Modelling rules of landslide susceptibility prediction considering the suitability of linear environmental factors and different machine learning models
-
摘要: 对于滑坡易发性预测中的水系、公路和断层等线状环境因子, 现有研究大多采用缓冲分析提取距离线状因子的距离。但缓冲分析得到的线距离属于离散型变量, 带有大小不等的随机波动性且对点或线要素的误差较为敏感, 导致滑坡易发性建模精度下降。提出了使用水系和公路的空间密度等连续型变量改进线状环境因子的适宜性。以江西省安远县为例, 选取高程、地形起伏度、距水系和公路距离等14个环境因子(原始因子), 再将距水系和公路距离2个线状因子改进为水系密度和公路密度(改进因子); 之后采用逻辑回归、多层感知器、支持向量机和C5.0决策树等机器学习模型, 分别构建了基于原始因子和改进因子的机器学习模型以预测滑坡易发性; 最后利用ROC曲线和易发性指数分布特征等来研究建模规律。结果表明: ①改进因子机器学习预测精度均高于原始因子机器学习模型, 表明空间密度对于易发性预测的适宜性更好; ②在4类机器学习模型中C5.0模型对于滑坡易发性预测性能最好, 其次是SVM、MLP和LR; ③水系和公路两类环境因子的重要性较高且使用改进因子机器学习后这两类环境因子重要性排名依然非常靠前。Abstract: For linear environmental factors such as river, road and geological fault networks, buffer analysis in GIS is commonly used to extract the buffer distances to the river and/or road networks. However, the line distances are discrete variables with random fluctuations of different grid sizes and are more sensitive to the errors of point and/or line elements, leading to a reduction in the accuracy of landslide susceptibility prediction (LSP). This study aims to use continuous environmental factors, such as the spatial density of river and road networks, to improve the suitability of linear environmental factors. Taking An'yuan County of Jiangxi Province as an example, 14 environmental factors, such as elevation, topographic relief, distances to river and road networks (original factors), are selected. Then, the two linear environmental factors of distances to river and road networks are improved to river and road density (improved factors). Based on machine learning models such as logistic regression (LR), multilayer perceptron (MLP), support vector machine (SVM) and C5.0 decision tree, original factor-based and improved factor-based machine learning models are built to carry out the LSP. The receiver operating characteristic (ROC) curves and the distribution characteristics of landslide susceptibility indexes are used to evaluate the LSP modelling rules. The results show that ① the LSP accuracy of the improved factor-based models are higher than those of the original factor-based models, indicating that the spatial density is more suitable for LSP; ② the C5.0 model has the best LSP performance among the four machine learning models, followed by the SVM, MLP and LR models; and ③ river and road factors are of great significance for landslide evolution, and their importance does not decrease underimproved factor-based machine learning models.
-
图 1 研究区地理位置及概况
Figure 1. Geographical location and general situation of the study area
图 2 滑坡易发性评价MLP结构图
Figure 2. Multilayer perceptron structure diagram of landslide susceptibility evaluation
图 4 各模型不同线密度参数改进因子组合及原始因子组合的ROC曲线
Figure 4. ROC curves of improved factors under different linear density parameters and original factors on each model
图 6 各模型下改进因子组合与原始因子组合机器学习模型的ROC曲线
Figure 6. ROC of improved factors and original factors of machine learning of each model
