基于智能算法优化支持向量机模型的滑坡稳定性预测  

Prediction of landslide stability based on SVM model optimized by intelligent algorithm

作　　者：胡安龙[1] 王孔伟[1] 李建林[1] 唐芳艳[2] 常德龙[1] 郭振[1] HU Anlong WANG Kongwei LI Jianlin TANG Fangyan CHANG Delong GUO Zhen(Key Laboratory of Geological Hazards on Three Gorges Reservoir Area of Ministry of Education,China Three Gorges University ,Yichang 443002, China College of Economics ＆ Management, China Three Gorges University,Yichang 443002, China)

机构地区：[1]三峡大学三峡库区地质灾害教育部重点实验室,湖北宜昌443002 [2]三峡大学经济与管理学院,湖北宜昌443002

出　　处：《自然灾害学报》2016年第5期46-54,共9页Journal of Natural Disasters

基　　金：国家自然科学基金项目(51309141);水利部公益基金项目(201401029);2015年三峡大学研究生科研创新基金(2015CX036)~~

摘　　要：影响滑坡稳定性的因素较多,利用滑坡稳定性影响因素快速预测滑坡稳定状态是当前滑坡研究的重要内容。利用相关系数、支持向量机、交叉验证法、遗传算法、粒子群优化算法等理论建立支持向量机模型对滑坡稳定性进行了研究。以湖北竹溪县197个滑坡为例,研究结果表明:遗传算法优化的支持向量机滑坡稳定性预测模型预测效果最好,与实际情况吻合得最好。最佳参数c为3.001 6、g为0.041 008,训练集滑坡稳定性预测的正确率为84%,测试集滑坡稳定性预测的正确率为79.32%。因此所提遗传算法优化的支持向量机滑坡稳定性预测模型对于滑坡稳定性分析具有一定参考价值。There are many factors affecting the stability of landslide, predicting the stability state of landslide rapidly through the influence factors of landslide stability is the important part of the study of landslide. Based on use of the correlation coefficient, support vector machines, cross - validation method, genetic algorithm, particle swarm optimi- zation algorithm theories, this paper establishes support vector machine model to study the stability of landslide. Tak- ing 197 landslides in Zhuxi County, Hebei Province as examples, the results show that the genetic algorithm optimi- zation support vector machine landslide stability prediction model works best, and best matches with the actual situ- ation. Optimal parameter e is 3. 0016, g is 0. 041008, the training set landslide stability prediction correct rate is 84%, the test set landslide stability prediction correct rates is 79.3196%. Therefore, genetic algorithm optimization＇s Support Vector Machine landslide stability prediction model has certain reference value for landslide stability analysis.

关 键 词：滑坡稳定性 相关系数 支持向量机 遗传算法 粒子群优化算法 

分 类 号：X43[环境科学与工程—灾害防治] P642.22[天文地球—工程地质学]