutions and other problems, particularly in densely populated urban areas. It has been debated for decades that parking behavior is affected by various factors. Among those, adjustment on parking fee is the most common measure to implement, which was also taken as a new policy to generally charge on-road parking in Taipei in 2015. In this study, the main analyzing dataset was acquainted by a published annual governmental survey, covering numbers on the demand, supply, and illegal parking in different traffic zone (TAZ). Through the application of spatial autocorrelation techniques, it is possible to identify the similarity, and capture the spatial heterogeneity of the patterns of price and parking ratio in overall situation. Due to the limitation of responding a city-wide parking policy, it requires a further discussion on local perspectives, which can be examined through a Geographically Weighted Regression (GWR). This thesis examines the relationship between potential factors, showing the influence level of different variables. Additionally, other advanced technique which target at optimizing weighting kernels and consider different scales of parameters are also applied to improve the estimated models. With these calculated coefficients, the major objective of this thesis, an approach taking into account dominating variables, conditions and characteristics is implemented in order to ultimately adjust the parking-policy based on interpretable units and to improve the parking situation in Taipei City. Eventually, this structural analysis is also expectable to be implemented to other cities to assist the policy review in a spatial perspective.