图 7 原始因子组合(a)与改进因子组合(b)在不同模型下的ROC曲线
Figure 7. ROC curve original of factors (a) and improved factors (b) under different models
图 8 改进因子和原始因子模型的滑坡易发性指数分布规律
Figure 8. Distribution rule of landslide susceptibility indexes of improved factor-based and original factor-based models
图 9 改进因子组合(a)及原始因子组合(b)环境因子重要性图
Figure 9. Importance diagram of environmental factors of improved factors (a) and original factors (b)
表 1 各基础环境因子FR值
Table 1. Frequency ratio of every basic environmental factor
基础环境因子 变量值 全区栅格/个 栅格比例/% 滑坡栅格/个 滑坡栅格比例/% FR值 高程/m
(连续型)[181.9, 288.3) 496 187 18.68 3 200 38.57 2.064 [288.3, 368.1) 685 240 25.80 2 366 28.52 1.105 [368.1, 451.7) 462 583 17.42 1 058 12.75 0.732 [451.7, 539.1) 382 432 14.40 667 8.04 0.558 [539.1, 630.3) 279 769 10.53 801 9.65 0.917 [630.3, 736.7) 190 734 7.18 149 1.80 0.250 [736.7, 873.5) 113 403 4.27 56 0.67 0.158 [873.5, 1 194.4] 45 624 1.72 0 0.00 0.000 坡向/(°)
(连续型)[-1, 44.3) 280 739 10.57 662 7.98 0.755 [44.3, 89.6) 304 482 11.46 1 048 12.63 1.102 [89.6, 133.5) 353 235 13.30 1 297 15.63 1.175 [133.5, 178.8) 353 072 13.29 1 269 15.29 1.151 [178.8, 225.5) 325 285 12.25 967 11.65 0.952 [225.5, 270.8) 324 212 12.21 1 145 13.80 1.131 [270.8, 314.7) 358 227 13.49 1 178 14.20 1.053 [314.7, 360.0] 356 720 13.43 731 8.81 0.656 坡度/(°)
(连续型)[0, 4.3) 415 986 15.66 337 4.06 0.259 [4.3, 8.2) 574 723 21.64 1 729 20.84 0.963 [8.2, 11.9) 533 748 20.10 2 352 28.35 1.411 [11.9, 15.5) 437 498 16.47 1 868 22.51 1.367 [15.5, 19.6) 338 458 12.74 1 167 14.07 1.104 [19.6, 24.2) 208 331 7.84 626 7.54 0.962 [24.2, 30.1) 110 358 4.16 197 2.37 0.571 [30.1, 58.2] 36 870 1.39 21 0.25 0.182 平面曲率
(连续型)[0, 10.2) 452 767 17.05 2 161 26.05 1.528 [10.2, 18.9) 525 698 19.79 2 208 26.61 1.345 [18.9, 27.8) 420 490 15.83 1 566 18.87 1.192 [27.8, 37.7) 332 054 12.50 958 11.55 0.924 [37.7, 48.3) 253 510 9.54 480 5.79 0.606 [48.3, 59.1) 205 977 7.76 254 3.06 0.395 [59.1, 70.6) 185 689 6.99 234 2.82 0.403 [70.6, 81.5] 279 787 10.53 436 5.25 0.499 剖面曲率
(连续型)[0, 1.8) 543 228 20.45 1 655 19.95 0.975 [1.8, 3.6) 715 475 26.94 2 495 30.07 1.116 [3.6, 5.4) 565 403 21.29 1 840 22.18 1.042 [5.4, 7.3) 392 134 14.76 1 112 13.40 0.908 [7.3, 9.5) 242 871 9.14 690 8.32 0.909 [9.5, 12.2) 127 410 4.80 380 4.58 0.955 [12.2, 15.9) 54 936 2.07 102 1.23 0.594 [15.9, 38.0] 14 515 0.55 23 0.28 0.507 地形起伏度/m
(连续型)[0, 29.1) 402 714 15.16 1 036 12.49 0.824 [29.1, 49.3) 669 778 25.22 3 105 37.42 1.484 [49.3, 68.4) 572 886 21.57 2 020 24.35 1.129 [68.4, 87.4) 424 833 16.00 1 026 12.37 0.773 [87.4, 108.7) 294 698 11.10 723 8.71 0.785 [108.7, 134.5) 177 688 6.69 354 4.27 0.638 [134.5, 169.2) 85 789 3.23 33 0.40 0.123 [169.2, 285.8] 27 586 1.04 0 0.00 0.000 地形湿度
(连续型)[4.0, 6.3) 739 489 27.84 2 430 29.29 1.052 [6.3, 7.8) 949 906 35.76 3 395 40.92 1.144 [7.8, 9.4) 554 597 20.88 1 759 21.20 1.015 [9.4, 11.2) 226 192 8.52 424 5.11 0.600 [11.3, 13.6) 113 458 4.27 127 1.53 0.358 [13.6, 16.4) 53 428 2.01 131 1.58 0.785 [16.4, 24.0] 18 829 0.71 31 0.37 0.527 [24.0, 40.845] 73 0.00 0 0.00 0.000 NDVI
