Stability Assessment of Deep Three-Soft Outburst Coal SeamRoof Based on Fuzzy Analytic Hierarchy Process

-e stability of deep “three-soft” coal seam roof has always been a key issue in coal mining. -ere are a lot of factors affecting the stability of deep three-soft coal seam outburst roof. However, there is currently no definite method able to draw an accurate assessment conclusion on roof stability. In order to accurately determine the main influencing factors of the stability of deep threesoft coal seam outburst roof and reduce the loss of coal production, this paper performed three-soft coal seam risk identification on Lugou Mine based on the introduction of the fuzzy analytic hierarchy process theory. 23 main risk factors were identified. -en, it established a hierarchical structure model of coal seam roof stability in accordance with experts’ opinions. -e analytic hierarchy process was used to calculate the weights of indicators at all levels. Next, the paper used the fuzzy comprehensive evaluation method and expert scoring to evaluate various risk factors in the indicator system, as well as the overall safety level. -e results showed that the deep three-soft coal seam stability of Lugou Mine ranks the third hazard level. -e main risk and harmful factors include safety awareness, safety monitoring system, roof weakness, ventilation system, fire-fighting system, and rock bolt quality. In response to the evaluation results, this paper formulated corresponding control measure in terms of ventilation risk, safety monitoring risks, construction personnel risks, and fire protection risk to reduce losses in the mining process, providing a new evaluation method for the stability assessment of deep outburst coal seam roof.


Introduction
China is a country with a huge coal resource storage. e total proven coal reserves are 5.57 trillion tons, ranking first in the world. In recent years, with the rapid development of the social economy, the demand for energy has increased in China [1]. Coal has always been one of China's main energy sources. About 70% of China's energy comes from coal. Coal and gas outburst has always been a major disaster plaguing China's efficient and safe production of coal mines [2][3][4]. Coal and gas outburst accidents occur every year in China. e causes are diverse. Due to these characteristics, "three-soft" coal seams are prone to coal and gas outburst accidents [5]. " ree-soft" coal seams refer to soft roof and floor stratum, soft main coal seams, and soft seam floor stratum encountered in coal mining. In general, the coal seam and roof and floor with three soft features are weak strata, the coal seam fracture is developed, and the structure is complex [6,7]. e strength of "three-soft" coal seams is very low. For this reason, the thickness of coal seams during mining operation varies greatly. e coal body is mostly scaly or powdery, resulting in unstable gas conditions. It is prone to gas outburst accidents, greatly affecting normal tunneling work [8][9][10].
At present, there is not much analysis of the stability of three-soft coal seam outburst roof. Coal seams are just evaluated by a single qualitative or quantitative evaluation method.
e analysis results often have large errors. e fuzzy analytic hierarchy process combines qualitative and quantitative methods, making evaluation results more reasonable and scientific. Laarhoven and Pedrycz [11] first proposed a comprehensive evaluation method combining the fuzzy mathematics theory and the analytic hierarchy process. Chen and Wu [12] applied the fuzzy analytic hierarchy process to 3D printer performance evaluations and proposed the fuzzy collaborative intelligent analytic hierarchy process. Wang et al. [13] used the fuzzy analytic hierarchy process to determine optimal stope structure parameters. In order to overcome the subjectivity and unity of traditional evaluation methods, Song et al. [14] put forward a new method of using the fuzzy analytic hierarchy process to select coal mining methods. Sridhar and Ganapuram [15] used the fuzzy analytic hierarchy process to analyze the factors affecting the maintenance function on a more sustainable manufacturing process. Peyman et al. [16] created an earthquake risk assessment (ERA) diagram for Sanandaj, Iran, through the fuzzy analysis analytic hierarchy process. Zhou et al. [17] conducted the fuzzy comprehensive evaluation on the stability of bedding slopes reinforced by prestressed anchor cables. e evaluation results obtained were consistent with the on-site monitoring results. Ince et al. [18] used the fuzzy analytic hierarchy process to determine metrics related to the sustainable performance in the construction industry. Mallick et al. [19,20] used FAHP and MCDM technologies, together with the geographic information technology, to perform a weighted overlay analysis of the groundwater potential area of the Itwad-Khamis watershed in Saudi Arabia. Rouyendegh et al. [21] used FAHP to construct a decision-making model to improve the operating performance of healthcare companies. Wijitkosum and Sriburi [22] used FAHP to analyze and assess the desertification risk in the upper reaches of the Bicebri River. Wang et al. [23] established a targeted groundwater environmental impact assessment indicator system specifically and determined the weight of each index, as well as the impact level through the fuzzy comprehensive analytic hierarchy process. Zhang et al. [24] applied FAHP to flood control and discharge to solve flood discharge conflicts between different areas within watershed when floods occur. Riibas et al. [25] used the fuzzy analytic hierarchy process (FAHP) to evaluate the project risk of a large hydropower project in the construction stage. Wu et al. used FAHP in the PBA construction risk analysis of subway stations [26]. Liang studied the highway cost risk assessment based on the fuzzy analytic hierarchy process [27]. Li optimized the mining method of the manganese mining area based on the fuzzy analytic hierarchy process [28].
On the basis of the above analysis, although scholars at home and abroad had applied the fuzzy analytic hierarchy process to many aspects, no one has analyzed the stability of deep three-soft coal seam roof in coal mines. Fuzzy mathematics can be used for qualitative analysis. e analytic hierarchy process can solve problems quantitatively. e fuzzy analytic hierarchy process combines the qualitative and quantitative analyses to make results more scientific and reasonable.

