您目前所在的位置:首页 - 期刊简介 - 详细页面

铁道科学与工程学报

JOURNAL OF RAILWAY SCIENCE AND ENGINEERING

第12卷    第1期    总第63期    2015年2月

[PDF全文下载]    [Flash在线阅读]

    

文章编号:1672-7029(2015)01-0072-07
基于遗传单纯形算法与RBF网络的地应力场反演方法
谢学斌1,罗海霞1,杨承祥2,李欣1

(1. 中南大学 资源与安全工程学院,湖南 长沙410083;
2. 安徽铜冠(庐江)矿业有限公司,安徽 铜陵 231561
)

摘 要: 为提升RBF神经网络性能,采用遗传单纯形算法优化RBF神经网络隐含层节点中心值,利用FLAC3D软件建立区域的数值计算模型来进行正分析计算,以若干测点的正应力值作为训练样本,用优化的RBF网络反演模型区域的岩体力学参数及初始地应力场。依据沙溪铜矿区的地质资料进行算例分析,该方法的反演计算值、实测值及其他RBF网络反演计算值对比表明:所采用的遗传单纯形算法优化RBF网络的反演方法是可行的,改善了计算精度,对类似工程有一定的参考价值。

 

关键字: 初始地应力场;反演;FLAC3D ;遗传算法;单纯形算法;RBF神经网络

Back analysis of geostress field with RBF neural network and genetic-simplex algorithm
XIE Xuebin1,LUO Haixia1,YNAG Chengxiang2,LI Xin1

1. School of Resources and Safety Engineering, Central South University, Changsha 410083, China;
2. Anhui Tongguan (Lujiang) Mining Co., Ltd, Tongling 231561, China

Abstract:To enhance the performance of RBF neural network, genetic- simplex algorithm(GSA) is used to optimize the central value in the hidden layer of RBF neural network. A numerical model of the area is established by FLAC3Dto conduct positive analysis. taken measured normal stresses of some specific points as the training sample, the optimized RBF neural network (GSA-RBF) was utilized to conduct back analyses for rock mechanical parameters and initial stress field. Based on geological data of Shaxi copper mine, the results obtained from back analysis were compared with those by GSA-RBF and the measured ones, and the comparison results show that the back analysis method of GSA-RBF is feasible, and improves the calculation accuracy, which provides certain reference value to the similar engineering projects.

 

Key words: initial stress field; back analysis; FLAC3D; genetic algorithm; simplex algorithm; RBF neural network

ISSN 1672-7029
CN 43-1423/U

主管:中华人民共和国教育部 主办:中南大学 中国铁道学会 承办:中南大学
湘ICP备09001153号 版权所有:《铁道科学与工程学报》编辑部
------------------------------------------------------------------------------------------
地 址:湖南省长沙市韶山南路22号 邮编:410075
电 话:0731-82655133,82656174   传真:0731-82655133   电子邮箱:jrse@mail.csu.edu.cn