(连续型)[-0.132, 0.085 3) 11 703 0.44 18 0.22 0.492 [0.085, 0.169) 59 295 2.23 143 1.72 0.772 [0.169, 0.229) 132 723 5.00 580 6.99 1.399 [0.229, 0.274) 312 822 11.78 1 265 15.25 1.294 [0.274, 0.311) 527 207 19.85 1 986 23.94 1.206 [0.311, 0.347) 659 184 24.82 2 135 25.73 1.037 [0.347, 0.389) 621 304 23.39 1 768 21.31 0.911 [0.389, 0.536] 331 734 12.49 402 4.85 0.388 NDBI
(连续型)[-0.530, -0.408) 414 877 15.62 377 4.54 0.291 [-0.408, -0.367) 733 961 27.63 1 541 18.57 0.672 [-0.367, -0.326) 672 071 25.30 2 335 28.14 1.112 [-0.326, -0.282) 383 889 14.45 1 851 22.31 1.543 [-0.282, -0.232) 211 736 7.97 1 120 13.50 1.693 [-0.232, -0.175) 130 136 4.90 661 7.97 1.626 [-0.175, -0.103) 77 499 2.92 320 3.86 1.322 [-0.103, 0.272] 31 803 1.20 92 1.11 0.926 MNDWI
(连续型)[0, 35) 156 772 5.90 554 6.68 1.131 [35, 71) 279 294 10.52 1 294 15.60 1.483 [71, 100) 393 799 14.83 1 663 20.04 1.352 [100, 127) 462 074 17.40 1 604 19.33 1.111 [127, 154) 456 527 17.19 1 336 16.10 0.937 [154, 183) 417 431 15.72 1 015 12.23 0.778 [183, 216) 315 966 11.90 592 7.14 0.600 [216, 255] 174 109 6.56 239 2.88 0.439 岩性
(离散型)变质岩 854 749 32.18 2 690 32.36 1.005 岩浆岩 1 110 912 41.83 2 235 26.89 0.643 碎屑岩 687 217 25.87 3 388 40.76 1.575 碳酸岩 3 094 0.12 0 0.00 0.000 断层密度/(km·km-2)
(连续型)[0, 0.96) 2 246 640 84.59 6 168 74.20 0.877 [0.96, 1.93) 181 229 6.82 811 9.76 1.430 [1.93, 2.89) 104 227 3.92 410 4.93 1.257 [2.89, 3.86) 67 033 2.52 501 6.03 2.388 [3.86, 4.82) 23 831 0.90 138 1.66 1.850 [4.82, 5.79) 21 856 0.82 222 2.67 3.245 [5.79, 6.75) 10 746 0.40 61 0.73 1.814 [6.75, 7.71] 410 0.02 2 0.02 1.559 距断层距离/m
(离散型)< 150 355 996 13.40 1 856 22.33 1.667 [150, 300) 332 812 12.53 1 336 16.07 1.283 [300, 450) 296 570 11.17 854 10.27 0.920 [450, 600) 258 642 9.74 793 9.54 0.980 [600, 800] 289 242 10.89 730 8.78 0.806 > 800 1 122 710 42.27 2 744 33.01 0.781 距公路距离/m
(离散型)< 150 743 873 28.01 4 740 57.02 2.036 [150, 300) 433 925 16.34 2 220 26.71 1.635 [300, 450) 305 927 11.52 802 9.65 0.838 [450, 600) 234 977 8.85 177 2.13 0.241 [600, 800] 240 829 9.07 40 0.48 0.053 > 800 696 441 26.22 334 4.02 0.153 距水系距离/m
(离散型)< 150 934 647 35.19 4 981 59.92 1.703 [150, 300) 763 304 28.74 2 431 29.24 1.018 [300, 450) 536 148 20.19 481 5.79 0.287 [450, 600) 258 803 9.74 171 2.06 0.211 [600, 800] 85 055 3.20 41 0.49 0.154 > 800 78 015 2.94 208 2.50 0.852 公路密度/(km·km-2)
(连续型)[0, 1.78) 1 658 283 62.44 3 215 38.67 0.619 [1.78, 3.56) 596 349 22.45 3 480 41.86 1.864 [3.56, 5.34) 218 100 8.21 765 9.20 1.121 [5.34, 7.12) 88 913 3.35 309 3.72 1.110 [7.12, 8.90) 50 822 1.91 251 3.02 1.578 [8.90, 10.68) 24 483 0.92 146 1.76 1.905 [10.68, 12.46) 16 911 0.64 147 1.77 2.777 [12.46, 14.23] 2 112 0.08 0 0.00 0.000 水系密度/(km·km-2)
(连续型)[0, 0.28) 907 613 34.17 818 9.84 0.288 [0.28, 0.81) 332 419 12.52 364 4.38 0.350 [0.81, 1.32) 330 831 12.46 701 8.43 0.677 [1.32, 1.82) 431 403 16.24 1 960 23.58 1.452 [1.82, 2.31) 254 193 9.57 1 655 19.91 2.080 [2.31, 2.82) 219 790 8.28 1 371 16.49 1.993 [2.82, 3.42) 125 981 4.74 922 11.09 2.338 [3.42, 5.45] 53 745 2.02 522 6.28 3.103 
