Research Article  Open Access
Zhenhua Luo, Li Zeng, Haize Pan, Qijun Hu, Bo Liang, Jianqiang Han, "Research on Construction Safety Risk Assessment of New Subway Station CloseAttached Undercrossing the Existing Operating Station", Mathematical Problems in Engineering, vol. 2019, Article ID 3215219, 20 pages, 2019. https://doi.org/10.1155/2019/3215219
Research on Construction Safety Risk Assessment of New Subway Station CloseAttached Undercrossing the Existing Operating Station
Abstract
With the largescale construction of urban rail transit, it will lead to the intersection and transfer of various lines, resulting in more transfer stations. The transfer station is a collection point for multiple subway lines, which is difficult to construct and has a high construction risk. The construction of the new subway station and the operation of the existing subway station are mutually influenced during the closeattached undercrossing construction. Considering the two objectives of ensuring the smooth operation of the existing subway station and the safe construction of the new subway station, this paper comprehensively analyzes the possible safety risk factors during the construction of the new subway station closeattached undercrossing the existing operating station and identifies 75 preliminary risk factors by means of literature review and onsite investigation. Then the Delphi Method and Entropy Weight Method are used to screen the preliminary risk factors, and the main risk factors with greater influence are retained, so that 49 key risk factors are obtained. According to the list of key risk factors, a safety risk assessment index system including 2 firstlevel indexes, 12 secondlevel indexes, and 49 thirdlevel indexes is established. Based on the index system, this paper establishes a safety risk assessment model by using Analytic Hierarchy Process (AHP) and Fuzzy Matter Element Method (FMEM). The model first calculates the weight of each index by using AHP, calculates the comprehensive correlation degree of each index by using FMEM, classifies the risk grade of each index according to the comprehensive correlation degree, and determines the risk grade of the project. Finally, the safety risk assessment model is applied to the Dongdalu Station project of Chengdu Rail Transit Line 8. The result shows that the risk grade of this project is moderate risk, which is basically consistent with the actual situation, indicating that the model has good practicability. In this paper, a new safety risk assessment model for subway closeattached undercrossing construction is proposed, which fills the gap in the field of safety risk assessment for the construction of the new subway station closeattached undercrossing the existing operating station.
1. Introduction
With the continuous progress of urbanization in China, the population density and traffic pressure are also increasing. In order to solve the problem of urban traffic congestion, subway construction has become the preferred solution for major cities in China. By the end of 2018, there were 37 cities with urban rail transit in mainland China, with a total length of 5539.19 km [1]. Compared with other projects, the subway project has more outstanding concealment, construction complexity, and uncertainty of stratum conditions and surrounding environment, which increases the difficulty of construction technology and the risk of construction. According to statistics, during the period from 2002 to 2016 (statistics to March 2016), 246 accidents occurred in the construction of subways in China. There were 238 accidents with clear and specific location records, of which subway station projects accounted for 57% and subway section projects accounted for 43% [2]. For example, on November 15 of 2008, the site collapse of Xianghu Station of Hangzhou Subway resulted in 21 deaths and a direct economic loss of 49.61 million yuan [3]. It can be seen that, in the construction of underground engineering, accidents occur frequently and the situation is very serious. Therefore, the necessity and urgency of implementing safety risk management in underground engineering construction is very obvious.
Risk analysis is a tool that can establish an active safety strategy by investigating potential risks [4]. The application and research on safety risk management of underground engineering have been actively carried out worldwide. Kampmann [5] used risk assessment technology to propose 10 types of risk for the Copenhagen Subway project, including more than 40 types of disasters, and proposed a specific classification system for the possibility of events and their impact results. Based on the actual project, through the safety risk analysis and assessment, Burland.JB [6, 7] calculated the possible damage caused by the buildings along the line during the design and planning stage and gave corresponding reinforcement measures. Later, the research result was applied to the Jubillee line in London, which ensured the smooth construction of the project. In 2004, the International Tunnel Association (ITA) wrote the Guidelines for Tunnel Risk Management [8], which provided a set of reference standards and methods for tunnel risk management and affirmed the practicability of safety risk assessment.
In terms of specific engineering examples, due to the safety risk assessment before construction and good construction control measures during construction, in a South Korean project where the subway station undercrossed the existing subway, the settlement of the existing subway line was only 3mm during the whole construction process [9], which is lower than the 10mm required by China’s national normative standards [10]. In the project where the northbound tunnel section of the I93 Interstate Highway in Boston, USA, passed through the South Station of Red Line, through the 24hour monitoring of the construction process, the safety risks of the entire construction process were accurately assessed and responded, and the normal operation of the existing station and its ancillary facilities was guaranteed [11]. Cao Zhen [12] formed a safety risk system for shield construction of Xi'an Subway Tunnel according to four aspects of engineering geology and hydrogeology, load factor, construction factor, and external factor. Through the analysis of shield construction safety of Xi’an Subway Line 2, three key risk sources, existing railway, moat arch bridge, and ancient city wall, were obtained. The construction safety risk grade can be reduced from grade I to grade IV by the proposed reasonable construction plan. Zhang L [13] conducted a risk assessment of the project based on the importance index of the possibility and the fuzzy importance to reveal the key method of reducing the risk limit. Wang ZZ [14] used the Fuzzy Comprehensive Bayesian Network (FCBN) which combined with Fuzzy Comprehensive Evaluation Method (FCEM) and Bayesian Network (BN) to analyze the risk of the subway project. The comparison between the analysis result and the example proved that this method can be effective to estimate the risk grade of the subway construction project. Peng Yuhan [15] formed a set of assessment system combining the risk factors of engineering characteristics and environmental characteristics, made a risk assessment on the Suzhou Street Station Project of Beijing Subway Line 16, and finally gave corresponding countermeasures for different risk grades. Wang F [16] identified the causes of land subsidence by statistical model based on adaptive Correlation Vector Machine (aRVM) and put forward risk countermeasures for highrisk factors, which effectively ensured the normal operation of the construction.
