APP下载

基于灰色关联故障树的运营期隧道病害风险研究

2016-03-24张国喜李晓钟

工程管理学报 2016年1期
关键词:故障树灰色关联

张国喜,李晓钟

(兰州交通大学 土木工程学院,甘肃 兰州 730070,E-mail:395374064@qq.com)



基于灰色关联故障树的运营期隧道病害风险研究

张国喜,李晓钟

(兰州交通大学 土木工程学院,甘肃 兰州 730070,E-mail:395374064@qq.com)

摘 要:隧道运营期的管理决定了铁路运行的安全,为保证运营安全,必须对隧道运营期病害的风险进行管理研究。用故障树方法对隧道病害进行分析,找出影响隧道运营安全作为底事件,分析出故障树系统任意一个底事件相关割集的重要度。利用灰色关联的思想得出任意最小割集的关联度,再依据关联度得出需要重视的底事件(既有排水设施年久失修阻塞,材料强度下降,隧道内照明不良),并提出具有针对性的管理改进建议。

关键词:隧道病害;灰色关联;故障树;运营期风险;底事件

随着经济的发展,人们对于出行提出了新的要求。2004年1月国务院颁布的中长期铁路网规划指出,建立省会城市间以及地区间如京津冀地区、长三角地区、环渤海湾地区和珠三角地区的城市快速铁路系统,到2020年全国铁路运营里程达到12万km,实现“四纵四横”铁路总体布局。截止到2013年底,我国运营的隧道达到11074座,受设计、施工、运营环境、运营时间等的影响,隧道病害日益严重[1]。目前隧道建设受气候条件、地质条件和设计要求等影响,建设难度增大。超长、超大跨径和超强水压隧道越来越多,使得后期隧道运营维护难度加大,很多隧道出现了不同程度和各种各样的病害。如水害、冻害、衬砌结构破坏、衬砌材料劣化、附属设施损坏以及洞口仰坡塌方落石等。随着铁路大提速后运营时间的增长,隧道病害的探测和维护时间缩短,加重了隧道病害的发生。

目前已有学者对上述问题进行了相关研究,如曾水长[2]针对具体的病害给出了对应的整治方法。袁超等[3]提出了隧道的劣化机制并根据劣化等级给出了整治方法。唐亮[4]对隧道衬砌结构的风险和可靠性进行了研究,建立了衬砌结构系统建设期风险估计的模型,研究了各风险因素对衬砌结构系统的影响程度的大小。从目前的研究现状可以看出,隧道运营期的风险分析比较缺乏,因此,本文主要对运营期的隧道病害进行风险研究,建立隧道运营期病害风险估计模型,根据风险重要度分析,找出各种因素对隧道运营的影响程度大小,为今后隧道维护提供借鉴。

1 建立隧道运营期故障树

通过大量阅读隧道运营期病害及病害整治方法的文献[5~8],以及对沈阳铁路局管辖的226座隧道探测得到的数据[4](见图1),构建了运营期隧道病害的故障树模型如图2所示。

图1 运营隧道病害统计

图2 运营期隧道病害的故障树模型

建立了隧道运营期故障树。顶事件是指故障树系统中最不希望发生的事件,因此把隧道运营安全事故(中断或停运)作为顶事件,用T表示。对故障树系统进行分析,造成顶事件T发生的事件有:A1(渗漏水或冻害)、A2(衬砌结构破坏)、A3(附属设施破坏)和X16(洞口仰坡塌方落石)。

对A1进行分析可知,X1(既有排水设施包括衬砌背后的暗沟、盲沟、无衬砌辅助坑道、排水孔和暗槽等年久失修阻塞)、X2(地区气温)、X3(围岩冻胀性)、X4(边墙下部排水孔周边及暗渠式排水孔周边围岩会产生超过固有限界流速的水,对周围软弱围岩渗透和冲刷)均可能会引起渗漏水或冻害。

对A2进行分析可知[9],B1(螺栓接头事故)、B2(管片破坏)、B3(密封垫损坏)都可能会造成衬砌破坏。其中X5(螺栓质量缺陷)、X6(螺栓剪力破坏)、X7(螺栓腐蚀)都可能会造成B1(螺栓接头事故)的发生。X8(管片质量缺陷)、X9(水土压力增大)、X10(材料强度下降)都可能会造成B2(管片破坏)的发生。X11(密封垫材料老化)、X12(腐蚀作用)都可能会造成B3(密封垫损坏)的发生。

对A3进行分析可知,X13(洞内外排水电子系统损坏)、X14(隧道内通风不良)、X15(隧道内照明不良)都可能会造成A3(附属设施破坏)的发生。

2 灰色关联故障树在隧道运营期的应用

2.1 灰色关联故障树的基本原理

故障树系统中,若干个底事件发生就会导致顶事件发生,这些事件的集合称为割集C,表示为{x1,x2,x3........xn},设xi表示底事件的状态变量,当割集C中取消任意一个状态变量xj,顶事件不发生,则称割集C为最小割集。故障树系统中,假设有m个底事件,n个最小割集,系统中包含第i个底事件的最小割集数记为gi[Q( t )],可以定义第i个底事件的相关割集重要度为,此时可以得到m个底事件的相关割集重要度的向量I={I1,I2,I3,....,Im}。

