多源城市数据驱动下城市设计中的空间句法模型校核及应用研究

Space Syntax Model Verification and Application in Urban Design with Multi-source Data

空间句法在我国城市设计研究及实践中的应用方兴未艾,然而其建模计算和辅助分析决策过程中带来的误差并未被充分研究。文章结合POI、大众点评等多源城市数据,以2017年珠三角部分城市及全国公路交通网络为实验对象,从空间句法模型校核的理论基础、空间句法分析的误差来源、处理误差的技术细节、数据驱动下基于空间句法的城市设计流程等方面探讨了空间句法的模型校核及其在城市设计中的应用问题。研究发现,空间句法的误差来源主要包括搜索半径误差、建模范围误差、建模细度误差和参数算法误差4种,而在多源城市数据的辅助下,可通过选取最优半径、适度外扩模型、选取合适细度和建模软件等方法减小误差。文章总结了数据驱动下基于空间句法的城市规划设计框架,认为多源城市数据在空间句法模型校核、空间句法辅助方案设计等阶段具有广阔的应用前景。

Space syntax has been used in urban design research and practice and has kept growing although its errors in model building and analysis aided decision-making were not fully studied. With multi-source data from POI and www.dianping.com, the paper studies transport of several cities in Pearl River Delta and national transport network, studies the verification and application of space syntax model in urban design from theoretical basis, error sources, technical details, urban design process with multisource data. The result shows that errors of space syntax are mainly from searching radius, modelling range, modelling accuracy,and parameter algorithm. The above errors could be diminished by the aid of multi-source data. The paper concludes urban design framework based on space syntax with multi-source data, and argues that multi-source data has wide application in model verification and aided design.