With the largescale construction of urban rail transit, it will bring about the intersection and transfer of various lines, resulting in more transfer stations. During the construction of the closeattached undercrossing engineering of transfer stations, the new subway construction and the existing subway structure will affect each other. The construction of the new subway project will affect the safety of the existing station structure, and the existing lines will also affect the construction safety of the new subway project [17].
With the increase of subway crossing projects, some scholars have begun to explore the research on risk management of subway crossing projects. Based on the risk dynamic analysis method, Guan Jifa [18] constructed a risk assessment system for a tunnel undercrossing the existing subway and put forward the idea that the settlement amount (the uplift amount) and settlement rate (the uplift rate) of the existing subway structure baseplate should be used as the main control indexes in the subway crossing project. The risk control system proposed by Yang Wendong [19] was mainly based on technical indexes and supplemented by nontechnical indexes. An underground line undercrossing a station of Subway Line 4 in Beijing was taken as an example to verify that the risk control system can effectively avoid and deal with risks. Li X [20] made a research on the safety control framework for shield tunneling in close proximity to the operating subway tunnels, proposed that the use of remedial grouting measures can effectively correct the excessive settlement problem in closeattached undercrossing projects, and suggested using the optical fiber measurement to dynamically monitor the project.
Because of the characteristics of underground engineering, such as large investment, long construction period, complex construction specialty, and many unpredictable risk factors, it is very important to carry out reasonable and effective risk management for underground engineering. In recent years, many scholars have studied the safety risk management of underground engineering, but there are few studies on the safety risk management of subway crossing engineering. At present, some scholars have studied the risk management of tunnels undercrossing existing subway stations, but there is no research on the risk management of the new subway station closeattached undercrossing the existing operating station. In the crossing project of new subway station closeattached undercrossing the existing operating station, both the normal operation of the existing station and the normal construction of the new station should be guaranteed, which makes the construction of the crossing project more complex and risky than the ordinary subway project. Therefore, it is necessary to put forward a risk assessment system suitable for the subway crossing project of new subway station closeattached undercrossing the existing operating station.
2. Establishment of the Construction Safety Risk Assessment Index System for New Subway Station CloseAttached Undercrossing the Existing Subway Station
2.1. Research Framework
The construction of new subway station closeattached undercrossing the existing station has the characteristics of complex technology and high risk, so it is of great significance to carry out risk assessment. How to successfully complete the construction of the new subway station under the normal operation of the existing subway station and how to ensure the two objectives of the smooth operation of the existing subway station and the safe construction of the new subway station are major technical problems faced by such crossing projects. Therefore, establishing a risk assessment model for the new subway station closeattached undercrossing the existing station can provide a theoretical basis for safety risk management of similar projects, help construction units to effectively control the risks of similar projects, and prevent or reduce the possible adverse effects in the construction process.
The research ideas of this paper are as follows: the first step is to identify preliminary risk factors by means of literature searching and onsite investigation. The second step is to use Delphi Method and Entropy Weight Method to screen the preliminary list of risk factors and retain the major risk factors that have a greater impact, so as to obtain the list of key risk factors. The third step is to use the list of key risk factors to construct a risk assessment index system and to divide the risk grade range of each index according to national norms and expert opinions. Fourthly, according to the established assessment index system, the appropriate assessment method is selected to establish a risk assessment model. In the assessment model, the weight of each index is calculated by Analytic Hierarchy Process (AHP), and then the comprehensive correlation degree of each index corresponding to different risk grades is calculated by Fuzzy Matter Element Method (FMEM). According to the comprehensive correlation degree, the risk grade of each index is graded, and the risk grade of the project is determined. Finally, the corresponding risk control measures are put forward for the indexes with higher risk grade in the assessment result. The specific research framework flow chart is shown in Figure 1.
2.2. Identifying Preliminary Risk Factors
To ensure the smooth operation of the existing subway station and the safe construction of the new subway station is the purpose of establishing and applying the risk assessment model. Therefore, the firstlevel indexes of the assessment index system are considered from two aspects: the existing subway station and the new subway station. Through onsite investigation of similar projects and consulting the “Technical Specification for Urban Rail Transit” promulgated by the Ministry of Housing and UrbanRural Development of the People's Republic of China [10, 21–28], combined with the opinions of relevant experts, a list of preliminary risk factors is finally determined, which includes 2 firstlevel indexes, 12 secondlevel indexes, and 75 thirdlevel indexes. The list of preliminary risk factors is shown in Table 1.

