论文部分内容阅读
滑坡的形成是众多非线性关系的影响因子相互作用的结果,传统滑坡预测方法需要大量实地勘查数据。利用Mamdani FIS(模糊推理系统)模型对三峡库区巴东-秭归段进行滑坡易发性预测,并对结果进行评价。通过地理信息系统(geographic information system,GIS)、遥感(remote sensing,RS)技术和区域地质背景资料获取地形类、生态环境类和地质背景类共3类7种滑坡影响因子,建立了192条相关的推理规则,在Matlab平台下基于Mamdani FIS模型得到研究区滑坡易发性预测指数,并生成滑坡易发性区划图。预测结果的受试者工作特征曲线下的面积值为82.8%,显示滑坡评估效果良好。结果证明,与其他模型相比,基于空间信息技术的Mamdani FIS模型,利用其非线性分析能力和基于专家意见的推理规则,评估滑坡易发性时不需要先验知识支撑,简化了模型使用时对数据的要求。另外,该模型只需通过专家意见改变推理规则就可以应用于不同的地质地理环境区域,显示其较强的适应性。
The formation of landslide is the result of the interaction of many nonlinear factors. The traditional landslide prediction method needs a lot of field survey data. Using Mamdani FIS (fuzzy inference system) model to predict the landslide susceptibility in Badong-Zigui Section of the Three Gorges Reservoir, and evaluate the results. Seven kinds of landslide influencing factors of terrain, ecological environment and geological background were obtained by geographic information system (GIS), remote sensing (RS) and regional geological background data, and 192 influential factors were established Based on the Mamdani FIS model, the predictor index of landslide susceptibility in the study area is obtained under the Matlab platform and the zoning map of the landslide susceptibility is generated. The area under the working characteristic curve of the predicted results was 82.8%, indicating good landslide assessment. The results show that, compared with other models, the Mamdani FIS model based on spatial information technology, using its non-linear analysis ability and expert-based reasoning rules, does not need prior knowledge to evaluate the landslide susceptibility, and simplifies the use of the model Data requirements. In addition, the model can be applied to different geo-geo-environment regions only by changing the rules of reasoning through expert opinion, which shows its strong adaptability.