基于自适应模糊神经网络模型的边坡形变预测应用研究
发布时间:2018-11-11 16:47
【摘要】:由于边坡变形所造成的地质灾害往往会对人们的日常生活及工程建设造成很大影响,因此边坡的变形预测成为近年来变形预测方面的一个重要研究方向,在对边坡变形进行监测的同时,对监测数据进行分析,然后做出及时准确预测,能在很大程度上减少灾害发生时造成的国家经济损失及人们的生命安全损失。本文在研究了已有的一些变形预测模型的基础上,利用模糊集理论,结合BP神经网络模型,建立了自适应模糊神经推理系统模型(Adaptive Neuro-Fuzzy Inference System)也称为基于网络的自适应模糊推理系统(Adaptive Network-based Fuzzy Inference System),简称为ANFIS,该模型同时具有模糊逻辑的易于表达人类知识能力和神经网络的分布式信息存储及学习能力,融合了模糊系统的语言推理能力和神经网络的学习机制能力,表现出很强的容错能力,具有很好地适应性,本文探索性的将该模型应用到边坡工程实例中,并用采集的监测数据,验证该模型在边坡变形预测中的可行性。本文的主要研究内容及成果总结如下: 1、简单介绍了目前常用的一些变形预测模型,重点讲述了BP神经网络模型的基本理论、结构、算法以及流程。 2、详细介绍了模糊系统理论,,以及模糊隶属度、模糊规则的选取,简单的介绍了模糊系统工具箱的各中参数的应用,比较了模糊系统与BP神经网络的的优缺点,将两种理论进行融合,建立了自适应模糊神经推理系统模型(ANFIS),并根据该模型的优点提出了的它在变形预测中的适用性。 3、基于自适应模糊神经网络模型,本文选取了黑龙江鹤岗市华鹤煤化股份有限公司大型煤化工基地边坡监测项目中深层水平位移的实测数据对模型进行验证,并对预测结果和精度进行对比分析,证明了该模型在此边坡变形预测中的可行性。
[Abstract]:Because the geological disasters caused by slope deformation often have a great impact on people's daily life and engineering construction, slope deformation prediction has become an important research direction of deformation prediction in recent years. When monitoring the slope deformation, analyzing the monitoring data and making timely and accurate prediction, the national economic loss caused by disasters and the loss of people's life and safety can be reduced to a great extent. On the basis of studying some existing deformation prediction models, this paper uses fuzzy set theory and BP neural network model. The adaptive fuzzy neural inference system model (Adaptive Neuro-Fuzzy Inference System) is established, which is also called the network based adaptive fuzzy inference system (Adaptive Network-based Fuzzy Inference System),). The model has the ability of fuzzy logic to express human knowledge easily and the distributed information storage and learning ability of neural network, which combines the language reasoning ability of fuzzy system and the learning mechanism of neural network. The model is applied to slope engineering examples and the monitoring data are collected to verify the feasibility of the model in slope deformation prediction. The main contents and achievements of this paper are summarized as follows: 1. Some commonly used deformation prediction models are briefly introduced, with emphasis on the basic theory, structure, algorithm and flow of BP neural network model. 2. The theory of fuzzy system, the degree of fuzzy membership and the selection of fuzzy rules are introduced in detail. The application of the parameters of fuzzy system toolbox is briefly introduced, and the advantages and disadvantages of fuzzy system and BP neural network are compared. The adaptive fuzzy neural inference system model (ANFIS),) is established by combining the two theories and its applicability in deformation prediction is proposed according to the advantages of the model. 3. Based on the adaptive fuzzy neural network model, the model is verified by the measured data of the deep horizontal displacement in the slope monitoring project of the large coal chemical base of Huahe Coal Chemical Company, Hegang City, Heilongjiang Province. The feasibility of the model in the prediction of the slope deformation is proved by comparing the prediction results with the precision.
【学位授予单位】:长安大学
【学位级别】:硕士
【学位授予年份】:2014
【分类号】:TU433;TP18
本文编号:2325499
[Abstract]:Because the geological disasters caused by slope deformation often have a great impact on people's daily life and engineering construction, slope deformation prediction has become an important research direction of deformation prediction in recent years. When monitoring the slope deformation, analyzing the monitoring data and making timely and accurate prediction, the national economic loss caused by disasters and the loss of people's life and safety can be reduced to a great extent. On the basis of studying some existing deformation prediction models, this paper uses fuzzy set theory and BP neural network model. The adaptive fuzzy neural inference system model (Adaptive Neuro-Fuzzy Inference System) is established, which is also called the network based adaptive fuzzy inference system (Adaptive Network-based Fuzzy Inference System),). The model has the ability of fuzzy logic to express human knowledge easily and the distributed information storage and learning ability of neural network, which combines the language reasoning ability of fuzzy system and the learning mechanism of neural network. The model is applied to slope engineering examples and the monitoring data are collected to verify the feasibility of the model in slope deformation prediction. The main contents and achievements of this paper are summarized as follows: 1. Some commonly used deformation prediction models are briefly introduced, with emphasis on the basic theory, structure, algorithm and flow of BP neural network model. 2. The theory of fuzzy system, the degree of fuzzy membership and the selection of fuzzy rules are introduced in detail. The application of the parameters of fuzzy system toolbox is briefly introduced, and the advantages and disadvantages of fuzzy system and BP neural network are compared. The adaptive fuzzy neural inference system model (ANFIS),) is established by combining the two theories and its applicability in deformation prediction is proposed according to the advantages of the model. 3. Based on the adaptive fuzzy neural network model, the model is verified by the measured data of the deep horizontal displacement in the slope monitoring project of the large coal chemical base of Huahe Coal Chemical Company, Hegang City, Heilongjiang Province. The feasibility of the model in the prediction of the slope deformation is proved by comparing the prediction results with the precision.
【学位授予单位】:长安大学
【学位级别】:硕士
【学位授予年份】:2014
【分类号】:TU433;TP18
【参考文献】
相关期刊论文 前10条
1 曹安照,田丽;自适应模糊神经网络在工业用水量预测中的应用[J];安徽工程科技学院学报(自然科学版);2005年03期
2 尹晖,陈永奇,张琰;贫信息条件下的多点变形预测模型及其应用[J];测绘学报;1997年04期
3 张志红;韩直;肖盛燮;常贵智;甘守武;;基于ANFIS交通流实时预测及在MATLAB中的实现[J];重庆交通学院学报;2007年03期
4 侯培国,杨靖,卢志刚,吴士昌;一种结构简单的四层模糊神经网络的设计[J];燕山大学学报;1998年04期
5 达飞鹏,宋文忠;用输入输出模型表示的非线性系统的模糊神经网络滑模自适应控制[J];电子学报;2000年07期
6 曹衍龙,刘保军,杨将新;基于模糊神经网络的公差-成本模型研究[J];工程设计学报;2004年02期
7 王翠红;刘金华;;基于聚类算法的模糊神经网络MCP预测方法研究[J];光盘技术;2009年06期
8 徐洪钟,胡群革,吴中如;自适应模糊神经网络在大坝安全监控中的应用[J];河海大学学报(自然科学版);2001年02期
9 李绍中;;基于自适应模糊神经系统的网络学习评价模型研究[J];计算技术与自动化;2012年02期
10 何习平;华锡生;何秀凤;冯小磊;;边坡变形预测研究现状与发展趋势[J];江西科学;2007年04期
本文编号:2325499
本文链接:https://www.wllwen.com/guanlilunwen/chengjian/2325499.html