These 12 secondlevel indexes in Table 1 include 75 thirdlevel indexes, which are composed as follows:
Station structure (A_{1}): A_{11}: radial displacement of the station structure, A_{12}: absolute settlement value of station structure, A_{13}: horizontal displacement of the station structure, A_{14}: station structure floating, A_{15}: differential settlement of station structure deformation joint, and A_{16}: crack width of station structure
Track structure (A_{2}): A_{21}: track gauge, A_{22}: orbital height difference (Vector value), A_{23}: vibration velocity, A_{24}: transverse elevation difference of track, and A_{25}: void amount of trackbed
Power supply system (A_{3}): A_{31}: area power outage renovation, A_{32}: circuit failure, and A_{33}: fault of power supply equipment
Water supply and drainage system (A_{4}): A_{41}: system blocking, A_{42}: mechanical fault, and A_{43}: pipeline damage
Ventilation and signal system (A_{5}): A_{51}: equipment sampling pass rate, A_{52}: maintenance pass rate, A_{53}: equipment failures, A_{54}: operating rate of workers with diseases, A_{55}: protective equipment completion rate, A_{56}: equipment renewal rate, and A_{57}: equipment pending repair rate
Construction monitoring of existing station (B_{1}): B_{11}: information monitoring equipment failures, B_{12}: accuracy rate of feedback realtime data, B_{13}: stress of column structure, B_{14}: horizontal and vertical displacement of retaining pile (wall) and top of slop, B_{15}: monitoring frequency, B_{16}: axial force of struts, B_{17}: anchor tension, B_{18}: ground settlement, B_{19}: clearance convergence of shaft support structure, B_{110}: monitoring hole of groundwater level, B_{111}: distance from excavation face to monitoring point or monitoring section, and B_{112}: vertical displacement of column structure
New space enclosing structure (B_{2}): B_{21}: survey spot arrangement spacing, B_{22}: stress of the retaining wall, and B_{23}: lateral displacement of space enclosing structure
Engineering dewatering (B_{3}): B_{31}: stratum filtering flow, B_{32}: precipitation depth, B_{33}: sand content control, and B_{34}: aquifer characteristics
Foundation pit excavation (B_{4}): B_{41}: foundation pit protection, B_{42}: foundation pit excavation plan, B_{43}: maintenance of foundation pit dewatering system, and B_{44}: depth of foundation pit
Surrounding rock and soil mass and surrounding environment (B_{5}): B_{51}: design of station structure, B_{52}: stratum and geological condition, B_{53}: site classification, B_{54}: soil liquefaction grade, B_{55}: buried depth, B_{56}: surface settlement, B_{57}: underground pipeline monitoring, B_{58}: effect of water level reduction on stratum, and B_{59}: impact of surrounding buildings on construction
Construction technology management of main structure (B_{6}): B_{61}: stiffness and deformation reinforcement of main structure, B_{62}: stiffness of initial support, B_{63}: quality of secondary lining, B_{64}: technical training for construction personnel, B_{65}: mechanical operation, B_{66}: construction personnel management, B_{67}: emergency plan, and B_{68}: safety drill
Green construction (B_{7}): B_{71}: public building green rate, B_{72}: ground traffic control, B_{73}: ratio of underground building area to total land area, B_{74}: noise control, B_{75}: sewage control, B_{76}: light pollution control, B_{77}: dust control, B_{78}: construction trash control, B_{79}: impact on business climate, B_{710}: underground space lighting, and B_{711}: the proportion of the weight of building materials produced locally to the total weight of building materials
2.3. Selecting Key Risk Factors
In the list of preliminary risk factors, the occurrence of some risk factors may have a greater adverse impact on the project, and even the consequences may be unacceptable to managers; such risks are the key risk factors. Project managers should put limited time, energy, and funds on the prevention and control of key risk factors, so how to sort out and select key risk factors from the list of many preliminary risk factors is crucial.
For multivariate and multilevel index system, Entropy Weight Method is an ideal method to objectively determine index weights, which can effectively use the information contained in each index to measure the difference between indexes, and avoid the subjectivity of weight assignment [29–31]. Based on the comprehensive consideration of the characteristics and the risk management objectives of the subway crossing project, this paper consults relevant technologies and management experts to sort out the preliminary risk factors and uses Entropy Weight Method to select and identify key risk factors.
2.3.1. Obtaining Raw Data
The risk factors in the preliminary risk factor list are classified according to different risk degrees, and the risk degree classification standards are shown in Table 2.

The subway project has the characteristics of high cost and complex technology, involving many specialties and participating units. In order to avoid the subjective onesidedness of the investigation caused by the small number of experts and the single source of experts, six types of units, such as owner, geological prospecting unit, designing unit, construction unit, supervision unit, and subcontractor, are investigated on the spot. A total of 30 experts are selected from these six types of units for consultation. The number of experts from various units is shown in Figure 2. In practice, if 30 experts cannot be invited for some objective reasons, at least 6 experts from different units should be invited.