灰色关联的思想是建立一个理想状态的标准,将待检验状态与理想状态进行关联,找出关联度大的待检验状态。假设第ni个最小割集Li含有mi个底事件,则可以构建一个理想状态下的矩阵[10]:

灰色关联故障树就是要将底事件的相关割集重要度向量I与理想状态矩阵L进行关联,找出关联度大的最小割集。而这一关联度大的最小割集就是隧道运营期可能出现病害的主要原因,为实际维护提供必要参考依据。

2.2 灰色关联故障树的应用过程

2.2.1 求底事件发生的概率

根据所建立的故障树,对16个底事件发生概率的求解分为3种方法:

(1)调查法。以沈丹线为例,沈丹线西起辽宁省沈阳市,东到辽宁省丹东市,属沈阳铁路局管辖内主要线路之一。调查数据显示,沈丹线的47座隧道中,有8座出现照明不良,6座隧道通风不良,21座隧道洞内外排水设施损坏。11处(长度约675.5 m)出现洞口仰坡坍方落石现象[3]。

(2)计算法。基于可靠度理论的衬砌结构损坏的失效概率计算方法,运用ANSYS软件计算底事件发生的概率[4]。

(3)专家打分法[11]。邀请专家对底事件的失效概率进行打分,打分依据如表1、表2所示。

表1 底事件发生的严重度A

表2 底事件发生的可能度B

由专家打出严重度A和可能度B的分数,通过P=(A+B)×0.5计算得出。

通过3种方法可以得出16个底事件的失效概率,见表3。

表3 底事件失效概率

2.2.2 求解最小割集

对故障树系统用上行法或者下行法,可以求得使顶事件发生的最小割集L,研究最小割集L可以得到系统的最薄弱环节,以便后续改进。设系统的结构函数为φ(X)=φ(x1,x2,x3........xn),结构函数表示了最小割集和顶事件的关系。

与门连接的系统:

或门连接的系统:根据建立的运营期隧道故障树系统的结构,可以得到该系统的结构函数:

由此可得运营期隧道故障树系统的最小割集,即16个底事件分别都是一个最小割集,得到的理想状态下的矩阵:

2.2.3 求解相关割集重要度向量

若故障树系统含有ci个最小割集,且它们相互独立,则顶事件失效的概率在计算中可以通过下列公式得出顶事件的失效概率:

式中,Pi表示最小割集;ci为发生的概率。

本系统中16个底事件分别都是一个最小割集,由此可以得出顶事件发生的概率P(T)=0.9930,则第i个底事件的相关割集重要度为其中g[Q(t)]=P(T),可以得到16个底事件的相关割集重要度的向量I={I1,I2,I3,.... I16},见表4。

表4 相关割集重要度

I=(0.4532,0.4028,0.3525,0.2014,0.1511,0.1511,0.1007,0.1511,0.1511,0.4532,0.1007,0.1007,0.1712,0.1309,0.4532,0.2316)=3.9579

2.2.4 对I做归一化处理得

I=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

2.2.5 求序列差

用归一化后的I与理想状态矩阵L每一行相减取正,得到序列差如下:

Δ1=(0.8346,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ2=(0.1654,0.8982,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ3=(0.1654,0.1018,0.9109,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.04 33,0.0331,0.1145,0.0585)

Δ4=(0.1654,0.1018,0.0891,0.9491,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ5=(0.1654,0.1018,0.0891,0.0509,0.9618,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0 433,0.0331,0.1145,0.0585)

Δ6=(0.1654,0.1018,0.0891,0.0509,0.0382,0.9618,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ7=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.9746,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ8=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.9618,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ9=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.9618,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ10=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.8855,0.0254,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ11=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.9746,0.0254,0.0433,0.0331,0.1145,0.0585)

Δ12=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145, 0.0254,0.9746,0.0433,0.0331,0.1145,0.0585)

Δ13=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.9567,0.0331,0.1145,0.0585)

Δ14=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.9669,0.1145,0.0585)

Δ15=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.8855,0.0585)

Δ16=(0.1654,0.1018,0.0891,0.0509,0.0382,0.0382,0.0254,0.0382,0.0382,0.1145,0.0254,0.0254,0.0433,0.0331,0.1145,0.9415)

2.2.6 求序列差最大值Δmax和最小值Δmin

根据上述Δi,i=1,2,3......16,可以得出序列差最大值和最小值:Δmax=0.9746,Δmin=0.0254。

2.2.7 求关联系数

关联系数由下式计算得出:

式中,ρ一般取0.5;Δij表示第i中的第j个元素。计算的关联系数见表5。

2.2.8 计算关联度

通过下列公式计算得出关联度,见表6。

根据计算结果排序得出:

3 建议

由此可得底事件X1(既有排水设施包括衬砌背后的暗沟、盲沟、无衬砌辅助坑道、排水孔和暗槽等年久失修阻塞)、X10(材料强度下降)、X15(隧道内照明不良)是隧道运营期故障树系统的薄弱环节,隧道运营维护时应高度重视。

