Although methods based on bottom-up approaches to delineating urban areas have been developed by focusing on the locations of buildings and road networks, they fail to account for the relationship between the delineated urban areas and the average total management cost of the road networks connecting the buildings. Focusing on building locations, a two-step method is proposed: clustering buildings with intervals that are shorter than a criterion, r_c, before delineating built clusters that are bigger than a criterion, m_c, as urban areas. Although r_c has been optimised in the literature, r_c and m_c have not simultaneously been offered as optimal solutions. This is the motivation for answering the following question: how can we simultaneously optimise r_c and m_c to minimise the average total management cost (per building)? The study’s main conclusions are that: (1) there is an r_c and m_c pair that can minimise the average total management cost and delineate optimal urban areas; and (2) in the empirical study region (Chiba prefecture), the optimal r_c and m_c are 48 m and 42, respectively. Such optimal urban areas are expected to provide urban planners with a norm that can be compared not only with current but also with future images of urban areas.

尽管基于自下而上的城市区域划分方法已经通过关注建筑物和道路网络的位置来开发，但它们未能考虑到所划定的城市区域与连接城市的道路网络的平均总管理成本之间的关系。建筑物。针对建筑物位置，提出了一种两步法：将间隔小于标准rc的建筑物聚类，然后将大于标准m _{c的已建成集群划分}为城市区域。虽然rc已经在文献中进行了优化_，但rc和m _c并没有同时作为最佳解决方案提供。这就是回答以下问题的动机：我们如何同时优化 rc和 m _c以最小化平均总管理成本（每栋建筑）_{？该研究的主要结论是：（ 1}）存在一个rc和m _c对，可以最小化平均总管理成本并划定最佳城市区域；(2) 在实证研究区域（千叶县），_最优rc和m _c分别为 48 m 和 42。这样的最佳城市区域有望为城市规划者提供一个规范，该规范不仅可以与当前的城市区域形象进行比较，而且还可以与城市区域的未来形象进行比较。