Firstly, 30 experts are invited to score and classify the risk degree of 75 preliminary risk factors according to the risk degree classification standards in Table 2. Secondly, according to the different units, the average value of experts’ scores in each unit is calculated, and six average values from six units are obtained. Thirdly, the evaluation value matrix ( denotes the ordinal number of secondlevel indexes, t = 1, 2, …, 12) is calculated. In the matrix, the number of rows is the number of expert source units, and the number of columns is the number of thirdlevel indexes included in each secondlevel index. Because the arithmetic mean of the expert scores is calculated separately according to the six types of units, is a matrix of six rows and n columns, of which the value of n is the number of thirdlevel indexes included in each secondlevel index.
For example, the secondlevel index “station structure” has six thirdlevel indexes, so its evaluation value matrix is
2.3.2. Standardization of Raw Data
The evaluation matrix is standardized to eliminate the dimensional differences among the indexes. If the evaluation value is bigger, the better, the standardized formula (2) can be used. If the evaluation value is smaller, the better, the standardized formula (3) can be used. is an element corresponding to the th row and the th column in the evaluation value matrix . is the standardization result of :
Because the evaluation value of risk factors in subway construction belongs to the smaller the better type, formula (3) is adopted to standardize the evaluation value matrix. For example, matrix is the standardization result of the evaluation value matrix (station structure), which is as follows:
2.3.3. Determining Index Entropy Weight
The evaluation value weight of the index is
The entropy value of the index is
Among them, , when the evaluation value , .
The entropy weight of the index is
In order to solve the problem of numerous indexes and complicated calculation, formulas (6) and (7) are compiled into Excel program. By inputting the evaluation value weight of each index into Excel program, the entropy value and entropy weight of each index can be obtained.
According to formula (5), the evaluation value weight matrix of the secondlevel index “station structure ()” is
By inputting the value of matrix into Excel program, the entropy value and entropy weight of each thirdlevel index can be obtained. The calculation results are shown in Table 3.

Referring to the calculation results in Table 3, according to the characteristics that the smaller the entropy weight is, the larger the weight is, those indexes with smaller entropy weight are selected as the key risk factors. For example, four indexes with smaller entropy weight in Table 3 are taken as key risk factors, namely, A_{12}, A_{13}, A_{15}, and A_{16}, and A_{11} and A_{14} are deleted. Similarly, the entropy value and entropy weight of other thirdlevel indexes are calculated, and the key risk factors are determined. From this, a list of key risk factors is obtained. The calculation results are shown in Table 4.

In summary, after the list of preliminary risk factors including 75 risk factors was screened by Delphi Method and Entropy Weight Method, 26 risk factors with relatively less impact were deleted, and a list of key risk factors including 49 risk factors is obtained.
2.4. Establishing the Construction Safety Risk Assessment Index System
According to the selected list of key risk factors (Table 4), a construction safety risk assessment index system of the new subway station closeattached undercrossing the existing operating station is established. The index system is divided into 2 firstlevel indexes: the existing subway station and the new subway station, including 12 secondlevel indexes and 49 thirdlevel indexes, as shown in Table 5.

By consulting to the relevant national technical specifications and engineering site data, the risk grade of the thirdlevel indexes in the construction safety risk assessment index system is divided into three levels: high risk, medium risk, and low risk, as shown in Table 5.
3. Assessment Approach
3.1. Superiority of the Method
After the establishment of the construction safety risk assessment index system, how to choose a reasonable and effective assessment method is very important to establish a construction safety risk assessment model. After consulting the relevant literature [32–38], this paper uses Analytic Hierarchy Process (AHP) and Fuzzy Matter Element Method (FMEM) to evaluate the construction safety risk of the new subway station closeattached undercrossing the existing station qualitatively and quantitatively. Firstly, the weights of the indexes are calculated by AHP, and then FMEM is used to determine the risk grade of each index and the project.
In order to facilitate the research of problems with multiple contents, it can be transformed into quantitative problems and then studied. AHP is a common method to solve such problems. For the complex index system, the weight of the index occupies an important position. Using AHP to calculate index weight can reduce subjective factors and improve the scientific nature of weight determination.
Fuzzy Matter Element Method (FMEM) was proposed by Chinese scholar Cai Wen in the 1980s. The main point of this method is to describe things with “object, features and quantized values.” Matter element refers to the basic elements of an orderly group composed of these three elements. If the object has n characteristics and corresponding quantized values, it is called ndimensional matter element. Matter element analysis is to study the matter element and its changes, mainly to study the incompatibility of the real world. The purpose of adding fuzzy mathematics to matter element analysis is to deal with the quantity ingredient with fuzziness in matter element, and the combination of these two is fuzzy matter element analysis [35–38].
