Festus

A Full-time Risk Assessment Method for Lifting Operation under Man-machine-environment Condition Zhang Qian1,2,3, Wang Fengqi1,2, Liu Shuo4, Fan Qinglong5, Wu Shaoru5 (1Hubei Provincial Key Laboratory of Hydropower Engineering Construction and Management, China Three Gorges University, Yichang 443002, Hubei, China; 2School of Hydraulic and Environmental Engineering, China Three Gorges University, Yichang 443002, Hubei, China; 3Guangxi Energy Co., Ltd., Hezhou 542800, Guangxi, China; 4Hubei Energy Group Nanzhang Zhangjiaping Pumped Storage Power Co., Ltd., Nanzhang 441500, Hubei, China; 5Sichuan Huaneng Luding Hydropower Co., Ltd., Luding 626100, Sichuan, China) Funded by: National Natural Science Foundation of China (Grant Nos- and-). Abstract To prevent and reduce the occurrence of lifting accidents, a full-time evaluation method for working condition risks in lifting operations is proposed based on the human-machine-environment system perspective. Firstly, granulation analysis technology is applied to decompose the safety risk factors in lifting operation conditions. Secondly, the coupling effect of safety risk factors in lifting operation conditions is analyzed, and the causal relationship of human-machine-environment safety risk coupling in lifting operations is established to clarify the coupling effect of safety risks. Then, a full-time evaluation model for human-machine-environment safety risks in lifting operation conditions is constructed. Finally, by analyzing 107 lifting accident reports, the full-time values of multi-condition safety risks are calculated, and the results show that the lifting condition is the most dangerous, with human unsafe behaviors having the greatest impact. The research results can provide theoretical support for the prevention and control of safety risks in lifting operations. PRACTITIONER SUMMARY This study provides a comprehensive risk assessment method for lifting working conditions from the systematic perspective of human behavior factors, equipment factors and environmental factors.The assessment shows that the lifting condition is the most dangerous, with human unsafe behaviors having the greatest impact. [Keywords] Lifting Operations; Working Condition Risk; Coupling Effect; Human-Machine-Environment System; Full-time Evaluation. 0 Preface Lifting operations are indispensable means in the construction of large-scale water conservancy and hydropower projects, long-span bridges, high-rise buildings, and other engineering projects. Characterized by multi-person coordination, complex mechanical operation interfaces, and intricate spatial environments, the human-machine-environment (HME) system faces significant safety risks, making lifting accidents highly prone to occur [1,2]. As a periodically intermittent activity, the lifting operation process consists of various working conditions such as lifting, hoisting, turning, and unhooking, which form a full-time relationship spanning the past, present, and future among adjacent conditions. Once a risk catastrophe occurs in a specific working condition, it disrupts the full-time balance of the lifting operation's working conditions, breaking the equilibrium state of the HME system [3]. Therefore, developing a full-time evaluation method for the HME system in lifting operations is crucial for tracing the causes of lifting accidents and formulating preventive measures. Research on lifting accident injuries, both domestically and internationally, primarily focuses on accident causation analysis, human factor safety theories, and safety risk assessment [4]. Shapira et al. [5] conducted a causation analysis of lifting accidents using expert interviews and concluded that 21 safety risk factors, including project conditions, environment, human factors, and safety management, affect the safety of lifting operations. Sun Shimei [6] utilized the "2-4" model of accident causation [7,8] to analyze unsafe human behaviors in lifting accidents, revealing the patterns of unsafe behaviors among frontline operators. Wang Yan [9] improved the traditional safety risk matrix by introducing safety risk probability and impact weights, providing valuable insights for safety risk assessment. Based on the Cost of Safety (COS) model and Analytic Hierarchy Process (AHP), Aminbakhsh et al. [10] established an evaluation system that prioritizes safety risks in engineering projects. Huang et al. [11] employed the accident domino effect theory and 4M theory to construct a safety level system structure. Lin Dayong et al. [12] introduced triangular fuzzy number theory to establish a safety risk assessment structural model based on AHP. Jin Kuiguang [13] studied the "human-machine-environment-management" safety assessment indicators and constructed a safety risk assessment model based on the cloud model. These studies indicate that the causes of lifting accident injuries have gradually developed and diversified, encompassing an increasing number of factors. However, previous research has only considered the static characteristics of human factors in the HME system, neglecting the dynamic transfer of safety risks. Moreover, most studies have focused solely on evaluating the "present" state, without conducting a comprehensive assessment of the entire process of "past," "present," and "future" in the safety risk system. In light of this, aiming at the ergonomic characteristics of lifting operations, this study identifies safety risk factors in the HME subsystems within lifting working conditions, establishes a scientific and reasonable safety risk assessment method for lifting operations, explores the full-time comprehensive level of safety risks in lifting working conditions, and constructs a full-time analysis model for the HME system spanning the past, present, and future. This research provides theoretical support for safety risk management and control in lifting operations. 