节点文献
基于聚类集成的地下空间地质环境质量三维评价
3D evaluation of geological environment quality in underground space based on clustering ensemble
【摘要】 城市地下空间开发利用是解决城市土地资源紧缺的重要手段,地下空间地质环境质量评价是地下空间合理安全利用和降低开发风险的前提和保障。为了降低评价过程中的主观性和评价结果中多种评价指标交叉交融的不确定性,文章基于三维地质模型,采用多种聚类模型的聚类集成算法对地下空间地质环境质量进行评价。利用K-means、高斯混合模型、自组织神经网络等聚类模型计算结果,结合重标记法的聚类集成算法实现地质环境质量评价。以厦门市某区为例,基于三维评价指标信息,利用上述分析方法进行评价,并与层次分析法结合多级指数叠加法评价结果进行对比分析。结果表明,基于聚类集成的评价方法能够有效应用于地下空间地质环境质量三维分类及评价研究,相关评价结果可以更客观地为地下空间的安全合理开发提供支持和保障,更好地服务于城市地下空间的建设规划和可持续发展。
【Abstract】 The development and utilization of urban underground space is an important means to solve the shortage of urban land resources. The evaluation of geological environment quality of underground space is the premise and guarantee of safe and rational utilization of underground space and reduction of development risks. In order to reduce the subjectivity in the evaluation process and the uncertainty of the cross integration of multiple evaluation indices in the evaluation results, this paper uses the clustering ensemble algorithm based on multiple clustering models to evaluate the geological environment quality of underground space based on the three-dimensional(3D) geological model. Using the results of K-means, Gaussian mixture model, self-organizing neural network and other clustering models, combined with the clustering ensemble algorithm based on re-labeling method, the geological environment quality evaluation is realized. Taking a district in Xiamen City as an example, based on the 3D evaluation index information, the above analysis method is used for evaluation, and compared with the analytic hierarchy process(AHP) combined with the multi-level index superposition method. The results show that the evaluation method based on clustering ensemble can be effectively applied to the 3D classification and evaluation of geological environment quality of underground space, and its related evaluation results can provide support and guarantee for the safe and rational development of underground space more objectively, and better serve the construction planning and sustainable development of urban underground space.
【Key words】 underground space; self-organizing neural network; K-means algorithm; Gaussian mixture model; clustering ensemble; three-dimensional(3D);
- 【文献出处】 合肥工业大学学报(自然科学版) ,Journal of Hefei University of Technology(Natural Science) , 编辑部邮箱 ,2025年01期
- 【分类号】TU984.113;X141
- 【下载频次】25