The application of FMEM in risk assessment has a number of advantages. Firstly, it can describe the complex risk assessment index comprehensively and clearly. Secondly, it can describe the fuzzy state of the assessed object in the form of fuzzy subset vector and objectively evaluate various indexes. Thirdly, it can be used not only for the comprehensive assessment of subjective risk factors to solve fuzzy phenomena with unclear boundaries and that are difficult to quantify in various indexes, but also for the comprehensive assessment of objective risk factors to make the assessment process clear and scientific.
3.2. Construction Safety Risk Assessment Model Based on AHP and FMEM
The construction of new subway station closeattached undercrossing the existing subway station should not only ensure the safe construction of the new subway station, but also ensure the normal operation of the existing subway station. Its risk assessment can be seen as a problem with two incompatible objectives. Combining AHP and FMEM, the construction safety risk assessment model is constructed. The detailed calculation process is shown in Figure 3. The main calculation steps are as follows.
3.2.1. Calculating Index Weight
AHP is used to calculate index weight. Firstly, the judgement matrix of all levels of indexes is constructed by synthesizing the judgement of experts. Secondly, whether the judgement matrix passes the consistency test is observed. If it passes the test, the weight of each index in the judgment matrix can be obtained.
3.2.2. Determining Matter Element
In the construction safety risk assessment model, FMEM is used to establish the matter element composed of “risk assessment, assessment indexes, and quantized values.” Risk assessment , risk assessment feature , and feature quantified value together constitute the matter element to be assessed, which is
3.2.3. Determining Matter Element Matrix of Classical Domain and Joint Domain
The classical domain matter element matrix of the construction safety risk assessment can be expressed aswhere is the classical domain matter element; is the th level of the classified risk grade, ; is the th assessment index; is the quantized value range corresponding to the th risk grade, that is, the classical domain.
The joint domain matter element matrix of the construction safety risk assessment can be expressed aswhere is the joint domain matter element; is the quantized value range of the feature ; is the overall level of risk grade.
3.2.4. Calculating the Correlation Function and Correlation Degree to Determine the Risk Grade of Each ThirdLevel Index
The index correlation function of construction safety risk assessment is defined aswhere is the distance between point and finite interval ; is the distance between point and finite interval ; is the quantized value of matter element to be assessed; is the quantized value range of the classical domain matter element; is the quantized value range of the joint domain matter element.
3.2.5. Calculating the Comprehensive Correlation Degree and Determining the Risk Grade of the Project
The comprehensive correlation degree of with respect to risk grade is as follows:where is the comprehensive correlation degree; is the single correlation degree ; is the weight of each index. Ifthen the th index belongs to the th grade of construction safety risk assessment. Ifthen belongs to the th grade of construction safety risk assessment.
When , it indicates that the unit to be assessed meets the requirements of the assessment standard; when , it indicates that the unit to be assessed does not meet the requirements of a certain grade of assessment standard, but the unit to be assessed has the conditions to convert to the assessment standard, and the smaller the value is, the easier it is to convert; when , it indicates that the unit to be assessed does not meet the requirements of a certain grade of assessment standard and does not have the conditions to convert to the assessment standard.
4. Case Study
4.1. Project Profile
Dongdalu Transfer Station of Chengdu Rail Transit Line 8 is located at Huiyuan Road with a road width of 30m. The transfer station is arranged in the eastwest direction, using a 13.5 m island platform. The station is a threestory, threespan boxtype underground frame structure. The main body of the station is constructed by opencut method, and the underground excavation method is used at the transfer node of the station and Chengdu Rail Transit Line 2. The total length of the transfer station is 306.2m, the width of the standard section is 23m, the covering soil is 2.6m, and the buried depth of the station baseplate is 24.2m. The station transfers with the operating Line 2 in the form of Ttype node transfer, as shown in Figure 4. At this Ttype node, the existing twostory underground station is on Line 2, the new threestory underground station is on Line 8, and the new station on Line 8 is closeattached undercrossing the existing operating station on Line 2.
4.2. Construction Safety Risk Assessment
4.2.1. Calculating Index Weight
As discussed above, 30 experts were consulted in the selecting of key risk factors. These 30 experts are invited to grade and judge each index of the risk assessment index system in Table 5, and the judgment matrix of each index at all levels is established. Then the judgment matrix is input into the Matlab program based on AHP to determine whether it passes the consistency test and get the weight of each index.
For example, the judgment matrix of the firstlevel index “existing subway station (A)” is established, as is shown in Table 6.

CI = 0.0170, CR = 0.0152 are obtained by AHP. CR < 0.1 means that the consistency test has been passed, and the calculated weights of each secondlevel index under the firstlevel index “existing subway station (A)” are 0.1599, 0.4185, 0.2625, 0.0618, and 0.0973, respectively.
Similarly, the weight of other secondlevel and thirdlevel indexes can be determined. The calculation results are shown in Table 7.

4.2.2. Determining Matter Element to Be Assessed
In the comprehensive assessment, the assessed unit is an ndimensional matter element, and its eigenvalues vary according to its influence. The 30 experts consulted during the selecting of key risk factors are invited to estimate the specific status of the indexes in the index system in Table 5. For example, experts estimate that the absolute settlement value of the station structure (A_{12}) of this project may be 4 mm based on the practical experience of similar projects, and the evaluation value of this index is 4. Similarly, the evaluation values of other indexes can be obtained, and the evaluation values are averaged arithmetically to get as shown below.
