| 基于PCA-PSO-SVM的凡口铅锌矿地区岩溶地表塌陷危险性预测 |
| |
| 引用本文： | 周爱红,牛鹏飞,袁颖,黄虎城.基于PCA-PSO-SVM的凡口铅锌矿地区岩溶地表塌陷危险性预测[J].中国岩溶,2020,39(4):622-628. |
| |
| 作者姓名： | 周爱红 牛鹏飞 袁颖 黄虎城 |
| |
| 作者单位： | 1.河北地质大学城市地质与工程学院/河北省高校生态环境地质应用技术研发中心 |
| |
| 基金项目： | 国家自然科学基金资助项目（41807231）；河北省自然科学基金项目（D201903182）；河北省教育厅青年基金项目（QN2019196）；河北省教育厅在读研究生创新能力培养资助项目（CXZZSS2019115）；山西省国土资源厅省级地质勘查项目（SXZDF20170820） |
| |
| 摘 要： | 岩溶地表塌陷是由多个影响因素共同作用导致地面形成塌陷坑（洞）的一种动力地质现象，具有隐蔽性和突发性的特点，常规简单数学模型难以对地表塌陷危险性准确预测。文章先通过主成分分析法（PCA）对选取的地下水位、地下水位波动幅度、给水度等11个影响因素提取5个主成分，对导致地表塌陷危险性的主成分进行全新的解释，同时引入粒子群算法（PSO）优化的支持向量机（SVM）方法，建立PCA-PSO-SVM岩溶地表塌陷危险性预测模型，并结合凡口铅锌矿地区工程实例，将预测结果与单一的SVM模型预测结果进行对比，表明PCA-PSO-SVM危险性预测模型精度更高，可以更好地为岩溶地表塌陷防治工作提供依据。
|
| 关 键 词： | 岩溶地表塌陷 主成分分析 粒子群 支持向量机 危险性预测 |
Prediction of karst surface subsidence risk in the Fankou lead-zinc mine area based on PCA-PSO-SVM |
| |
| ZHOU Aihong,NIU Pengfei,YUAN Ying,HUANG Hucheng.Prediction of karst surface subsidence risk in the Fankou lead-zinc mine area based on PCA-PSO-SVM[J].Carsologica Sinica,2020,39(4):622-628. |
| |
| Authors: | ZHOU Aihong NIU Pengfei YUAN Ying HUANG Hucheng |
| |
| Institution: | 1.School of Urban Geology and Engineering， Hebei Geologic University/Center of Applied Technology Development for Eco-environmental Geology， Hebei Geologic University2.School of Urban Geology and Engineering， Hebei Geologic University3.Shanxi Institute of Geological Survey |
| |
| Abstract: | Karst surface subsidence is a dynamic geological phenomenon with the characteristics of concealment and suddenness， which results from the joint effect many factors. Thus， it is difficult to accurately predict the risk of surface subsidence by the conventional simple mathematical model. In this paper， the Principal Component Analysis （PCA） is used to extract five principal components from 11 influencing factors， including groundwater level， fluctuation range of groundwater level and water supply， so as to make a new interpretation of the principal components leading to the risk of surface subsidence. Additionally， the Support Vector Machine （SVM） method optimized by Particle Swarm Optimization （PSO） is introduced to establish a PCA-PSO-SVM model for prediction of risk of karst surface subsidence. Finally， combined with the engineering example of the Fankou lead-zinc mine， the prediction results by the above proposed model are compared with those obtained by the single SVM model. The results show that the PCA-PSO-SVM risk prediction model has higher accuracy， which can provide a basis for prevention and control of karst surface subsidence. |
| |
| Keywords: | |
|
| 点击此处可从《中国岩溶》浏览原始摘要信息 |
| 点击此处可从《中国岩溶》下载免费的PDF全文 |
