基于数据的滑坡预测与模糊控制研究

发布时间：2018-08-20 07:53

【摘要】：近年来,在人类社会现代化与城镇化的进程当中,自然灾害随着生态环境的破坏愈演愈烈。滑坡是一种非常常见的自然灾害,往往直接或者间接危害着人们的生命财产安全,加强对滑坡的研究具有非常重要的经济意义和现实意义。文章首先总结了滑坡预测、滑坡控制及滑坡三维可视化的研究现状;随后以白水河滑坡与凉水井滑坡为研究对象,引入二次指数平滑和BP神经网络两种预测方法,成功实现了比较精确的滑坡位移预测;基于滑坡预测的研究成果,以减缓凉水井滑坡位移变化速率为目的,将模糊控制引入到滑坡控制中并进行了模糊控制系统设计;根据白水河滑坡区域等高线地图及监测数据,利用Creator建立了滑坡地理模型,并基于Vega Prime平台成功展现了滑坡演化的全过程。最后,针对文中三个方面研究的不足之处和后续需要拓宽与深化的研究点进行了总结,就滑坡预测方法、模糊控制精确度与三维可视化仿真的动态实现三个方面给出了自己的建议。本文探讨了当今滑坡研究方面的三个主要问题。利用滑坡预测对前期监测数据进行处理,以滑坡控制手段控制滑坡未来走向,通过三维可视化动态地呈现滑坡演化过程,三者先后承接连成一个完整系统,对于滑坡研究具有一定的参考意义。
[Abstract]:In recent years, in the process of modernization and urbanization of human society, natural disasters become more and more serious with the destruction of ecological environment. Landslide is a very common natural disaster, often directly or indirectly endangering the safety of people's lives and property. It is of great economic and practical significance to strengthen the study of landslide. This paper first summarizes the research status of landslide prediction, landslide control and 3D visualization of landslide, then takes Baishuihe landslide and Liangshui landslide as research objects, and introduces two prediction methods: quadratic exponential smoothing and BP neural network. Based on the research results of landslide prediction, fuzzy control is introduced into landslide control and fuzzy control system is designed to slow down the change rate of landslide displacement in Liangshui well. According to the contour map and monitoring data of Baishuihe landslide area, the landslide geography model is established by using Creator, and the whole process of landslide evolution is successfully demonstrated based on Vega Prime platform. Finally, aiming at the shortcomings of the three aspects of the research and the further research points that need to be broadened and deepened in the paper, the method of landslide prediction is summarized. Three aspects of fuzzy control accuracy and dynamic realization of 3D visual simulation are given. Three main problems in landslide research are discussed in this paper. The early monitoring data are processed by landslide prediction, the future direction of landslide is controlled by landslide control means, and the process of landslide evolution is presented dynamically through 3D visualization. The three aspects are successively connected into a complete system. It has certain reference significance for landslide research.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P642.22;TP273.4

【相似文献】