4.2.3. Determining Matter Element Matrices of Classical Domain and Joint Domain
The determination of the classical domain is the basis of Fuzzy Matter Element Method. According to the extensibility of construction safety risk assessment of new subway station closeattached undercrossing the existing subway station, the risk grade of each index is divided into three grades, namely, , , and , which are qualitatively described as high risk, medium risk, and low risk. The classic domain matter element matrices , , and and the joint domain matter element matrix established based on Table 5 are shown in Table 8.

4.2.4. Calculating the Correlation Degree to Determine the Risk Grade of Each ThirdLevel Index
The values of the matter element are input into Excel program which has been compiled based on the FMEM, and the corresponding calculation results are shown in Table 9.

4.2.5. Calculating the Comprehensive Correlation Degree and Determining the Risk Grade of the Project
The comprehensive correlation degree is calculated according to FMEM, and the comprehensive correlation degree and risk grade of each secondlevel index can be obtained as shown in Table 10.

Combining the correlation degree of each index and the corresponding index weight in Table 7, the comprehensive correlation degree of each firstlevel index is obtained, as shown in Table 11.

The correlation degree of each index and the corresponding index weight in Table 7 are input into formula (18) to obtain the comprehensive correlation degree of the matter element to be assessed, as shown in Table 12.

4.3. Result Analysis
4.3.1. Result Analysis of Existing Subway Station
The risk grade of the firstlevel index “existing subway station” is medium risk, which indicates that the closeattached undercrossing project is easy to affect the existing station. According to Table 10, in the secondlevel indexes, the risk grade of “station structure” and “track structure” is medium risk, while the risk grade of “power supply system,” “water supply and drainage system,” and “ventilation and signal system” is low risk.
Table 9 shows that, in the thirdlevel indexes, the risk grade of “absolute settlement value of station structure,” “crack width of station structure,” and “void amount of trackbed” is medium risk. To guard against these risk factors, it is necessary to have strict control of each working procedure that may cause settlement in the construction process. As the structural settlement and deformation caused by construction are the main factors affecting the deformation of station structure and track structure, good construction measures to control the value of station settlement within the allowable range can ensure the normal operation of existing operating station.
The results of risk assessment are compared with the actual engineering data. The risk assessment results of the qualitative indexes in the thirdlevel indexes are consistent with the actual engineering measurement results. The risk assessment results of quantitative indexes and the actual engineering measurement results are compared and analyzed as shown in Table 13. There are 15 quantitative indexes in Table 13. Among them, the risk grade corresponding to the risk assessment results of 13 indexes is consistent with the risk grade corresponding to the actual engineering measurement results, and 2 indexes are inconsistent. For instance, the risk grade corresponding to the expert assessment results of “pipeline damage” and “equipment pending repair rate” is medium risk, while the risk grade corresponding to the actual engineering measurement results is low risk, which indicates that the construction unit has strengthened the management of pipelines and equipment on site according to the result of risk assessment and has taken effective control measures to reduce the risk grade of these indexes.

4.3.2. Result Analysis of New Subway Station
The risk grade of firstlevel index “new subway station” is medium risk, which indicates that the operation of existing subway station will have a great impact on the construction of new subway station. From Table 10, it can be concluded that, in the secondlevel indexes, the risk grade of “construction monitoring of existing station,” “engineering dewatering,” “foundation pit excavation,” “surrounding rock and soil mass and surrounding environment,” and “green construction” is medium risk, while the risk grade of “new space enclosing structure” and “construction technology management of main structure” is low risk.
Table 9 shows that, in the thirdlevel indexes, the risk grade of “accuracy rate of feedback realtime data,” “depth of foundation pit,” “buried depth,” “ground traffic control,” and “construction trash control” is high risk. In order to prevent the occurrence of these risk factors, it is necessary to strengthen the protective work and monitoring of the station structure and line of the existing station and analyze the monitoring data in time. For the alarm data issued by the monitoring device in the construction process, the operator should respond in time, report to the superior department immediately, and take effective control measures. Then, it is necessary to pay attention to the monitoring of various indicators in the construction process of new subway station and to timely analysis of monitoring data to ensure the accuracy rate of data feedback. At the same time, the construction unit should reasonably guide the ground transportation, and the trash generated in the construction process should be transported to the designated location in time, and protective measures should be taken in the transportation process to prevent dust, soil pollution, and other problems.
The results of risk assessment are compared with the actual engineering data. The risk assessment results of the qualitative indexes in the thirdlevel indexes are consistent with the actual engineering measurement results. The comparative analysis of the risk assessment results of the quantitative indexes and the actual engineering measurement results is shown in Table 14. There are 23 quantitative indexes in Table 14. Among them, the risk grade corresponding to the risk assessment results of 19 indexes is consistent with the risk grade corresponding to the actual engineering measurement results, and 4 indexes are inconsistent.

