神经网络方法评价泥石流危险度  

Assessment of Debris Flow Hazards Using Neural NetworksLI Yang-yang

作　　者：李杨杨[1] 

机构地区：[1]河海大学水利水电工程学院,南京210098

出　　处：《黑龙江水专学报》2008年第1期87-90,共4页Journal of Heilongjiang Hydraulic Engineering College

摘　　要：近年来,泥石流问题的研究得到了广泛重视,泥石流危险度评价领域也取得了前所未有的发展。总结了泥石流危险度的评价方法,介绍了模型试验、数值模拟方法、灰色关联度法、神经网络方法及GIS和RS方法。计算了神经网络算例,并与已得泥石流危险度评价成果进行对比,验证了神经网络在危险度评价领域的实用性。最后认为以GIS和RS技术为基础,神经网络方法在定量评价的单沟和区域泥石流危险度评价领域具有广泛应用前景。The problem about debris flow was emphased widely. Assessment for dangerous degree of debris flow hazards were studied. In this paper, some of the assessment methods are introduced, such as the numerical simulation, the method of grey relational degree, neural networks, RS and GIS. Using neural networks, an example has been given. The result testifies the rationality and practicability of the method. Based on GIS and RS, in the field of quantitative assessment in the solo channel as well as the regional debris flow hazard with the neural networks are proposed.

关 键 词：泥石流 危险度 数值模拟 灰色关联度 神经网络 

分 类 号：P642.23[天文地球—工程地质学]