下载: 导出CSV
表 2 原始因子相关性系数矩阵
Table 2. Correlation coefficient matrix diagram of the original factors
地形起伏度 剖面曲率 坡度 NDVI NDBI 平面曲率 高程 坡向 MNDWI 地形湿度 岩性 断层密度 距水系距离 距公路距离 地形起伏度 1 剖面曲率 0.123 1 坡度 0.172 -0.074 1 NDV 0.109 0.037 -0.100 1 NDBI 0.124 0.042 -0.206 0.313 1 平面曲率 -0.222 0.07 0.208 -0.091 -0.102 1 高程 0.265 0.093 -0.237 0.208 0.317 -0.107 1 坡向 0.042 0.013 -0.003 -0.211 -0.142 0.136 0.035 1 MNDWI 0.086 0.041 -0.131 -0.286 0.094 -0.014 0.181 0.221 1 地形湿度 -0.056 -0.027 0.280 -0.02 -0.122 0.270 -0.202 0.069 -0.169 1 岩性 0.098 0.028 0.004 -0.009 0.026 -0.038 0.013 0.026 0.052 -0.011 1 断层密度 0.037 0.004 -0.021 0.032 0.062 -0.004 0.139 0.012 0.046 -0.025 0.002 1 距水系距离 0.152 -0.001 -0.097 0.064 0.164 -0.054 0.274 0.011 0.148 -0.215 0.012 0.085 1 距公路距离 0.193 0.074 -0.195 0.270 0.400 -0.075 0.474 -0.021 0.183 -0.143 0.075 0.094 0.258 1
下载: 导出CSV
表 3 改进因子相关性系数矩阵
Table 3. Correlation coefficient matrix diagram of the improved factors
地形起伏度 剖面曲率 坡度 NDVI NDBI 平面曲率 高程 坡向 MNDWI 地形湿度 岩性 断层密度 水系密度 公路密度 地形起伏度 1 剖面曲率 0.123 1 坡度 0.172 -0.074 1 NDVI 0.109 0.037 -0.100 1 NDBI 0.124 0.042 -0.206 0.313 1 平面曲率 -0.222 0.070 0.208 -0.091 -0.102 1 高程 0.265 0.093 -0.237 0.208 0.317 -0.107 1 坡向 0.042 0.013 -0.003 -0.211 -0.142 0.136 0.035 1 MNDWI 0.086 0.041 -0.131 -0.286 0.094 -0.014 0.181 0.221 1 地形湿度 -0.056 -0.027 0.280 -0.020 -0.122 0.270 -0.202 0.069 -0.169 1 岩性 0.098 0.028 0.004 -0.009 0.026 -0.038 0.013 0.026 0.052 -0.011 1 断层密度 0.037 0.004 -0.021 0.032 0.062 -0.004 0.139 0.012 0.046 -0.025 0.002 1 水系密度 0.130 0.052 -0.148 0.183 0.283 -0.047 0.341 -0.000 0.145 -0.106 -0.053 0.076 1 公路密度 0.128 -0.002 -0.139 0.115 0.211 -0.054 0.322 -0.011 0.123 -0.159 0.019 0.110 0.262 1
下载: 导出CSV
表 4 原始因子与改进因子回归系数
Table 4. Regression coefficient table of the original factor and improved factor
环境因子 原始因子 改进因子 环境因子 原始因子 改进因子 高程 -0.002 0.143 地形湿度 1.105 0.984 坡度 1.474 1.509 地形起伏度 -0.167 -0.066 坡向 0.756 0.655 岩性 4.034 4.253 平面曲率 1.153 1.233 断层密度 0.126 0.080 剖面曲率 0.311 0.370 距水系距离 0.849 - NDVI 0.148 0.359 距公路距离 0.967 - NDBI 0.515 0.613 水系密度 - 0.715 MNDWI 0.346 0.427 公路密度 - 0.790
下载: 导出CSV
表 5 各模型不同线密度参数下改进因子组合及原始因子组合的AUC值
Table 5. AUC values of the improved factors under different linear density parameters and original factors on each model
工况 水系密度 公路密度 MLP LR SVM C5.0 汇流累积阈值 搜索半径/m 搜索半径/m AUC值 1 300 300 200 0.913 0.881 0.942 0.965 2 300 300 250 0.912 0.884 0.943 0.962 3 300 300 300 0.922 0.888 0.949 0.975 4 300 300 400 0.920 0.886 0.946 0.975 5 400 300 400 0.924 0.892 0.951 0.976 6 400 400 400 0.922 0.898 0.951 0.983 7 原始因子组合 0.896 0.888 0.919 0.939
下载: 导出CSV
-
