Improved Hybrid Reasoning Approach to Safety Risk Perception under Uncertainty for Mountain Tunnel Construction

被引:21
作者
Cai, Qijie [1 ]
Hu, Qijun [2 ]
Ma, Guoli [2 ]
机构
[1] Southwest Jiaotong Univ, Sch Transportat & Logist, Chengdu 611756, Peoples R China
[2] Southwest Petr Univ, Sch Civil Engn & Geomat, Chengdu 610500, Peoples R China
基金
中国国家自然科学基金;
关键词
Tunnel collapse; Safety risk; Uncertainty analysis; Extended cloud model (ECM); Dempster-Shafer (D-S) evidence theory; Information fusion; COMBINING BELIEF FUNCTIONS; ANALYTIC HIERARCHY PROCESS; NETWORK-BASED APPROACH; DECISION-ANALYSIS; MANAGEMENT; COMBINATION; CONFLICT; RULES; MODEL;
D O I
10.1061/(ASCE)CO.1943-7862.0002128
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Tunnel projects in mountainous areas are considered a kind of high-risk work due to the complexity of geological conditions, as well as the diversity of construction and management factors. Numerous research efforts have been made to estimate the risk level of tunnel collapse by multicriteria decision-making (MCDM) methods considering a variety of risk factors, aiming to provide a decision basis for risk management. However, the uncertainty of fuzziness and randomness existing in evaluation factors gradation are usually neglected. The information fusion method based on extension theory (ET) is an effective tool for uncertain information reasoning. Nevertheless, the potential contribution of each piece of evidence to the fusion result has not been given enough attention. To overcome these deficiencies, this paper presents a novel hybrid reasoning approach to perceive the risk level of mountain tunnel collapse under uncertainty by integrating the extended cloud model (ECM) and improved Dempster-Shafer (D-S) evidence theory at the construction stage. ET and cloud model (CM) are merged to construct the basic probability assignments (BPAs) of the influenceable variables corresponding to different risk states under uncertainty, i.e., the randomness and fuzziness. The overall risk level is derived from a fusion of all the influenceable factors using the improved D-S theory developed in this paper. Specifically, the weight of each piece of evidence is assessed considering its importance to risk perception results and used to improve the evidence source by the discounting method before evidence fusion. Evidence conflicts are resolved through a weighted-average fusion process based on the evidence distance. Besides, a confidence indicator is introduced to assess the reliability of the fusion results. Finally, a case study in Canglongxia Tunnel is employed to verify the validity of the developed approach. The method proposed in this paper has the potential to reduce the deviation of risk assessment results for tunnel collapse in mountain tunnel construction induced by the uncertainty of risk factors.
引用
收藏
页数:15
相关论文
共 55 条
[1]  
An Yong-lin, 2011, Journal of Central South University (Science and Technology), V42, P514
[2]   Risk assessment and risk management: Review of recent advances on their foundation [J].
Aven, Terje .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2016, 253 (01) :1-13
[3]   Tunnelling in difficult ground: a case study from Dranaz tunnel, Sinop, Turkey [J].
Aydin, A ;
Ozbek, A ;
Cobanoglu, I .
ENGINEERING GEOLOGY, 2004, 74 (3-4) :293-301
[4]   Engine fault diagnosis based on multi-sensor information fusion using Dempster-Shafer evidence theory [J].
Basir, Otman ;
Yuan, Xiaohong .
INFORMATION FUSION, 2007, 8 (04) :379-386
[5]   Extension theory and its application [J].
Cai, W .
CHINESE SCIENCE BULLETIN, 1999, 44 (17) :1538-1548
[6]  
Chen JJ, 2009, ROCK SOIL MECH, V30, P2365
[7]   Risk assessment methodology for underground construction projects [J].
Choi, HH ;
Cho, HN ;
Seo, JW .
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT, 2004, 130 (02) :258-272
[8]   Fuzzy logic-based fault-type identification in unbalanced radial power distribution system [J].
Das, B .
IEEE TRANSACTIONS ON POWER DELIVERY, 2006, 21 (01) :278-285
[9]   UPPER AND LOWER PROBABILITY INFERENCES BASED ON A SAMPLE FROM A FINITE UNIVARIATE POPULATION [J].
DEMPSTER, AP .
BIOMETRIKA, 1967, 54 :515-&
[10]  
Dempster AP, 2008, STUD FUZZ SOFT COMP, V219, P57