Fuzzy Analytic Hierarchy Process.
e fuzzy analytic hierarchy process is a systematic analysis method combining qualitative and quantitative analysis. It is a comprehensive application of the analytic hierarchy process and the fuzzy evaluation method [29]. e essence of the analytic hierarchy process is a way of decision-making thinking. First, a complex problem is considered as a system. e system is decomposed into multiple small aspects. en, each small aspect is decomposed to establish a hierarchical structure model, which is generally divided into three layers from high to low: target layer, criterion layer, and scheme layer. e relative importance of each two factors in each level will be compared to calculate the weight of each factor in the evaluation system [30]. e fuzzy comprehensive evaluation method aims to transform qualitative evaluation into quantitative evaluation. Its basic idea is to apply the fuzzy transformation principle and the maximum membership principle to consider the comprehensive evaluation of all factors related to the item to be evaluated. e focus of the evaluation is all relevant factors considered. e evaluation set of each factor is determined in accordance with the evaluation results of experts. e membership degree vector can be obtained by combining the weight set. Finally, the fuzzy comprehensive evaluation result can be obtained according to the evaluation standard [31].

Operation Process of Fuzzy Analytic Hierarchy Process.
For specific operations of the fuzzy analytic hierarchy process, different items have diverse operations. e risk assessment of the stability of deep three-soft coal seam outburst roof can be implemented in four stages [32], including model design, expert consultation, matrix establishment, calculation analysis, and analysis report formation, as shown in Figure 1.

Decision-Making Steps of Fuzzy Analytic Hierarchy Process
Step 1. Create a hierarchical structure diagram.
Establish an analytic hierarchy process diagram to identify the risks of the project.
Step 2. Calculate the weights of primary and secondary indicators.
Compare each factor in the criterion layer with each factor (scheme) in the scheme layer and score the importance with the number within the interval [1,9] to form a judgment matrix. After being tested by the consistency ratio C × R � λ max − n/(n − 1)R × I < 0.1, the weights w i ′ � (w 1 , w 2 , . . . , w i ) are normalized to get w i , which satisfies 0 ≤ w i ≤ 1, and k i�1 w i � 1. e values of RI in the consistency check list can be found in Table 1.
A matrix W with k rows and n columns can be obtained by combing several weight vectors:

Shock and Vibration
Step 3. Determine the fuzzy vector and comment set. Each factor in the indicator layer was scored by 10 experts.
e score vector R i � (R i1 , R i2 , . . . , R in i � 1, 2, . . . , k) of the i-th factor C i can be obtained, and the comment [33] is also determined.
Step 4. Perform the comprehensive risk assessment. e weight of each risk factor is W � W 1 , W 2 , W 3 , . . . W n }, and the evaluation matrix is B ij , where en, the comprehensive evaluation grade score of A is After the grade score corresponding to each factor of the criterion layer is calculated, the corresponding risk level can be obtained in accordance with the score interval in the    branch faults in the southern part of No. 32141 working face. e fault drop is relatively large. Based on the geological structure analysis above, the geological structure of this area is relatively complex. It is expected that there may be hidden structures in the working face. It is necessary to strengthen the detection of the geological structure during the excavation.

