State Affine Transfer Learning Method for Hydrodynamic Pressure-Driven Landslide
-
摘要: 三峡库区的动水驱动型滑坡具有阶梯式变形特征,在监测数据不足的情况下,难以准确、合理地完成滑坡分析与预测预报等相关研究.针对监测数据不足的情况,设计了一种状态仿射迁移学习方法(State affine transfer learning method,SATLM),通过学习相似滑坡的知识完成对数据量不足的滑坡状态分析.为验证SATLM对滑坡状态分析的有效性,设计了一种状态相似分析方法,完成对库区多个滑坡的知识学习后实现对另一个数据量不足的滑坡地表位移预测.结果表明,完成状态仿射迁移后,本方法与BPNN和SVM相比,万州塘角1号滑坡地表位移预测的平均绝对误差和均方根误差都实现了较大降低.白家包滑坡、白水河滑坡、八字门滑坡知识的成功迁移,证明了SATLM在相似动水驱动型滑坡的知识迁移上具有较好效果.Abstract: The hydrodynamic pressure-driven landslides in the Three Gorges reservoir area have the characteristics of stepped deformation,and it is difficult to complete the analysis and prediction of landslides accurately and reasonably under the condition of insufficient monitoring data. In view of insufficient monitoring data,a state affine transfer learning method (SATLM) was designed in this paper to analyze the state of landslides with insufficient data by learning similar landslide knowledge. In order to verify the effectiveness of SATLM in landslide state analysis,a state similarity analysis method was designed in this paper. After learning the knowledge of multiple landslides in the reservoir area,another landslide displacement prediction with insufficient data was realized.The results show that compared with BPNN and SVM,the mean absolute error and root mean square error of landslide displacement prediction of Wanzhou Tangjiao No.1 landslide are greatly reduced after state affine migration.The successful knowledge transfer of Baijiabao landslide,Baishuihe landslide,Bazimen landslide proves that the state affine transfer learning method has a good effect on the knowledge transfer of similar hydrodynamic pressure-driven landslides.
-
图 5 白家包滑坡(a)、白水河滑坡(b)、八字门滑坡(c)、万州塘角1号滑坡(d)剖面图(Long et al., 2022)
Fig. 5. The cross-section of Baijiabao landslide (a), Baishuihe landslide (b), Bazimen landslide (c), Wanzhou Tangjiao No.1 landslide (d) (Long et al., 2022)
图 6 白家包滑坡(a)、白水河滑坡(b)、八字门滑坡(c)、万州塘角1号滑坡(d)GPS监测点分布图(Long et al., 2022)
Fig. 6. The locations of GPS monioring sites of (a) Baijiabao landslide, (b) Baishuihe landslide, (c) Bazimen landslide, (d) Wanzhou Tangjiao No.1 landslide (Long et al., 2022)
图 7 白家包滑坡(a)、白水河滑坡(b)、八字门滑坡(c)、万州塘角1号滑坡(d)累积地表位移、库水位、降雨量关系(Li et al., 2021)
Fig. 7. Relationship between the cumulative displacement, reservoir water level and precipitation of Baijiabao landslide (a), Baishuihe landslide (b), Bazimen landslide (c), Wanzhou Tangjiao No.1 landslide (d) (Li et al., 2021)
表 1 滑坡地表位移与影响因子灰色关联度统计
Table 1. Grey correlation degree statistics of landslide displacements and influence factors
监测点
名称Rt‒2 Rt‒1 Rt Kt‒2 Kt‒1 Kt △Wt‒2 △Wt‒1 源域 ZG323 0.74 0.76 0.75 0.74 0.76 0.76 0.84 0.88 ZG324 0.73 0.76 0.74 0.75 0.76 0.75 0.85 0.88 ZG325 0.73 0.75 0.74 0.74 0.76 0.75 0.84 0.88 ZG326 0.73 0.75 0.73 0.74 0.76 0.75 0.84 0.88 ZG93 0.75 0.75 0.76 0.72 0.74 0.75 0.77 0.80 ZG110 0.73 0.72 0.71 0.75 0.71 0.69 0.70 0.67 ZG111 0.67 0.73 0.72 0.72 0.72 0.67 0.65 0.61 目标域 WZ13-09 0.67 0.66 0.73 0.81 0.81 0.81 0.75 0.75 表 2 滑坡状态的分类情况