1 Granular Analysis of Human-Machine-Environment Safety Risk Factors Adopting a granular approach to analyzing the safety risk factors of the human-machine-environment system in lifting operations involves breaking down the temporal characteristics of lifting operation conditions. This approach allows for a detailed analysis of the elements within the human-machine-environment system and a granular decomposition of the associated safety risk factors[14]. Lifting operation conditions are defined according to the operational sequence, with the process cyclically progressing through lifting condition→transporting condition→turning condition→unhooking condition. As lifting operations proceed, various operations are interconnected and influence each other, forming past, present, and future relationships between adjacent conditions. The specific temporal segmentation of these conditions is illustrated in Figure 1. Fig. 1 Temporal division of lifting operation conditions The safety risk factors in lifting operations are organized from top to bottom. The first level is the coarse-grained layer, which corresponds to the macro safety risk factors of the lifting operation safety risk system. Following the principle of subsystem independence and based on human-machine safety engineering theory, the coarse-grained safety risk factors in lifting operations are divided into human factors, machine factors, and environmental factors. The second level is the fine-grained layer, which provides a further breakdown of the safety risk subsystems in lifting operations[15]. By compiling 132 crane injury accident reports from the past decade and integrating all reports into a single TXT file, we utilized KH Coder software for text mining analysis and word frequency statistics. The distribution of word frequencies was found to be consistent with the long tail theory, indicating that a limited number of high-frequency words can effectively reflect the key information of lifting operation safety risk factors, enabling their granular analysis. Based on the coarse-grained analysis of lifting operation safety risks, we categorized the fine-grained word frequencies into three aspects: unsafe human behaviors, unsafe machine states, and unsafe environmental factors. The specific classification is shown in Table 1. Using the high-frequency words obtained from the text analysis, we conducted a fine-grained decomposition according to the three coarse-grained subsystems: unsafe human behaviors, unsafe machine states, and unsafe environmental factors. The organized and fine-grained safety risk factors are presented in Table 2. Tab. 1 Classification of high-frequency words related to safety risks in lifting operation Dimension High-Frequency Words (Frequency) Vocabulary Proportion (%) Unsafe Human Behaviors Overload (30), Violation (112), Non-compliance (66), Training and Examination (103), No Certificate (61), Command (99), Safety Awareness (174), Physical Quality (9), Personality (27), Operation (84), Hidden Danger Investigation (120), Safety Protection Device (39), Technical Disclosure (57), Positioning (103), Inspection (74) 65% Unsafe Mechanical Conditions Wire Rope (150), Hook (66), Maintenance (76), Hoisting Mechanism (25), Corrosion (54), Substandard Quality (135) 28% Unsafe Environmental Factors Working Environment (37), Weather (29), Foundation (10), Lighting (10), Safety Warning Signs (28) 7% Tab. 2 Risk factors for man-machine-environment safety under lifting conditions Dimension Classification No. Safety Risk Factors Unsafe Human Behaviors (H) Working Without Certification Poor Mental State Poor Physical Fitness Lack of Education and Training Inadequate Operational Skills Command Violation Non-compliant Operation Failure to Inspect and Rectify Hazards Failure to Wear Safety Protection Equipment Lack of Communication Weak Safety Awareness Improper Positioning Absence of Professional Personnel Unsafe Mechanical Conditions (M) Non-compliant Lifting Equipment Insufficient Component Strength Safety Device Failure Operating System Malfunction Lack of Maintenance Non-compliant Parts Design and Installation Defects Unsafe Environmental Factors (E) Working in Adverse Weather Conditions Unstable Working Foundation Poor Working Visibility Narrow and Complex Working Space No Warning Signs in the Working Area 2 Construction of a Full-Time State Safety Risk Evaluation Model 2.1 Coupling Effects of Human-Machine-Environment Safety Risk Factors The coupling effect of safety risks is a measure of the interactions and influences among various safety risk factors within the safety risk system[16,17]. From the perspective of the lifting operation safety management system, coupling is defined as the interaction between subsystems of unsafe human behaviors, unsafe mechanical states, and unsafe environmental factors, leading to crane safety risks with a combined effect of “1+1>2”. By borrowing from the capacity coupling model in physics, a more effective evaluation of the coupling degree between the human-machine-environment subsystems can be achieved. By defining a coupling function using an effect function, let Xij (1≤ i ≤n,1≤ j ≤m) represent the value of the j-th safety risk factor in the i-th safety risk dimension. Ai and Bij are the upper and lower limits in the lifting operation safety risk coupling system, where Aij=max(Xij) and