基于聚类算法的传统街区建筑分类及风貌改造策略研究——以董市镇老正街为例  

Character Rehabilitation Strategies and Traditional Block Building Classification Based on Clustering Algorithm:A Case Study of Laozheng Street in Dongshi Town

作　　者：黄艳雁 吾提克尔·吐尔逊 涂瑞鑫 张旭 陈小康 Huang Yanyan;Tuerxun Wutikeer;Tu Ruixin;Zhang Xu;Chen Xiaokang

机构地区：[1]湖北工业大学土木建筑与环境学院

出　　处：《华中建筑》2025年第4期120-123,共4页Huazhong Architecture

基　　金：湖北省教学研究项目“迈向新工科的建筑教育全过程‘通专融合’人才培养模式建构研究”(编号:2020451);枝江市董市镇“擦亮小城镇”建设项目(编号:ZJGJS0475-202202-01-2)。

摘　　要：该研究聚焦于快速城镇化背景下传统街区建筑风貌的保护与改造问题。鉴于传统街区建筑风貌改造的复杂性和限制性,以董市镇老正街街区为例,选取10个建筑风貌特征,从类型学的观点出发,统计老正街街区556栋建筑的风貌特征数据,并运用K-means聚类算法对建筑进行聚类分析,最终识别出4类建筑聚类。针对各聚类的特点和存在的问题提出了针对性的风貌改造策略,旨在为传统街区建筑风貌改造提供科学化分类与新的研究视角和实践路径。This study focuses on the protection and rehabilitation of traditional block architectural character under the background of rapid urbanization.Given the complexity and limitations of the transformation of traditional block architectural character,using the Laozheng Street block in Dongshi Town as a case study,ten architectural character features are selected.From a typological perspective,the architectural character feature data of 556 buildings in Laozheng Street block are statistically analyzed.The K-means clustering algorithm is applied to conduct clustering analysis on the buildings,ultimately identifying four classes of architectural clusters.Targeted character transformation strategies are proposed for each cluster based on their characteristics and existing issues,aiming to provide a more scientific classification and new research perspectives and practical paths for the transformation of traditional block architectural character.

关 键 词：聚类算法 传统街区 风貌改造策略 特征指标 

分 类 号：TU984.18[建筑科学—城市规划与设计]