基于模糊层次分析的CFG桩复合地基辅助神经网络设计  

Design method of CFG piles composite foundation with assistant neural networks based on fuzzy analytical hierarchy process

作　　者：张京穗[1] 彭勇波[2] 刘章军[1] 

机构地区：[1]三峡大学土木水电学院,湖北宜昌443002 [2]同济大学土木工程学院,上海200092

出　　处：《西安建筑科技大学学报（自然科学版）》2006年第3期360-367,共8页Journal of Xi'an University of Architecture & Technology(Natural Science Edition)

摘　　要：建立了面向CFG桩复合地基承载力与沉降量的辅助BP网络设计模型.利用模糊层次分析法对桩周土分层各指标的综合影响进行了评价,不仅逻辑概念明确,而且计算大为简化.通过网络学习与预测分析,说明基于模糊层次分析法的CFG桩复合地基辅助神经网络设计模型预测效果良好、可靠,具有较好的工程实用价值.Fuzzy analytical hierarchy process is introduced into the assistant design model with back propagation neural networks according to the problems about the bearing capacity and settlement value of CFG piles composite foundation, and the effect of soil around piles has been evaluated, with which not only the logical concept of design model is clear but also the workload is greatly reduced. Moreover, the design model is proved to be steady and fine as for as the results are concerned. This has certain value for projects.

关 键 词：BP神经网络 模糊层次分析法 CFG桩复合地基 承载力 沉降量 

分 类 号：TU473.11[建筑科学—结构工程]