[1] 周超, 殷坤龙, 曹颖, 等. 基于诱发因素响应与支持向量机的阶跃式滑坡位移预测[J]. 岩石力学与工程学报, 2015, 34(增刊2): 4132-4139. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX2015S2061.htmZhou C, Yin K L, Cao Y, et al. Displacement prediction of step-like landslide based on the response of inducing factors and Support Vector Machine[J]. Chinese Journal of Rock Mechanics and Engineering, 2015, 34(S2): 4132-4139 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX2015S2061.htm [2] Chang Z L, Gao H X, Huang F M, et al. Study on the creep behaviours and the improved Burgers model of a loess landslide considering matric suction[J]. Natural Hazards, 2020, 103: 1479-1497. doi: 10.1007/s11069-020-04046-0 [3] 吴常润, 赵冬梅, 刘澄静, 等. 基于GIS的华宁县滑坡灾害影响因子分析及易发性评价[J]. 水土保持研究, 2019, 26(6): 212-218, 225. https://www.cnki.com.cn/Article/CJFDTOTAL-STBY201906034.htmWu C R, Zhao D M, Liu C J, et al. GIS-based analysis on impact factors and susceptibility evaluation of landslide hazard in Huaning County[J]. Research of Soil and Water Conservation, 2019, 26(6): 212-218, 225 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-STBY201906034.htm [4] Li S, Bai Y, Long Y, et al. GIS-supported evaluation of landslide susceptibility in the karst mountainous area: A case study in Wudang, Guiyang[J]. E3S Web of Conferences, 2020, 143(3/4): 02032. [5] Huang F M, Chen J W, Yao C, et al. SUSLE: A slope and seasonal rainfall-based RUSLE model for regional quantitative prediction of soil erosion[J]. Bulletin of Engineering Geology and the Environment, 2020, 79(10), 5213-5228. doi: 10.1007/s10064-020-01886-9 [6] 郭子正, 殷坤龙, 唐扬, 等. 库水位下降及降雨作用下麻柳林滑坡稳定性评价与预测[J]. 地质科技情报, 2017, 36(4): 260-265, 270. https://www.cnki.com.cn/Article/CJFDTOTAL-DZKQ201704035.htmGou Z Z, Yin K L, Tang Y, et al. Stability evaluation and prediction of Maliulin landslide under reservoir water level decline and rainfall[J]. Geological Science and Technology Information, 2017, 36(4): 260-265, 270 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-DZKQ201704035.htm [7] 熊十力. 基于岩性因子的滑坡易发性评价研究[D]. 武汉: 湖北工业大学, 2020.Xiong S L. Study on landslides susceptibility mapping based on lithology factors[D]. Wuhan: Hubei University of Technology, 2020 (in Chinese with English abstract). [8] 李文彬, 范宣梅, 黄发明, 等. 不同环境因子联接方法和数据驱动模型对滑坡易发性预测建模的影响规律[J/OL]. 地球科学: 1-20[2021-06-25]. http://kns.cnki.net/kcms/detail/42.1874.P.20210506.1457.004.html.Li W B, Fan X M, Huang F M, et al. Influence law of different environmental factor connection methods and data-based models on landslide susceptibility prediction modeling[J/OL]. Earth Science: 1-20[2021-06-25]. http://kns.cnki.net/kcms/detail/42.1874.P.20210506.1457.004.html (in Chinese with English abstract). [9] 田乃满, 兰恒星, 伍宇明, 等. 人工神经网络和决策树模型在滑坡易发性分析中的性能对比[J]. 地球信息科学学报, 2020, 22(12): 2304-2316. doi: 10.12082/dqxxkx.2020.190766Tian N M, Lan H X, Wu Y M, et al. Performance comparison of BP Artificial Neural Network and CART Decision Tree Model in landslide susceptibility prediction[J]. Journal of Geo-information Science, 2020, 22(12): 2304-2316 (in Chinese with English abstract). doi: 10.12082/dqxxkx.2020.190766 [10] Dou J, Yunus A P, Bui D T, et al. Assessment of advanced random forest and decision tree algorithms for modeling rainfall-induced landslide susceptibility in the Izu-Oshima Volcanic Island, Japan[J]. Science of the Total Environment, 2019, 662: 332-346. doi: 10.1016/j.scitotenv.2019.01.221 [11] Tsangaratos P, Ilia I, Hong H Y, et al. Applying information theory and GIS-based quantitative methods to produce landslide susceptibility maps in Nancheng County, China[J]. Landslides, 2017, 14(3): 1091-1111. doi: 10.1007/s10346-016-0769-4 [12] 黄发明, 曹中山, 姚池, 等. 