For instance, the risk grade corresponding to risk assessment result of “accuracy rate of feedback realtime data” is high risk, while the risk grade corresponding to the actual measurement result is medium risk. The risk grade corresponding to risk assessment result of “monitoring hole of groundwater level,” “survey spot arrangement spacing,” and “dust control” is medium risk, while the risk grade corresponding to the actual measurement result is low risk. It shows that the results of the risk assessment model have a certain warning effect on the construction unit. The construction unit has strengthened the feedback of realtime data on site, strengthened the layout of groundwater level monitoring holes, the spacing control of survey spot arrangement, and dust control, and adopted effective prevention and control measures to reduce the risk grade of these indexes.
In summary, the risk grade of the closeattached undercrossing project is medium risk. The assessment result is basically consistent with the actual engineering situation, which indicates that the construction safety risk assessment model has certain practicability.
5. Conclusion
In order to solve the problem of safety risk assessment in the construction of new subway station closeattached undercrossing the existing station, this paper considers the two objectives of ensuring the smooth operation of existing subway station and the safe construction of new subway station and uses Delphi Method and Entropy Weight Method to select the list of key risk factors from the list of preliminary risk factors. According to the list of key risk factors, the construction safety risk assessment index system is established. Based on the assessment index system, a construction safety risk assessment model is established by using AHP and FMEM. The assessment result of the model is basically in line with the actual situation through case analysis, which shows that the model has good practicability.
In this paper, Delphi Method and Entropy Weight Method are used to select key risk factors, AHP and FMEM are applied to assess construction safety risk, and a new construction safety risk assessment model for subway crossing projects is proposed, which provides a theoretical basis for the construction safety risk assessment of new subway station closeattached undercrossing the existing operating station.
Data Availability
The data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
The authors acknowledge financial support from the National Natural Science Foundation of China (No. 51574201) and the China Scholarship Council (No. 201808515050).
References
 “Statistical Analysis Report on the Most Whole Rail Transit in Mainland China in 2018,” DB/OL, 2019, http://www.sohu.com/a/287064842_729676. View at: Google Scholar
 H. Li, Q. Li, and Y. Lu, “Statistical analysis on regularity of subway construction accidents from 2002 to 2016 in China,” Urban Rapid Rail Transit, vol. 30, no. 1, pp. 12–19, 2002. View at: Google Scholar
 R. Wan, “The site collapse accident of Xianghu Station of Hangzhou Metro was recognized as a major responsibility accident,” Urban Mass Transit, no. 3, p. 74, 2010. View at: Google Scholar
 A. Labib and M. Read, “Not just rearranging the deckchairs on the Titanic: learning from failures through risk and reliability analysis,” Safety Science, vol. 51, no. 1, pp. 397–413, 2013. View at: Publisher Site  Google Scholar
 J. Kampmann, S. D. Eskesen, and J. W. Summers, “Risk assessment helps select the contractor for the Copenhagen Metro System,” in Proceedings of the World Tunnel Congress on Tunnels and Metropolise, vol. 98, São Paulo, Brazil, April 1998. View at: Google Scholar
 J. B. Burland and C. P. Wroth, “Settlements on buildings and associated damage,” in Proceedings of Conference on Conference on Settlement of Structures, pp. 611–654, Cambridge, Uk, April 1974. View at: Google Scholar
 J. B. Houlsby and C. E. Houlsby, “Modeling tunnelinginduced settlement of masonry buildings,” Institution of Civil Engineers and Geotechnical Engineering, p. 1729, 2000. View at: Google Scholar
 S. D. Eskesen, P. Tengborg, J. Kampmann, and T. Holst Veicherts, “Guidelines for tunnelling risk management: international tunnelling association, working group no. 2,” Tunnelling and Underground Space Technology, vol. 19, no. 3, pp. 217–237, 2004. View at: Publisher Site  Google Scholar
 S. Chang and S. Moon, “A case study on instrumentations of a large tunnel crossing under the existing subway structure,” in Proceedings of the KGS 2000 Spring Conference, pp. 56–59, 2000. View at: Google Scholar
 MOHURD China, “Code for monitoring measurement of urban rail transit engineering,” Tech. Rep., China Architecture & Building Press, Beijing, China, 2013. View at: Google Scholar
 B. Brenner and T. Neff, “Desing and management for underground uncertainty on the central artery/tunnel projectct,” in North American Tunneling, 2000. View at: Google Scholar
 Z. Cao, The Research on Safety Risk Assessment and Disater Control Technology of Metro Shield Construction in Xi’an, Xi’an University of Science and Technology, 2013.
 L. Zhang, M. J. Skibniewski, X. Wu, Y. Chen, and Q. Deng, “A probabilistic approach for safety risk analysis in metro construction,” Safety Science, vol. 63, no. 3, pp. 8–17, 2014. View at: Publisher Site  Google Scholar
 Z. Z. Wang and C. Chen, “Fuzzy comprehensive Bayesian networkbased safety risk assessment for metro construction projects,” Tunnelling and Underground Space Technology, vol. 70, pp. 330–342, 2017. View at: Publisher Site  Google Scholar
 Y. Peng, Beijing Subway Line 16 SuZhouJie Station Construction Project Risk Analysis, Beijing Institute of Technology, 2016.