Risks and Harmful Factors
(1) Risks of Geological Environment. e coal seam in No. 32141 working face of Lugou Coal Mine has a simple structure, which is black scaly powder. e lump coal is hard with a metallic luster. Influenced by the oblique structure, the roof and floor of the coal seam have local undulating changes, belonging to unstable coal seams; the geological environment is complex, so there are inherent risks such as coal and rock gas content, geological structure, deep stress concentration, coal outburst characteristics, and gas emission.
(2) Risks of Safety Management. Since the Lugou Mine project is large and involves many people, making its management difficult, safety education, safety rules and regulations, and emergency drilling have become the main contents of management. (3) Risks of Safety Facilities. Due to the existence of gas, CO, and other gases, in order to prevent personnel poisoning and gas explosion hazards during the excavation process, the ventilation requirements of the coal mine must be quite high; the ventilation issue has become the primary risk for three-soft coal seams. In addition, monitoring systems, fire-fighting systems, transportation systems, and protective rescue facilities are also risk sources. (4) Risks of Construction Personnel. For construction personnel, the risks mainly come from the unreasonable arrangement of personnel, weak safety awareness, and weak ideological and political qualities.

Establishment of a Risk Assessment Indicator System.
From the identification of main risk factors in Section 3.2, it can be found that there are 23 levels of main risks and harmful factors affecting the project, which can be divided into three layers: target layer, primary indicator layer, and secondary indicator layer. e corresponding factor set is A � C1, C2, ..., C23, as shown in Figure 2.  Table 2. e square root method is used to solve the approximate value w i ′ of the weight vector of the evaluation factor:

Weight Calculation of Indicators
After the normalization of w i ′ , the evaluation factor weight vector w i can be obtained. e relationship is Get Python is used to obtain the value of λ and find λ max � 5.3930. From Table 1, it can be obtained that R · I � 1.12. After being substituted into the formula, the following can be obtained: It shows that the discriminant matrix satisfies the consistency requirement, that is, the obtained eigenvector is valid.

Weight Calculation of Secondary Indicators.
e analytic hierarchy process is still used to calculate the weight of indicators.
(1) First, establish the judgment matrix for B 1 geological environment risk. Similarly, the following can be obtained from the data calculation in Table 3: From   Shock and Vibration 5 It shows that the discriminant matrix satisfies the consistency requirement, that is, the obtained eigenvector is valid. (2) Establish a judgment matrix for B2 safety management. e following can be obtained from the data calculation in Table 4: From Table 1, it can be obtained that R * I � 0.89. After being substituted into formula (6), it can be obtained that C × R � 0.0077 < 0.1. It shows that the discriminant matrix satisfies the consistency requirement, that is, the obtained eigenvector is valid. (3) Establish a judgment matrix for B3 roof factors. e following can be obtained from the data calculation in Table 5: From Table 1, it can be obtained that R × I � 1.12. After being substituted into formula (6), it can be obtained that C × R � 0.0998 < 0.1. It shows that the discriminant matrix satisfies the consistency requirement, that is, the obtained eigenvector is valid. (4) Establish a judgment matrix for B4 safety facilities. e following can be obtained from the data calculation in Table 6: From Table 1, it can be obtained that R × I � 1.12. After being substituted into formula (6), it can be obtained that C × R � 0.0030 < 0.1. It shows that the discriminant matrix satisfies the consistency requirement, that is, the obtained eigenvector is valid. (5) Establish a judgment matrix for B5 construction personnel. e following can be obtained from the data calculation in Table 7: From Table 1, it can be obtained that R × I � 0.89. After being substituted into formula (6), it can be obtained that C × R � 0 < 0.1. It shows that the discriminant matrix satisfies the consistency requirement, that is, the obtained eigenvector is valid.
According to the total weight of the C layer calculated above, all risk factors are ranked based on the weight value, as shown in Table 8. e ranking above shows the evaluation results of risk factors of Lugou Mine's three-soft coal seam outburst roof stability. During the evaluation process, 10 experts were invited to score various risk factors in the Anping Gas Tunnel Construction Safety Project. Statistical methods were used to calculate the final weight value of different factors. Based on the weight value, it can be concluded that safety awareness, safety monitoring system, roof weakness, ideological and political qualities, fire-fighting system, and rock bolt quality dominate the whole project; among them, the riskiest factors are safety awareness and safety monitoring system. In other words, if gas emission cannot be effectively controlled and detected, the safety awareness of workers cannot be improved, and coal mines will have accidents such as poisoning and gas outbursts, which may eventually cause coal mines to stop production. Moreover, roof weakness is also a risk factor, which means that if the tunnel support measures are not done well, roof fall accidents may occur.