基于决策树和有效降雨强度的滑坡危险性预警[J]. 浙江大学学报: 工学版, 2021, 55(3): 472-482. https://www.cnki.com.cn/Article/CJFDTOTAL-ZDZC202103007.htmHuang F M, Cao Z S, Yao C, et al. Landslideshazard warning based on decision tree and effective rainfall intensity[J]. Journal of Zhejiang University: Engineering Science, 2021, 55(3): 472-482 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-ZDZC202103007.htm [13] Reichenbach P, Rossi M, Malamud B D, et al. A review of statistically-based landslide susceptibility models[J]. Earth-Science Reviews, 2018, 180: 60-91. doi: 10.1016/j.earscirev.2018.03.001 [14] 黄发明, 殷坤龙, 张桂荣, 等. 基于相空间重构和小波分析-粒子群向量机的滑坡地下水位预测[J]. 地球科学: 中国地质大学学报, 2015, 40(7): 1254-1265. https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX201507013.htmHuang F M, Yin K L, Zhang G R, et al. Landslide groundwater level time series prediction based on phase space reconstruction and wavelet analysis-support vector machine optimized by PSO algorithm[J]. Earth Science: Journal of China University of Geosciences, 2015, 40(7): 1254-1265 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX201507013.htm [15] 黄发明, 殷坤龙, 张桂荣, 等. 多变量PSO-SVM模型预测滑坡地下水位[J]. 浙江大学学报: 工学版, 2015, 49(6): 1193-1200. https://www.cnki.com.cn/Article/CJFDTOTAL-ZDZC201506031.htmHuang F M, Yin K L, Zhang G R, et al. Prediction of groundwater level in landslide using multivariable PSO-SVM model[J]. Journal of Zhejiang University: Engineering Science, 2015, 49(6): 1193-1200 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-ZDZC201506031.htm [16] 马彦彬, 李红蕊, 王林, 等. 机器学习方法在滑坡易发性评价中的应用(英文)[J/OL]. 土木与环境工程学报(中英文): 1-14[2021-06-25]. http://kns.cnki.net/kcms/detail/50.1218.TU.20210608.1649.008.html.Ma Y B, Li H X, Wang L, et al. Machine learning algorithms and techniques for landslide susceptibility investigation literature review[J/OL]. Journal of Civil and Environmental Engineering: 1-14[2021-06-25]. http://kns.cnki.net/kcms/detail/50.1218.TU.20210608.1649.008.html (in Chinese with English abstract). [17] 曾帅. 基于多模型对比的盐边县滑坡易发性评价[D]. 成都: 成都理工大学, 2020.Zeng S. Evaluation of landslide susceptibility in Yanbian County based on multi-model comparison[D]. Chengdu: Chengdu University of Technology, 2020 (in Chinese with English abstract). [18] 韩继冲, 张朝, 曹娟. 基于逻辑回归的地震滑坡易发性评价: 以汶川地震、鲁甸地震为例[J]. 灾害学, 2021, 36(2): 193-199. doi: 10.3969/j.issn.1000-811X.2021.02.034Han J C, Zhang C, Cao J. Assessing earthquake-induced landslide susceptibility based on logistic regression in 2008 Wenchuan Earthquake and 2014 Ludian Earthquake[J]. Journal of Catastrophology, 2021, 36(2): 193-199 (in Chinese with English abstract). doi: 10.3969/j.issn.1000-811X.2021.02.034 [19] Huang F M, Cao Z S, Jiang S H, et al. Landslide susceptibility prediction based on a semi-supervised multiple-layer perceptron model[J]. Landslides, 2020, 17(12): 2919-2930. doi: 10.1007/s10346-020-01473-9 [20] 陈飞, 蔡超, 李小双, 等. 