直交及非直交幹道架構的都市空間結構動態模擬

都市發展的型態會受到道路系統建設的影響，形成了線性都市(linear city)、星型城市(star city)與矩形網格城市(rectangular grid city)等，傳統都市動態理論著重於描述性討論如Friedman(1966)的核心－腹地理論(Core-Periphery Theory)，Doxiadis（1969；源自鄧清乾，1983）的理想動態都市型態(Ideal Dynamic Urban Style)與Tanabe(1970)的渦輪多角發展(Turbo-cornutus)。Batty(1994)以生命遊戲(Live Game)建立都市空間動態模擬(urban spatial dynamic simulation)的基礎，後繼學者使用細胞自動體(Cellular Automata (CA))作都市發展的模型。標準的CA缺乏都市經濟理論，它很難建立核心們層級性關係及展現城市的租金狀況。經觀察多核心城市(multi-core city)的變化，本研究考慮不同道路層級的運輸成本建立初步的架構，並以需求門檻(threshold demand)，近鄰模式(neighborhood model)與效用函數(utility function)在網格地理資訊(grid geography information)中的系統架構下建立多核心城市的模擬模式。它發現接近渦輪多角(Turbo-cornutus)與中地理論(Center Place Theory)（Christaller, 1933；源自薛益忠，2006，p.142)）的城市發展。分析發現幹道的對稱現象與周邊的凸出而較長使核心發展位於幹道的機會變大，有通勤需求下次核心－腹地對高層核心的環繞仍存在，幹道運輸成本低使腹地範圍大及向外凸出的腹地大使幹道上易產生高層級核心，受租金環線影響非直交道路的核心易隨著環狀道路發展。

關鍵字

空間結構；競標地租；細胞自動機；需求門檻；直交與非直交道路

並列摘要

The patterns of city development are susceptible to be impacted by the road frame to form linear city, star city, and rectangular grid city and the like. The traditional theories of urban spatial dynamics often focused on descriptive discussions, such as Friedman's (1966) Core-Periphery Theory, Doxiadis' (1969) Ideal Dynamic Urban Style of and the Turbo-cornutus of Tanabe (1970). Batty (1994) utilized the Live Game to create the foundation of urban spatial dynamic simulation, and subsequent scholars adopted the Cellular Automata (CA) modeling as an urban development model. With standard CA lacking urban economics theories, it was highly difficult to create a core's hierarchy correlation and to demonstrate the sate of urban rentals distribution. By observing multi-core city changes, this study attempts to instill a preliminary framework by taking into account the transportation cost on varied roadway hierarchies, and to establish a multi-core city simulation model under a grid geography information system utilizing the threshold demand, neighborhood model and utility functions. It discovers the city of Turbo-cornutus and Center Place Theory (Christaller, 1993) in urban development. The analysis indicates that a majority of cores distributed on arteries, the highest hierarchy core shift from the original spots to other places, newer arteries replaced the old ones, and the reasons why the number of cores on artery became more, and larger in scale.