Establishment of Safety Rating.
After on-site survey by experts, the Anping Gas Tunnel single-factor score comment set is V � v 1 , v 2 , v 3 , v 4 , v 5 , corresponding to five grades, namely, "very good," "good," "average," "poor," and "very bad." e evaluation set is represented by (95, 85, 65, 45, 30), as shown in Table 9. Relationship between scores and safety grades is shown in Table 10.
Generally, basic methods such as expert survey method, CIM method, and Monte Carlo simulation method are used for risk assessment [34]. In order to simplify the calculation and obtain the corresponding data, this chapter used the expert survey method. e expert evaluation method was adopted. 10 experts were invited to make judgments on possible risks and their grades, and then they scored according to indicators. e statistics based on the judgment values given by the experts are shown in Table 11.
(1) Establish the fuzzy evaluation matrix R i .
Obtain the fuzzy evaluation matrix R i after being scored by experts.
(2) Solve the evaluation matrix B i of each factor. According to the formula, e evaluation matrix of each factor is as follows: (3) Obtain the total evaluation matrix R: (5) Grading: From the above calculations, it can be known from Table 10 that the safety grades of these four types of factors are as follows: the safety grade of the geological environment is "average," the safety grade of the safety management is "average," the safety grade of the roof itself is "average," the safety grade of the safety facilities is "average," and the safety grade of the construction personnel is "average." e total score of the system is f � 95 × 0.19 + 85 × 0.28 + 65 × 0.24 + 45 × 0.20 + 30 × 0.08 � 69, and the safety grade of Lugou Mine's three-soft coal seam outburst roof stability is "average." (6) Evaluation results: from the fuzzy analytic hierarchy process, it can be concluded that safety awareness, safety monitoring system, roof weakness, ventilation system, fire-fighting system, and rock bolt quality are dominant in the entire project. e safety grade of the entire system is Grade III.