基于LR-ANN-SVM的滑坡易发性评价[J]. 有色金属科学与工程, 2020, 11(4): 82-90. https://www.cnki.com.cn/Article/CJFDTOTAL-JXYS202004013.htmChen F, Cai C, Li X S, et al. Evaluation of landslide susceptibility based on LR-ANN-SVM[J]. Nonferrous Metals Science and Engineering, 2020, 11(4): 82-90 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-JXYS202004013.htm [21] Saito H, Nakayama D, Matsuyama H. Comparison of landslide susceptibility based on a decision-tree model and actual landslide occurrence: The Akaishi Mountains, Japan[J]. Geomorphology, 2009, 109(3): 108-121. [22] Tanyu B F, Aiyoub A, Yashar A, et al. Landslide susceptibility analyses using Random Forest, C4.5, and C5.0 with balanced and unbalanced datasets[J]. Catena, 2021, 203: 105355. doi: 10.1016/j.catena.2021.105355 [23] 邓念东, 崔阳阳, 郭有金. 基于频率比-随机森林模型的滑坡易发性评价[J]. 科学技术与工程, 2020, 20(34): 13990-13996. doi: 10.3969/j.issn.1671-1815.2020.34.006Deng N D, Cui Y Y, Guo Y J. Frequency ratio-random forest-model-based landslide susceptibility assessment[J]. Science Technology and Engineering, 2020, 20(34): 13990-13996 (in Chinese with English abstract). doi: 10.3969/j.issn.1671-1815.2020.34.006 [24] Youssef A M, Pourghasemi H R. Landslide susceptibility mapping using machine learning algorithms and comparison of their performance at Abha Basin, Asir Region, Saudi Arabia[J]. Geoscience Frontiers, 2021, 12(2): 639-655. doi: 10.1016/j.gsf.2020.05.010 [25] 冯杭建. 浙西淳安降雨型滑坡发育规律及危险性评价研究[D]. 武汉: 中国地质大学(武汉), 2016.Feng H J. Rainfall-triggered landslide development regularity analysis and hazard assessment in Chun'an, West Zhejiang[D]. Wuhan: China University of Geoscience(Wuhan), 2016 (in Chinese with English abstract). [26] 王倩, 薛云, 张维, 等. 基于支持向量机的滑坡易发性评价[J]. 湖南城市学院学报: 自然科学版, 2021, 30(1): 22-28. doi: 10.3969/j.issn.1672-7304.2021.01.0005Wang Q, Xue Y, Zhang W, et al. Landslide susceptibility mapping based on support vector machine models[J]. Journal of Hunan City University: Natural Science, 2021, 30(1): 22-28 (in Chinese with English abstract). doi: 10.3969/j.issn.1672-7304.2021.01.0005 [27] 胡涛, 樊鑫, 王硕, 等. 基于逻辑回归模型和3S技术的思南县滑坡易发性评价[J]. 地质科技通报, 2020, 39(2): 113-121. doi: 10.19509/j.cnki.dzkq.2020.0212Hu T, Fan J, Wang S, et al. Landslide susceptibility evaluation of Sinan County using logistics regression model and 3S technology[J]. Bulletin of Geological Science and Technology, 2020, 39(2): 113-121 (in Chinese with English abstract). doi: 10.19509/j.cnki.dzkq.2020.0212 [28] 杨永刚, 殷坤龙, 赵海燕, 等. 基于C5.0决策树-快速聚类模型的万州区库岸段乡镇滑坡易发性区划[J]. 地质科技情报, 2019, 38(6): 189-197. https://www.cnki.com.cn/Article/CJFDTOTAL-DZKQ201906023.htm [29] 孙丽, 徐卫东. 江西省安远县滑坡灾害发育特征与形成条件[J]. 山西建筑, 2010, 36(5): 95-96. doi: 10.3969/j.issn.1009-6825.2010.05.059Sun L, Xü W D. The development characteristics and formation conditions of landslide hazard in Anyuan County in Jiangxi Province[J]. Shanxi Architecture, 2010, 36(5): 95-96 (in Chinese with English abstract). doi: 10.3969/j.issn.1009-6825.2010.05.059 [30] 王文坡, 韩爱果, 任光明, 等. 