Acta Scientiarum Naturalium Universitatis Pekinensis ›› 2023, Vol. 59 ›› Issue (4): 547-554.DOI: 10.13209/j.0479-8023.2023.043

Previous Articles     Next Articles

Regression Analysis of Building Scale Data and Estimation of Demolition Rate

LIU Debin1, JIN Zeyu2, KE Zhifa1, XU Fangfang3,†   

  1. 1. Center for Data Science, Peking University, Beijing 100871 2. Alibaba Damo Academy, Hangzhou 311100 3. College of Mathematics and Systems Sciences, Shandong University of Science and Technology, Qingdao 266590
  • Received:2022-06-24 Revised:2022-09-20 Online:2023-07-20 Published:2023-07-20
  • Contact: XU Fangfang, E-mail: xuff(at)


刘德斌1, 金泽宇2, 柯志发1, 徐芳芳3,†   

  1. 1. 北京大学大数据科学研究中心, 北京 100871 2. 阿里巴巴达摩院, 杭州 311100 3. 山东科技大学数学与系统科学学院, 青岛 266590
  • 通讯作者: 徐芳芳, E-mail: xuff(at)
  • 基金资助:
    国家重点研发计划(2018YFC0704300)和国家自然科学基金数学天元基金项目(12126348, 12126370)资助


Since the statistical data of the completed area in different statistical yearbooks are quite different, this paper proposes a weighted regression model to analyze the statistical values of the completed area in order to analyze and confirm the data sources by combining the relevant data in the statistical yearbooks, such as urban land demand, social development level, etc. On the other hand, there are only the cumulative demolition area in the statistical yearbooks, and the annual statistical data of demolition area is missing. To estimate the demolition rate, a constrained regression model is established to calculate the annual demolition area of each province based on the national population census data, urban land demand, social development level, and other relevant data. The numerical results can provide a reference for understanding the actual situation of building scale in China over the last two decades, as well as making policies of energy conservation and emission reduction.

Key words: building scale data, completed area, demolition area, demolition rate, regression analysis


针对不同统计年鉴中竣工面积统计数据差异较大的问题, 结合统计年鉴中关于城市用地需求和社会发展水平等相关数据, 运用加权回归分析模型, 分析不同数据源民用建筑的竣工面积统计值, 并对数据源进行分析与确认。针对全国及分省只有累积拆除面积数据以及缺失逐年拆除面积统计值的问题, 利用全国人口普查、城市用地需求和社会发展水平等相关数据, 建立约束回归分析模型, 提出分省逐年拆除面积计算方法, 估算拆除面积和拆除率。研究结果可为确定我国近20年建筑规模的实际情况和相关节能减排政策的制定提供科学参考。

关键词: 建筑规模数据, 竣工面积, 拆除面积, 拆除率, 回归分析