Risk Control Measures
According to the risk analysis result, in the analysis of Lugou Mine's three-soft coal seam outburst roof stability, the following five major risks have the highest probability of occurrence, including ventilation risks, safety monitoring risks, roof own risks, safety management risks, and firefighting risks, which have become the key control objects for Lugou Mine's three-soft coal seam outburst roof stability. In order to reduce coal seam risks and ensure the smooth completion of the project, attention should be focused on risk prevention from the following aspects. Corresponding control measures have been formulated.
(1) Ventilation Risks. e gas in the coal mine will continuously overflow. When the gas accumulates to a certain concentration, it will burn and explode in the fire source. In order to ensure the safety of the coal mine, the ventilation issue is particularly crucial.
(i) During the extraction period, the air distribution on the working face should not be less than 1500 m 3 /min. e ventilation system is independent and the return air side is unblocked. e inlet and return air does not pass through the goaf or roof falling area. At the same time, the wind measurement should be strengthened and the personnel should be fixed. (ii) If the gas exceeds the limit, the power supply of all nonintrinsically safe electrical equipment in No. 32081 working face as well as the upper and lower suballeys can be disconnected; the poweroff is sensitive and reliable. A total of 2 feed sensors are installed on the load side of the controlled switch, which is maintained by a dedicated person every day. e sensors are portable for comparison. e calibration is performed every 15 days with a full-time gas inspector. (iii) Set up backup power supply and ventilator: when one power supply stops working or the ventilator fails, the backup power supply or the ventilator can be quickly used to ensure the normal ventilation in tunnels.
(2) Safety Monitoring Risks. e most important thing of safety monitoring is to control the gas limit and prevent gas accumulation. During the construction, when the gas concentration is detected to exceed the limit, the gas monitoring system will automatically cut off the power supply in the overlimit area, and the system can still work normally. At this time, measures should be taken to deal with the situation in accordance with the data provided by the system.
(i) It is strictly forbidden to allow people to enter the overlimit area. e variable air volume supply is used to control the air intake, and the gas in the overlimit area is discharged gradually. (ii) When the gas is discharged, it is necessary to ensure that the gas concentration in the working area does not exceed the limit. Moreover, it should be ensured that the gas discharged from the coal mine does not exceed the limit. e gas inspectors must frequently inspect the gas concentration in the return air flow. When the gas concentration is less than 0.75%, the air supply should be reduced. (iii) After the normal ventilation is restored, the electromechanical equipment in the power-off area shall be inspected first. e power supply can be manually restored for normal construction only after the equipment is confirmed to be in good condition.
(3) Construction Personnel Risks. e main construction personnel risks in this project refer to the personnel physical condition and professional competence, as well as the safety awareness.
(i) Conduct safety education and training regularly. Safety education training can improve people's safety production knowledge and enhance their safety awareness, which can effectively prevent from unsafe behaviors. (ii) Safety education training must be comprehensive with prominent focus and strong systemicity. (iii) Conduct physical examinations on employees regularly.
(i) After the fire report is received, the on-site team leader and technical person in charge should first cut off the power supply and actively organize personnel to extinguish the fire. When the on-site personnel are threatened by disasters, they should be organized to evacuate in time in accordance with the gas combustion accident plan. (ii) In the process of fire-fighting and disaster relief, the concentration, wind direction, and air volume of various toxic and harmful gases should be detected in time. Safety measures should be taken to prevent gas and coal dust explosion, as well as personnel poisoning timely when the situation is identified. (iii) e electrical equipment should be inspected frequently to avoid electrical sparks, thus ensuring the continuous normal operation of the mechanical equipment and avoiding friction sparks.

Conclusions
Based on the engineering background of stability assessments of three-soft coal seam roof outburst coal seams, this paper used the fuzzy analytic hierarchy process combining the analytic hierarchy process and the fuzzy mathematics to conduct risk analysis of Lugou Mine's No. 32141 working face. It built an analytic hierarchy model. e influencing factors of primary and secondary indicators were determined and ranked by scores. e conclusions are as follows: (1) e analytic hierarchy process is a method combining human subjective judgment and quantitative calculation. It has the characteristics of systematization and hierarchicalization. e quantitative information needed for the analysis is relatively less; the fuzzy mathematics evaluation method combines the fuzzy transformation principle and the maximum membership principle. e comprehensive evaluation of all factors related to the thing to be evaluated focuses on all relevant factors considered. e fuzzy analytic hierarchy process is a combination of the analytic hierarchy process and the fuzzy mathematics method. It has the advantages of qualitative analysis and quantitative evaluation, which is a new method of evaluating coal seam roof. It can obtain a high accuracy of three-soft coal seam roof stability assessments.
(2) Based on the characteristics of the coal seams of the Lugou Mine, combined with historical data and after discussions with experts, the risks in each layer are divided into five categories: geological environment risks, safety management risks, roof own risks, safety facility risks, and construction personnel risks. Each category contains several basic risk factors. e analytic hierarchy process can be used to solve the weight of each factor. It can be concluded that safety awareness, safety monitoring system, roof weakness, ventilation system, fire-fighting system, rock bolt quality, and other risk factors dominate the entire project.
(3) After the analytic hierarchy process was used to obtain the weight of each risk factor, based on the fuzzy mathematics analysis, it is concluded that the safety grade of Lugou Mine's three-soft coal seam outburst roof stability is Grade III: "average." e main risks and harmful factors include ventilation risks, safety monitoring risks, roof own risks, safety management risks, and firefighting risks. (4) According to the conclusion drawn from the fuzzy analytic hierarchy process, corresponding measures are proposed for factors with high risks, such as ventilation problems, safety monitoring systems, and safety management, to reduce the probability of accidents and ensure the safe production of coal mines. e fuzzy analytic hierarchy process can also be used to evaluate the stability of deep roof in other similar coal mines to reduce the probability of coal mine accidents.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.