基于改进BP网络算法的隧洞围岩分类

周翠英; 张亮; 黄显艺

基于改进BP网络算法的隧洞围岩分类

周翠英^{1, 2, 3,},
张亮^{2, 4},
黄显艺^{2, 4}

1.
中山大学应用力学与工程系, 广东广州 510275
2.
中山大学规划设计研究院岩土工程研究所, 广东广州 510275
3.
中山大学地下工程与信息技术研究中心, 广东广州 510275
4.
中山大学地球科学系, 广东广州 510275

详细信息

作者简介:
周翠英(1963一)，女，博士生导师，主要从事岩土工程与环境地质的教学与研究工作.E-mail：eeszcy@ZSU.edu.cn

中图分类号: P642
计量
- 文章访问数: 3267
- HTML全文浏览量: 113
- PDF下载量: 3
- 被引次数: 0
出版历程
- 收稿日期: 2004-12-07
- 刊出日期: 2005-07-25

Classification of Rocks Surrounding Tunnel Based on Improved BP Network Algorithm

ZHOU Cui-ying^{1, 2, 3
,},
ZHANG Liang^{2, 4},
HUANG Xian-yi^{2, 4}

1.
Department o f Applied Mechanics and Engineering, Sun Yat-sen University, Guangzhou 510275, China
2.
Institute o f Geotechnical Engineering, Sun Yat-sen University, Guangzhou 510275, China
3.
Research Centre o f Underground Engineering and In formation Technology, Sun Yat-sen University, Guangzhou 510275, China
4.
Department o f Earth Sciences, Sun Yat-sen University, Guangzhou 510275, China

摘要

摘要: 围岩分类对指导地下工程的设计和施工具有非常重要的意义.引入人工神经网络的方法, 进行隧洞围岩分类, 在传统BP算法的基础上, 通过改进学习算法、优化传递函数和网络结构进行神经网络方法优化.采用附加动量法和学习速率自适应调整的策略改进学习算法, 使得当误差大于上临界值时, 则降低学习率, 当误差小于下临界值时, 则适当提高学习率, 这样可加快网络的训练速度, 确保网络的稳定性; 通过引入调整学习率参数, 使得传递过程更加敏感, 加快了传递函数的收敛速度, 提高了训练函数的计算精度; 通过给定隐含层节点模型的取值范围, 对网络结构进行优化, 提高了泛化精度.将改进的BP网络模型应用于广东省东深供水改造工程的隧洞围岩分类中, 分类结果与根据《水工隧洞设计规范(SL279-2002) 》的分类结果完全一致, 表明该方法具有良好的工程实用性.
- 围岩分类 /
- 人工神经网络 /
- 改进BP网络算法 /
- 工程应用
Abstract: The classification of rocks surrounding a tunnel has an important significance for guiding design and construction in underground engineering. This paper introduces an artificial neural network method into the classification of these rocks. Based on traditional back propagation (BP) arithmetic, an enhanced neural network method is obtained by improving the training algorithm, transfer function and network structure. By combining the additive momentum method with the self-adjusting learning speed method, the algorithm has been improved: when the error is bigger than the upper critical limits the learning speed automatically decreases; when the error is smaller than the lower critical limits the learning speed automatically increases. Thus, the training speed can be fast yet at the same time the stability of the network can be ensured. By introducing the parameter of adjusting learning speed, the transfer process becomes more sensitive and the convergent speed becomes faster, thus, increasing the calculating precision of the training function. By giving a data range for a certain implicit layer joint model, the structure of the network is optimized; correspondingly, the functional precision is improved. The improved BP network model is tested in example classifications of some typical rocks surrounding tunnels in the Dong Shen Water Supply Reconstruction Project. The results fit well with the classification according to the code of hydraulic tunnel design in China, which indicates that this improved method has a high practical application.
- classification of rocks surrounding a tunnel /
- artificial neural network /
- improved back propagation (BP) network algorithm /
- engineering application

HTML全文

图 1 BP网络模型结构

Fig. 1. Structure of BP network model

下载: 全尺寸图片幻灯片

图 2 不同形式Sigmoid型函数比较

Fig. 2. Comparison of different modalities of Sigmoid function

下载: 全尺寸图片幻灯片

图 3 隧洞围岩稳定性分类模型

Fig. 3. Stability classification model of rocks surrounding tunnels

下载: 全尺寸图片幻灯片

图 4 原始模型(a) 与改进模型(b) 误差曲线

Fig. 4. Error curves of the original model (a) and improved model (b)

下载: 全尺寸图片幻灯片

图 5 优化后的BP网络结构

Fig. 5. Structure of BP network after optimized

下载: 全尺寸图片幻灯片

表 1 围岩分类的神经网络学习样本

Table 1. Learning samples of neural network on classification of surrounding rocks

表 2 归一化及变换处理后的学习样本

Table 2. Learning samples after normalizing and transforming

表 3 待分级的走马岗隧洞围岩分类指标实测值

Table 3. Measurement classification indexes of surrounding rocks on Zoumagang tunnel to be classified