(1)加强对排水设施的维修更换。水害是隧道运营期常见的病害,所谓“十隧九害”正是反映了水害多发性。隧道设计过程中因地制宜的采用了“截、排、堵相结合”的原则,根据这一原则可以看出,隧道运营期维护应该做好“排”的工作,每隔一段时间有针对性地重点巡查排水设施,发现因泥沙或杂质阻塞的排水设施及时清理,遭到破坏的排水设施及时维修更换,确保排水设施性能良好。

表5 关联系数表

表6 关联度

(2)早发现早治理。随着运营时间增长,隧道衬砌受到水害、冻胀和材料自身等原因的影响,材料强度下降,整体老化严重,造成衬砌开裂、脱落等病害,严重的甚至衬砌整体垮塌,造成运行安全事故。因此平时营运排查时,对发现的衬砌渗水引起注意,查明渗水原因,制定治理方案,对老化和腐蚀严重的部位剔除干净,将周围混凝土凿毛,沿裂缝两侧注浆。

(3)优化隧道内部环境。隧道内部空间狭小,照明不良和通风不良给排查人员的排查工作带来了不便,甚至影响排查的结果,对一些潜在的病害不能及时发现,错过了最佳的治理时间。

4 结语

对隧道运营期病害的风险管理,是实现铁路安全运行的关键。本文通过建立灰色故障树模型,对影响隧道安全的风险进行了识别和评价,找出了隧道运营期安全的薄弱环节,针对这些薄弱环节提出了加强对排水设施的维修更换、早发现早治理、优化隧道内部环境的建议。

参考文献:

[1] 刘 强,王翠莲.中国交通概况[EB/OL].新华网:http://www.xinhuanet.com.

[2] 曾水长.既有铁路隧道病害整治[J].铁道建筑,2006(8):40-41.

[3] 袁 超,李树忱,李术才,等.寒区老旧隧道病害特征及治理方法研究[J].岩石力学与工程学报,2011,30 (2):3354-3361.

[4] 唐 亮.隧道病害调查分析及衬砌结构的风险分析与控制研究[D].浙江大学,2008.

[5] 李 彬,雷明锋,李文华.运营公路隧道病害对衬砌结构安全性的影响[J].铁道科学与工程学报,2011(5):40-45.

[6] 危 强,舒中文.兰新线妖魔山隧道病害整治[J].铁道建筑,2014(5):69-71.

[7] 姜 洲,丁保军,张立茂,吴贤国,陈跃庆.基于贝叶斯网络的运营隧道渗漏水病害风险评估[J].环境工程,2014(32):992-997.

[8] 高菊如,张 博,袁 玮,涂文轩.既有线铁路隧道病害综合整治技术与设备配套研究[J].现代隧道技术,2013,50(6):24-31.

[9] 刘思峰,党耀同.灰色系统理论及其应用[M].北京:科学出版社,2010.

[10] 刘静乐,王恩茂.基于灰色关联故障树的绿色建筑设计风险研究[J].工程管理学报,2014,28(2):17-21.

[11] 周继忠.灰色关联故障树分析法在基坑工程放坡开挖中应用[J].土木工程学报,2009,42(8):109-114.

张国喜(1991-), 男,硕士研究生,研究方向:土木工程建造与管理;

李晓钟(1976-), 男,副教授,硕士生导师,研究方向:土木工程建造与管理。

Application of Gray Correlation Fault Tree on Risks of the Defects of the Operational Tunnel

ZHANG Guo-xi,LI Xiao-zhong
(School of Civil Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China,E-mail:395374064@qq.com )

Abstract:Management of the operational tunnel determines the safety of the railway. In order to ensure the safety of operation,

risks of the defects of the operational tunnel must be controlled. Firstly,utilizing the method of fault tree,the tunnel defects are analyzed and found out the factors which affect the safety of the tunnel as the bottom events. Secondly, the importance of the correlation of the cut set is obtained. Finally,according to the grey relation,the great important minimum cut sets of the bottom events including existing drainage facility out of repair and block,reduction in strength of material,the tunnel poor lighting.

Keywords:tunnel defects;gray correlation;fault tree;operate risk;bottom events

作者简介:

收稿日期:2015-11-04.

中图分类号:U451

文献标识码:A

文章编号:1674-8859(2016)01-077-05

DOI:10.13991/j.cnki.jem.2016.01.014

猜你喜欢

故障树灰色关联
某型发动机喘振故障的研究
数控机床液压系统故障诊断专家系统的研究
人民币国际化背景下外汇储备影响因素的灰色关联分析
电子商务对上海低碳城市建设的影响
基于灰色关联模型的风雹灾害经济损失影响因素研究
系统安全性分析技术在空空导弹中的应用
基于熵权法的“互联网+”农业发展影响因素权重确定
民用飞机强度试验故障处理流程研究
基于故障树的中专电梯维修教学研究
基于故障树的大型设备机组故障诊断研究