四川省普格县滑坡孕灾环境因子敏感性分析[J]. 长江科学院院报, 2018, 35(9): 63-67, 97. https://www.cnki.com.cn/Article/CJFDTOTAL-CJKB201809015.htmWang W P, Han A G, Ren G M, et al. Sensitivity analysis of hazard-brewing environmental factors of landslides in Puge County of Sichuan Province[J]. Journal of Yangtze River Scientific Research Institute, 2018, 35(9): 63-67, 97 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-CJKB201809015.htm [31] 黄发明, 叶舟, 姚池, 等. 滑坡易发性预测不确定性: 环境因子不同属性区间划分和不同数据驱动模型的影响[J]. 地球科学, 2020, 45(12): 4535-4549. https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX202012017.htmHuang F M, Ye Z, Yao C, et al. Uncertainties of landslide susceptibility prediction: Different attribute interval divisions of environmental factors and different data-based models[J]. Earth Science, 2020, 45(12): 4535-4549 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX202012017.htm [32] Merghadi A, Yunus A P, Dou J, et al. Machine learning methods for landslide susceptibility studies: A comparative overview of algorithm performance[J]. Earth-Science Reviews, 2020, 207: 103225. [33] Hong H Y, Liu J Z, Zhu A X. Landslide susceptibility evaluating using artificial intelligence method in the Youfang district (China)[J]. Environmental Earth Sciences, 2019, 78(15): 488. doi: 10.1007/s12665-019-8415-9 [34] Li D Y, Huang F M, Yan L G, et al. Landslide susceptibility prediction using particle-swarm-optimized multilayer perceptron: Comparisons with multilayer-perceptron-only, BP neural network, and information value models[J]. Applied Sciences, 2019, 9(18): 3664. doi: 10.3390/app9183664 [35] 黄发明, 殷坤龙, 蒋水华, 等. 基于聚类分析和支持向量机的滑坡易发性评价[J]. 岩石力学与工程学报, 2018, 37(1): 156-167. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX201801016.htmHuang F M, Yin K L, Jiang S H, et al. Landslide susceptibility assessment based on clustering analysis and support vector machine[J]. Chinese Journal of Rock Mechanics and Engineering, 2018, 37(1): 156-167 (in Chinese with English abstract). https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX201801016.htm [36] Guo Z Z, Shi Y, Huang F M, et al. Landslide susceptibility zonation method based on C5.0 decision tree and K-means cluster algorithms to improve the efficiency of risk management[J]. Geoscience Frontiers, 2021, 12(6): 101249. https://www.sciencedirect.com/science/article/pii/S1674987121001134 [37] Pham B T, Bui D T, Prakash I, et al. Hybrid integration of multilayer perceptron neural networks and machine learning ensembles for landslide susceptibility assessment at Himalayan area (India) using GIS[J]. Catena, 2017, 149: 52-63. https://www.sciencedirect.com/science/article/pii/S034181621630368X [38] Li W B, Fan X M, Huang F M, et al. Uncertainties analysis of collapse susceptibility prediction based on remote sensing and GIS: Influences of different data-based models and connections between collapses and environmental factors[J]. Remote Sensing, 2020, 12(24): 4134. -
下载:
点击查看大图
图(9) / 表(5)
计量
- 文章访问数: 531
- PDF下载量: 51
- 被引次数: 0