 F. Wang, B. Gou, Q. Zhang, Y. Qin, and B. Li, “Evaluation of ground settlement in response to shield penetration using numerical and statistical methods: a metro tunnel construction case,” Structure and Infrastructure Engineering, vol. 12, no. 9, pp. 1024–1037, 2016. View at: Publisher Site  Google Scholar
 W. Ke, Z. Wen, W. Haotian, W. Yajun, and L. Jinlong, “Study of impact of metro station sidecrossing on adjecent existing underground structure,” Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology, vol. 31, no. 4, pp. 2291–2298, 2016. View at: Publisher Site  Google Scholar
 J. Guan, The Safety Risk Assessment of Tunnel Construction during Crossing The Metro and ZHE study of ZHE Core Congtrol Technic, Xi’an University of Architecture and Technology, 2008.
 W. Yang, Research on Risk Management of Tunnel Construction Crossing the Existing Subway, Beijing Jiaotong University, 2011.
 X. Li and D. Yuan, “Development of the safety control framework for shield tunneling in close proximity to the operational subway tunnels: case studies in mainland China,” SpringerPlus, vol. 5, no. 1, p. 527, 2016. View at: Google Scholar
 MOHURD China, “Technical code for protection structures of urban rail transit,” Tech. Rep., China Architecture & Building Press, 2013. View at: Google Scholar
 MOHURD China, “Technical code for temporary support structures in construction,” Tech. Rep., China Architecture & Building Press, Beijing, China, 2013. View at: Google Scholar
 MOHURD China, “Code for hydropower engineering geological investigation,” Tech. Rep., China Planning Press, Beijing, China, 2013. View at: Google Scholar
 MOHURD China, “Technical code for groundwater lowering engineering in building and municipal,” Tech. Rep., China Architecture & Building Press, Beijing, China, 2013. View at: Google Scholar
 MOHURD China, “Technical specification for safety construction of deep building foundation pits,” Tech. Rep., China Architecture & Building Press, Beijing, China, 2013. View at: Google Scholar
 MOHURD China, “Code for green construction of building,” Tech. Rep., China Architecture & Building Press, Beijing, China, 2014. View at: Google Scholar
 MOHURD China, “Code for green construction of building,” Tech. Rep., China Planning Press, Beijing, China, 2010. View at: Google Scholar
 MOHURD China, “Assessment standard for green building,” Tech. Rep., China Architecture & Building Press, Beijing, China, 2014. View at: Google Scholar
 M. Chen, H. Pan, and J. He, “The application of fuzzy entropy in metro tunnel water leakage,” Chinese Journal of Underground Space and Engineering, vol. 12, no. A1, pp. 62–65, 2016. View at: Google Scholar
 B. Qi, Y. Zhu, and W. Fan, “Risk factor identification method of the whole life cycle in prefabricated construction,” Journal of Shenyang Jianzhu University (Social Science), vol. 18, no. 3, pp. 257–261, 2016. View at: Google Scholar
 H. Jingyu, “Study of fuzzy comprehensive evaluation method for metro operation safety based on entropy weight method,” Tunnel Construction, vol. 36, no. 12, pp. 1465–1470, 2016. View at: Google Scholar
 A. Nicolae and A. Nedelcu, “Security risk analysis and management,” Matec Web of Conferences, vol. 178, p. 8015, 2018. View at: Google Scholar
 L. Zhenhua, B. Bo, and H. Li, “Research on environmental impact evaluation model for shale gas development in China,” Modern Chemical Industry, vol. 37, no. 11, pp. 10–16, 2017. View at: Google Scholar
 Z. Luo, J. He, H. Pan, and Y. Yang, “Research on the selection strategy of green building parts supplier based on the catastrophe theory and kent index method,” Advances in Materials Science and Engineering, vol. 2016, Article ID 4673526, 12 pages, 2016. View at: Publisher Site  Google Scholar
 C. Wen and Y. Chunyan, “Basic theory and methodological system of extension,” Chinese Science Bulletin, vol. 58, no. 13, pp. 1190–1199, 2013. View at: Google Scholar
 H. Xuepeng and X. Wei, “Method of information security risk assessment based on improved fuzzy theory of evidence,” International Journal of Online Engineering, vol. 14, no. 3, pp. 188–196, 2018. View at: Publisher Site  Google Scholar
 L. Zhigao, M. Zhan, and J.P. Xu, “Dynamic comprehensive evaluation and speed characteristics analysis of a matterelement method,” Chinese Journal of Management, no. 11, pp. 119–127, 2015. View at: Google Scholar
 Z. H. Luo, Y. F. Yang, H. Z. Pan, and M. F. Zhong, “Research on performance evaluation system of shale gas PPP project based on matter element analysis,” Mathematical Problems in Engineering, vol. 2018, Article ID 4657383, 18 pages, 2018. View at: Publisher Site  Google Scholar
Copyright
Copyright © 2019 Zhenhua Luo et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.