基于模糊神经网络的边坡稳定性评价方法
Stability Assessment of Rock Slope Based on Fuzzy Neural Network
李元松 1陈文峰 2李新平 3田昌贵 1夏进 1郭运华3
作者信息
- 1. 武汉工程大学环境与城市建设学院,武汉430073
- 2. 长江勘测规划设计研究有限责任公司,武汉430010
- 3. 武汉理工大学道路桥梁与结构工程湖北省重点实验室,武汉430070
- 折叠
摘要
针对模糊逻辑推理和人工神经网络各自的优势与局限,提出模糊逻辑与神经网络相结合的边坡稳定性评价方法.基于实测数据的回归分析,找出影响边坡稳定性的主要因素,建立模糊神经网络评价模型.通过对输入知识的预处理和输出知识的后处理,将模糊逻辑推理融入神经网络的非线性计算中;将模糊规则隐含分布于神经网络中,实现了模糊映射算法.选定将整个边坡的稳定状态视为同一变形规律的评价模式,利用所建模型对施工期岩质边坡稳定性进行综合评价.结果表明,该方法用于岩质边坡的稳定性评价取得了较好的效果.
Abstract
An assessing approach to slope stability based on fuzzy logical and neural network was presented in view of their advantages and limitations in respective. An assessing model was set up based on measured data regression analysis used to identify the main factors which affect the stability of rock mass slope during construction. The fuzzy inference process is involved in nonlinear computing of neural net by means of pre-processing of input information and past-processing output results. The fuzzy algorithm is contained in neural net structure. The assessing process to slope stability in construction was implemented. The results show that this approach is applicable to assess stability of rock slope.
关键词
模糊推理/神经网络/岩质边坡/稳定性/评价方法Key words
fuzzy logic/neural network/rock slope/stability/assessing approach引用本文复制引用
出版年
2013
NETL
NSTL
维普
万方数据