表 4 走马岗隧洞围岩分类结果

Table 4. Classification results of Zoumagang rocks surrounding tunnel

参考文献(23)

[1]	Cong, S., 1998. MATLAB toolbox oriented theory and appli- cation of neural network. University of Science andTechnology of China Press, Hefei(in Chinese).
[2]	Feisi Research and Development Center of Science and Technology Product, 2003. Auxiliary analysis and design of neural networks by MATLAB 6.5. Electronic Industry Publishing House, Beijing(in Chinese).
[3]	Feng, X.T., 2000. Introduction to intelligent rock mechanics. Science Press, Beijing(in Chinese).
[4]	Hao, Z., Liu, B., 2003. Back analysis of mechanical parameters of rocks surrounding openings on the basis of calculus of difference and neural network. Rock and Soil Mechanics, 24(Suppl. ): 78 -79(in Chinese with English abstract).
[5]	Hu, Y.B., Huang, X.B., 2002. Effect of the Three Gorges Project storing on stabil ity of T8-T12 fracture segment of hazardous rocks in Lianziya. Earth Sicence Journal of China University o f Geosciences, 27(2): 193 -198(in Chinese with English abstract).
[6]	Huo, R.K., Liu, H.D., 1998. An application of neural net work to surrounding rock stability classification. Journal o f North China Institute o f Water Conservancy and Hydroelectric Power, 19(2): 62 -63(in Chinese with English abstract).
[7]	Lou, S.T., Shi, Y., 2000. System analysis and design based on MATLAB -Neural Networks. Xidian University Press, Xi'an(in Chinese).
[8]	Rao, Y.Z., Hou, Y.B., 2001. Application of artificial neural network method in the wall rock stability classification evaluation. Gold, 22(10): 15 -16(in Chinese with English abstract).
[9]	Standards of Water Conservancy in People's Republic of China, 2003. Specification for design of hydraulic tunnel(SL279-2002). China Water Power Press, Beijing(in Chinese).
[10]	Tan, C.X., Wang, R.J., Sun, Y., 2000. Present-day structural activity of Shenzhen fracture zone and its effect on crustal stability of water diversion tunnel in Shenzhen. Earth Sicence -Journal of China University of Geo-sciences, 25(1): 51 -56(in Chinese with English abstract).
[11]	Wang, X.C., Zhang, H., Liu, Z.H., 2001. Geological hazards about west-line project of water diversion from upper Yangtze River into upper Yellow River. Earth Sicence Journal o f China University o f Geosciences, 26(3): 297 -303(in Chinese with English abstract).
[12]	William, C.C., Margery, E.H., 1997. Guidelines for the selection of network architecture. Artificial Intelligence for Engineering Design Analysis and Manufacturing, 11(5): 395 -408. doi: 10.1017/S0890060400003322
[13]	丛爽, 1998. 面向MATLAB工具箱的神经网络理论与应用. 合肥: 中国科学技术大学出版社.
[14]	飞思科技产品研发中心, 2003. MATLAB 6.5辅助神经网络分析与设计. 北京: 电子工业出版社.
[15]	冯夏庭, 2000. 智能岩石力学导论. 北京: 科学出版社.
[16]	郝哲, 刘斌, 2003. 基于差分法及神经网络的硐室围岩力学参数反分析. 岩土力学, 24 (增刊): 78-79.
[17]	霍润科, 刘汉东, 1998. 神经网络法在地下洞室围岩分类中的应用. 华北水利水电学院学报, 19 (2): 62-63. https://www.cnki.com.cn/Article/CJFDTOTAL-HBSL802.016.htm
[18]	胡亚波, 黄学斌, 2002. 长江三峡工程蓄水对链子崖危岩体T8-T12缝段稳定性影响研究. 地球科学———中国地质大学学报, 27 (2): 193-198. https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX200202017.htm
[19]	楼顺天, 施阳, 2000. 基于MATLAB的系统分析与设计———神经网络. 西安: 西安电子科技大学出版社.
[20]	饶运章, 侯运炳, 2001. 神经网络方法在围岩稳定性分级评价中的应用. 黄金, 22 (10): 15-16. https://www.cnki.com.cn/Article/CJFDTOTAL-HJZZ200110003.htm
[21]	谭成轩, 王瑞江, 孙叶, 2000. 深圳断裂带现今构造活动性及其对深圳市输水隧洞工程地壳稳定性影响. 地球科学———中国地质大学学报, 25 (1): 51-56. https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX200001010.htm
[22]	王学潮, 张辉, 刘振红, 等, 2001. 南水北调西线工程地质灾害初步研究. 地球科学———中国地质大学学报, 26 (3): 297-303. https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX200103012.htm
[23]	中华人民共和国水利行业标准, 2003. 水工隧洞设计规范(SL279-2002). 北京: 中国水利水电出版社.