市區道路橫斷面空間配置最佳化

OPTIMIZING CROSS-SECTIONAL SPACE ALLOCATION OF URBAN STREETS

市區道路的寬度及橫斷面配置決定一個城市發展的效率性及永續性。然而市區道路的寬度和橫斷面配置常因都市規劃者和交通工程師分別決策而缺乏整合性及一致性。因此，如何根據毗鄰土地使用狀況及道路在路網中所屬之道路定位，設計最佳的道路寬度及橫斷面配置是相當重要的議題。基此，本研究發展一個道路橫斷面配置最佳化模式，在給定道路寬度的情況下，同時考慮人本、安全、效率、環保及延續五個面向的目標，並求解汽車道、機車優先道、自行車道、人行道以及取決於道路等級的道路附屬設施（如中央分隔島、公共設施帶）之最佳配置數量及寬度。藉由設定0.25公尺作為道路元素設計之基本寬度，將原先混合整數規劃模式，轉換成純整數規劃模式，並透過簡單加權法以及層級分析法取得之不同土地使用（如商業區、住宅區、工業區等）各目標權重組合，以窮舉法和套裝軟體LINGO進行求解。為展示模式之應用性，本研究以北投士林科技園區為例，挑選園區內三個路段（承德路、福國路、承平路）進行模式求解與比較。結果顯示承德路目前的人行空間配置不足，將無法滿足未來園區開發完成時之行人需求；福國路延伸段建議未來可單向配置四車道、人行道淨寬4.75公尺；至於承平路，則建議可單向配置二車道及自行車道。結果證明本研究提出之模式可以針對五個面向設計最佳化之道路橫斷面配置。

The width and cross-sectional allocation of city streets determine how efficiently and sustainably a city can move forward. Unfortunately, the width and allocation of urban streets are usually separately determined by urban planners and traffic engineers lacking of integration and consistency. Therefore, how to optimally design the width and cross-sectional allocation of city streets simultaneously according to their neighboring land use and level in the city roadway network is particularly essential, which can provide valuable guidelines for urban planning review process. Based on this, this study aims to develop an optimization model for cross-sectional allocation of urban streets under given road width. The model simultaneously optimizes five objectives, including human-oriented, safety, efficiency, environmentally friendly and consistency. The model optimizes the numbers and widths of lane types (for motor vehicles, motorcycles, bikes and pedestrians) as well as road ancillary facilities in association with the road level (i.e., median, public facilities strip). By setting a minimum design unit of 0.25 m, the mixed integer model can then be converted into a pure integer programming model. By using the simple weighted sum method and sets of weights for various land use types (commercial, residential, industrial…) obtained by Analytic Hierarchy Process (AHP) survey, the model are solved by total enumeration method and package software (LINGO), respectively. To show the applicability of the proposed model, a case study on Beitou-Shilin Technological Park is performed and three selected streets (Chengde Road, Fuguo Road, Chengping Road) in the park is chosen for comparisons. The results of Chengde Road show that the current allocation lacks of pedestrian space and fails to meet the future pedestrian demand. The model suggests an allocation of four motor vehicle lanes and a 4.75m sidewalk for the extension of Fuguo Road. As for Chengping Road, it suggests an allocation of two motor vehicle lanes and a bike lane. The results suggest that the proposed model can design an optimal cross-sectional allocation towards five objectives.