Table 2. Classification of landslide state
ZG323 ZG324 ZG325 ZG326 突变状态时间段 2009.06
2010.07
2011.06
2012.06
2013.062009.06
2010.07
2011.06
2012.06
2013.062009.06
2010.07
2011.06
2012.06
2013.062009.06
2010.07
2011.06
2012.06
2013.06蠕变状态时间段 2008.03~2009.05
2009.10~2010.06
2010.10~2011.05
2011.10~2012.05
2012.10~2013.05
2013.10~2013.122008.03~2009.05
2009.10~2010.06
2010.10~2011.05
2011.10~2012.05
2012.10~2013.05
2013.10~2013.122008.03~2009.05
2009.10~2010.06
2010.10~2011.05
2011.10~2012.05
2012.10~2013.05
2013.10~2013.122008.03~2009.05
2009.11~2010.06
2010.11~2011.05
2011.10~2012.05
2012.10~2013.05
2013.10~2013.12ZG93 ZG110 ZG111 WZ13-09 突变状态时间段 2010.07
2011.07
2012.06
2013.072009.06
2010.07
2011.06
2012.07
2013.062008.09
2009.06
2010.07
2011.06
2012.07
2013.062012.04
2013.05蠕变状态时间段 2010.03~2010.06
2010.12~2011.06
2011.10~2012.05
2012.11~2013.06
2013.10~2013.122008.03~2009.05
2009.10~2010.06
2010.10~2011.05
2011.10~2012.06
2012.10~2013.05
2013.09~2013.122008.03~2008.08
2008.12~2009.05
2009.10~2010.06
2010.10~2011.05
2011.10~2012.06
2012.10~2013.05
2013.09~2013.122011.03~2012.03
2012.07~2013.04
2013.08~2014.05表 3 3种方法预测结果的RMSE值和MAE值
Table 3. The RMSE and MAE values of three methods
RMSE MAE SSAM 2.5 1.9 BPNN 8.3 5.4 SVM 8.5 5.1 -
Ai, X., Sun, B. T., Chen, X. Z., 2022. Construction of Small Sample Seismic Landslide Susceptibility Evaluation Model Based on Transfer Learning: A Case Study of Jiuzhaigou Earthquake. Bulletin of Engineering Geology and the Environment, 81(3): 81-116. doi: 10.1007/s10064-022-02601-6?utm_source=xmol&utm_content=meta Fu, Z., Long, J., Chen, W., et al., 2021. Reliability of the Prediction Model for Landslide Displacement with Step⁃Like Behavior. Stochastic Environmental Research and Risk Assessment, 35(11): 2335-2353. doi: 10.1007/s00477-021-02029-6 Gao, D. X., Li, K., Cai, Y. C., et al., 2021. Predicting of Landslide Displacement Based on Time Series and Pso⁃Bp Model in Three Georges Reservoir, China. Journal of Earth Science: 1-17. https://doi.org/10.1007/s12583-021-1575-z Guo, C., Xu, Q., Dong, X. J., et al., 2021. Geohazard Recognition and Inventory Mapping Using Airborne LiDAR Data in Complex Mountainous Areas. Journal of Earth Science, 32(5): 1079-1091. doi: 10.1007/s12583-021-1467-2 Guo, J. W., Li, Y. S., Li, Z., et al., 2016. An Automatic Interpretation Model for Mountains Landslide Disaster of High⁃Resolution Remote Sensing Images Based on Transfer Learning. Journal of Geomatics Science and Technology, 33(5): 496-501 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTotal-JFJC201605011.htm Huang, F. M., Chen, J. W., Fan, X. M., et al., 2022. Logistic Regression Fitting of Rainfall⁃Induced Landslide Occurrence Probability and Continuous Landslide Hazard Prediction Modelling. Earth