Evaluation of urban underground space via automated constraint identification and hybrid analysis

As urbanization progresses, the exploration and development of urban underground space resources have become imperative. Assessments of urban underground space are conducive to understanding the quantity and potential of urban underground resources that can be developed and utilized. The practice of combining urban three-dimensional geological models for suitability assessments of urban underground space is commonly used and has been proven to be effective. However, existing assessment methods struggle to automatically consider the impact of existing facilities, necessitating extensive preliminary investigation. In addition, despite the abundance of evaluation methods, there is a noticeable gap in research applying both subjective and objective evaluation techniques in tandem. To address these problems, this study introduces an innovative framework to automate the identification of existing constraints via a semantic segmentation deep learning method. In addition, a hybrid evaluation method integrating the entropy weight method, the CRITIC method, and the Analytic Hierarchy Process is proposed. This methodology not only fills a gap in existing studies by providing a comprehensive framework for urban underground space evaluations but also offers a novel approach to integrating technological advances into urban planning research. The application of this study in the Sanlong Bay area of Foshan City further demonstrates its practicality and effectiveness, showcasing a significant advancement in the field of urban underground space evaluation.