滑坡综合集成智能分析方法研究进展

本文综合课题组在滑坡方面的研究,以具体的滑坡工程为实例,详细介绍了滑坡现场监测、基于GIS的滑坡三维可视化、综合集成智能滑坡分析方法,包括滑坡变形预测的进化支持向量机建模、加固方案优化设计的进化神经网络-数值方法、基于动态聚类的进化神经网络危险性分区、滑体参数反演的进化支持向量机-有限元方法、滑动面搜索的免疫连续蚁群-极限平衡分析、滑动面参教反演的粒子群优化算法等。结果表明,所提出的滑坡研究综合集成智能分析方法是适用、科学和合理的。

The research progress of intelligent analysis with comprehensive and integrated methodology towards landslide study

This paper reports the nearly research progress on landslide of our research group. On the basis of a specific landslide, the in situ monitoring, three dimensional visualization and compressive integrated intelligent analysis methods are introduced. The intelligent analysis methods include the following aspects: deformation time series prediction model of support vector machine, analysis and optimum design of landslide, optimum design of reinforcement of landslide using evolutionary neural network-numerical method, hazard zonal approach based on dynamic cluster analysis using evolutionary support vector machine, back analysis of mechanical parameters using evolutionary support vector machine-finite element method, search for slip surface using immunized continuous ant colony algorithm with limiting equilibrium method, parameters back analysis of slip surface using particle swarm optimization algorithm, et al. The results indicated that the suggested methodology is scientific and applicable.