Science, 47(12): 4609-4628 (in Chinese with English abstract). http://www.sciencedirect.com/science/article/pii/S0169555X22001295 Huang, F. M., Yin, K. L., Yang, B. B., et al., 2018. Step⁃Like Displacement Prediction of Landslide Based on Time Series Decomposition and Multivariate Chaotic Model. Earth Science, 43(3): 887-898 (in Chinese with English abstract). Li, L., Wu, Y., Miao, F., et al., 2021. A Hybrid Interval Displacement Forecasting Model for Reservoir Colluvial Landslides with Step⁃Like Deformation Characteristics Considering Dynamic Switching of Deformation States. Stochastic Environmental Research and Risk Assessment, 35: 1089-1112. doi: 10.1007/s00477-020-01914-w Lian, C., Zhu, L., Zeng, Z., et al., 2018. Constructing Prediction Intervals for Landslide Displacement Using Bootstrapping Random Vector Functional Link Networks Selective Ensemble with Neural Networks Switched. Neurocomputing, 291: 1-10. doi: 10.1016/j.neucom.2018.02.046 Lin, Q. G., Liu, Y. Y., Liu, L. Y., et al., 2017. Earthquake⁃Triggered Landslide Susceptibility Assessment Based on Support Vector Machine Combined with Newmark Displacement Model. Journal of Geo⁃Information Science, 19(12): 1623-1633 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTOTAL-DQXX201712011.htm Liu, P. Y., Chang, M., Wu, B. B., et al., 2022. Route Selection of Landslide Prone Area in Wenchuan Section of Chengdu⁃Wenchuan Expressway Based on SBAS⁃ InSAR. Earth Science, 47(6): 2048-2057 (in Chinese with English abstract). Liu, Y., Hu, B. D., Chen, Z., 2019. A Similarity Measurement Method for Multiple Information Data of Landslide. Rock and Soil Mechanics, 40(10): 4001-4010 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTotal-YTLX201910036.htm Liu, Y., Xu, C., Huang, B., et al., 2020. Landslide Displacement Prediction Based on Multi⁃Source Data Fusion and Sensitivity States. Engineering Geology, 271: 105608. doi: 10.1016/j.enggeo.2020.105608 Long, J., Li, C., Liu, Y., et al., 2022. A Multi⁃Feature Fusion Transfer Learning Method for Displacement Prediction of Rainfall Reservoir⁃Induced Landslide with Step⁃Like Deformation Characteristics. Engineering Geology, 297: 106494. doi: 10.1016/j.enggeo.2021.106494 Lu, H., Ma, L., Fu, X., et al., 2020. Landslides Information Extraction Using Object⁃Oriented Image Analysis Paradigm Based on Deep Learning and Transfer Learning. Remote Sensing, 12: 752. doi: 10.3390/rs12050752 Qin, S., Guo, X., Sun, J., et al., 2021. Landslide Detection from Open Satellite Imagery Using Distant Domain Transfer Learning. Remote Sensing, 13: 3383. doi: 10.3390/rs13173383 Rajakumar, R., 2021. One⁃Dimensional Quaternionic Special Affine Fourier Transform. Advances in Applied Clifford Algebras, 31(5): 13. doi: 10.1007/s00006-021-01174-z Shi, K. Y., Zhang, D. X., Han, X. Q., et al., 2022. Digital Twin Model of Photovoltaic Power Generation Prediction Based on LSTM and