摘要
由于康泽恩城市形态学划分形态区域的方法在量化上的不足,适用规模、效率以及客观性有待提升,针对形态区域依据的三个要素——平面类型、建筑类型和土地利用,分别提出了量化分析的措施。其中平面单元主要采用空间句法,并对组构值的选择和计算进行了比较、辨析;最后应用K均值聚类法结合GIS技术进行整合、可视化。使用所构建的方法对典型的广州传统街区进行了实证研究。能为城市形态研究的继续深入提供数据支撑和可个性化调整的研究方法。受限于研究领域,技术方法还有进一步优化的空间。
The Conzenian School, as one of the leading schools of western urban morphology, is highly relevant to urban development and historical protection in China. Nevertheless, morphological region, the core achievement of the Conzenian School, lacks nuance and thus influences objective reasonability. Therefore, this study aims to quantitatively improve morphological regionalization by combining new technological means,attempting to improve objectivity and application efficiency.Although classical works of the Conzenian School have developed conscious quantitative morphologies, the scope of their descriptive statistical analysis remained small due to the limitations of artificial efficiency and calculation tools. So far, quantitative exploration continues to face limitations in single element quantitative model and computation methods, as well as shortages in multi-element integration methods.Therefore, from the three elements of morphological regionalization described by the Conzenian School(plane type, building type, and land use),this study firstly analyzed keys in the quantification process of elements one by one and described correlations. Secondly, quantification of morphological elements was performed by combining with relatively mature single-element quantification methods such as space syntax, thus forming a morphological database. Subsequently, data was optimized and integrated by centering a K-means clustering method. Data was connected to space using GIS to realize visualization alongside further adjustment and iteration. In this way, a relatively complete morphological regionalization quantification model was established.Specific research areas have their own appropriate quantitative index systems. Indexes that can distinguish different types and morphologies in specific regions will be chosen according to local characteristics and status data. Hence, this research utilized a case study based on a traditional block in Guangzhou. It also analyzed specific quantitative indexes from three morphological elements. During quantification of plane type, the space model,configurational index selection, and calculation methods of space syntax were also analyzed. Moreover, a more reasonable connection and assignment method was provided, which laid foundations for element integration.Finally, the Zhuangyuanfang Block was chosen in the case study for database construction, data integration, and visualization. Correlations between main expressions of data reported by clustering and spatial forms were analyzed, which provided an opportunity to interpret data integration results of morphological elements. Based on the results, the morphological regions gained from this quantitative regionalization method were similar with those of classical methods, thus verifying the feasibility of the quantitative regionalization method on the microscale. This method could provide not only data supports and research paths that can be adjusted for continuous studies on urban morphology, but also more objective morphological regionalization for plan formation.Nevertheless, the studied quantitative method gave little attention to quantification of historical factors based on status data, resulting in some differences with classical methods in term of high-level morphological regionalization. The K-means clustering method also uses some subjective factors in determining number of clusters. As a result, the proposed quantitative method continues to have space for optimization and further exploration.
作者
朱知麟
田银生
ZHU Zhilin;TIAN Yinsheng
出处
《南方建筑》
CSCD
北大核心
2022年第9期8-18,共11页
South Architecture
基金
国家社会科学基金资助重大项目(18ZDA161):西南地区传统村落保护与利用研究。