Transfer Learning. Power System Technology, 46(4): 1363-1371 (in Chinese with English abstract). Tsung, F., Zhang, K., Cheng, L. W., et al., 2018. Statistical Transfer Learning: A Review and Some Extensions to Statistical Process Control. Quality Engineering, 30(1): 115-128. https://doi.org/10.1080/08982112.2017.1373810 Xiao, T., 2020. Landslide Risk Assessment in Wanzhou District and a Key Section, Three Gorges Reservoir (Dissertation). China University of Geosciences, Wuhan, 146 (in Chinese with English abstract). Xu, S. L., 2018. Study on Dynamic Landslide Susceptibility Mapping Based on Multi⁃Source Remote Sensing Imagery (Dissertation). China University of Geosciences, Wuhan (in Chinese with English abstract). Yang, Y., Mei, G., 2021. Deep Transfer Learning Approach for Identifying Slope Surface Cracks. Applied Sciences, 11: 11193. doi: 10.3390/app112311193 Yao, Y., Zhang, Z. X., Ni, X., et al., 2022. Cgnet: Detecting Computer⁃Generated Images Based on Transfer Learning with Attention Module. Signal Processing: Image Communication, 105: 116692. doi: 10.1016/j.image.2022.116692 Yu, M. L., Mei, H. B., Li, J. H., et al., 2016. Landslide Displacement Prediction Based on Varying Coefficient Regression Model in Three Gorges Reservoir Area. Earth Science, 41(9): 1593-1602 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTotal-DQKX201609014.htm Zhou, C., Yin, K., Cao, Y., et al., 2016. Application of Time Series Analysis and PSO⁃SVM Model in Predicting the Bazimen Landslide in the Three Gorges Reservoir, China. Engineering Geology, 204: 108-120. doi: 10.1016/j.enggeo.2016.02.009 Zou, Z., Yang, Y., Fan, Z., et al., 2020. Suitability of Data Preprocessing Methods for Landslide Displacement Forecasting. Stochastic Environmental Research and Risk Assessment, 34(8): 1105-1119. doi: 10.1007/s00477-020-01824-x 郭加伟, 李永树, 李政, 等, 2016. 迁移学习支持下的高分影像山地滑坡灾害解译模型. 测绘科学技术学报, 33(5): 496-501. https://www.cnki.com.cn/Article/CJFDTOTAL-JFJC201605011.htm 黄发明, 陈佳武, 范宣梅, 等, 2022. 降雨型滑坡时间概率的逻辑回归拟合及连续概率滑坡危险性建模. 地球科学, 47(12): 4609-4628. doi: 10.3799/dqkx.2021.164 黄发明, 殷坤龙, 杨背背, 等, 2018. 基于时间序列分解和多变量混沌模型的滑坡阶跃式位移预测. 地球科学, 43(3): 887-898. doi: 10.3799/dqkx.2018.909 林齐根, 刘燕仪, 刘连友, 等, 2017. 支持向量机与Newmark模型结合的地震滑坡易发性评估研究. 地球信息科学学报, 19(12): 1623-1633. https://www.cnki.com.cn/Article/CJFDTOTAL-DQXX201712011.htm 刘沛源, 常鸣, 武彬彬, 等, 2022. 基于SBAS⁃InSAR技术的成汶高速汶川段滑坡易发区选线研究. 地球科学, 47(6): 2048-2057. doi: 10.3799/dqkx.2022.069 刘勇, 胡宝丹, 陈喆, 2019. 滑坡监测点多信息相似性度量方法研究. 岩土力学, 40(10): 4001-4010. https://www.cnki.com.cn/Article/CJFDTOTAL-YTLX201910036.htm 史凯钰, 张东霞, 韩肖清, 等, 2022. 基于LSTM与迁移学习的光伏发电功率预测数字孪生模型. 电网技术, 46(4): 1363-1371. https://www.cnki.com.cn/Article/CJFDTOTAL-DWJS202204014.htm 肖婷, 2020. 三峡库区万州区及重点库岸段滑坡灾害风险评价(博士学位论文). 武汉: 中国地质大学. 许石罗, 2018. 基于多源遥感影像的动态滑坡灾害空间预测模型研究(博士学位论文). 武汉: 中国地质大学. 喻孟良, 梅红波, 李冀骅, 等, 2016. 基于变系数回归模型的三峡库区滑坡位移预测. 地球科学, 41(9): 1593-1602. doi: 10.3799/dqkx.2016.118 -