Bij=min(Xij). The effect function can represent the impact of dynamic changes in the human-machine-environment safety risk subsystems on the lifting operation safety system. The effect function for each safety risk indicator in the lifting operation safety risk indicator system on the overall safety risk system is as follows: (1) In the formula, Uij represents the contribution size of Xij to the safety risk system, and is referred to as the effect coefficient. When Uij →0, it indicates that Xij’s contribution to the safety risk system is almost negligible; when Uij →1, it indicates that Xij’s contribution to the safety risk system is maximal. Therefore, 0 ≤ Uij ≤ 1. Since the coupling of human-machine-environment safety risks in lifting operations increases the likelihood of accidents, this safety risk coupling is considered an amplifying coupling. Hence, only the positive effects of safety risk coupling need to be considered. 2.2 Human-Machine-Environment Safety Risk Evaluation Value 2.2.1 Temporal Transfer of Operational Conditions In the continuous process of lifting operations, if the transporting condition is viewed as the present state, then the lifting condition represents the past state and the turning condition represents the future state. The temporal transfer process of conditions is lifting → transporting → turning. If the unhooking condition is considered the present state, then the turning and lifting conditions represent the past and future states respectively, with the temporal condition sequence being: turning → unhooking → lifting. This process continues in a similar manner. The temporal transfer probability of lifting operation conditions is represented by matrix P, where Pu,v denotes the one-step transfer probability of safety risk from ru→rv. If xu→xv is unreachable, then Pu,v=0. The condition temporal transfer matrix P satisfies: (2) The specific details of the temporal transfer of lifting operation conditions are illustrated in Figure 2. Fig. 2 Temporal transfer diagram of operational safety risk According to Figure 2, the past state of a certain lifting operation condition is influenced by the first-order and second-order past states, and this condition will only impact the first-order future state. Thus, the probability matrix for the transfer of safety risk from the present to the past state, Pp(k), and the probability matrix for the transfer from the present to the future state, Pf (k), can be derived as follows: （3） （4） In the formula, P11, P22, P33 and P44 represent the present state transfer probabilities of the human-machine-environment system for the four types of conditions.P12, P23, P34 and P41 represent the transfer probabilities between different temporal states of adjacent conditions. 2.2.2 Temporal Evaluation Values To dynamically, objectively, and scientifically characterize the safety risks of lifting operation conditions, we use the past-present-future axis. Let the lifting operation conditions be denoted as ci (i=1,2,3,4), and the human-machine-environment safety risk indicators be denoted as rj (j=1,2,…,25). The present state evaluation values rj for the human-machine-environment safety risk indicators xij(T) in lifting operations are assigned according to the magnitude of their impact. Based on the temporal transfer relationships between conditions, the past state evaluation values of lifting operation safety risk indicators can be represented using the present state values and the present-to-past state transfer probability matrix, as shown in the following formula: xij (t)= xij (T)×Pp(k) (5) In the formula, xij(t) represents the past state evaluation value of the human-machine-environment safety risk. Similarly, the future state evaluation values are given by the following formula: ij (t)= xij (T)×Pf (k) (6) In the formula, ij represents the future state evaluation value. 2.3 Temporal Values of Human-Machine-Environment Safety Risks 2.3.1 Past State Values To analyze the past state of lifting operation safety risks, a linear evaluation model is used to characterize the human-machine-environment safety risks for lifting operation condition cic_ici​ at a past time ttt. The characterization function is given by: (7) In the formula, yi (t) represents the safety risk characterization value of lifting operation condition ci at past time t; represents the weight at past time t. For any t, the following conditions hold: Wj (t)≥0, ∑Wj (t)=1. The past time can be considered as a time period, denoted by the interval [t0 , t0 + T-1]. The overall safety risk characterization for lifting operation condition ci during the past time period is given by: (8) In the formula, t∈[t0 , t0 + T-1], and T is a known positive integer, and representing the weight of the past state of human-machine-environment safety risk. 2.3.2 Present State Values The present state values characterize the current status of lifting operation condition ci at time T. At the present time t = T, the overall safety risk characterization for lifting operation condition ci is given by: (9) In the formula, represents the weight of the present state of human-machine-environment safety risk for lifting operation conditions at time T. 2.3.3 Future State Values The future state values characterize the future development trend of lifting operation condition ci. Let N be an appropriate known constant. Within the future time interval [T,T+N], the overall safety risk characterization for lifting operation condition ci during the future time period is given by: (10) In the formula, represents the weight of the future state of human-machine-environment safety risk, and represents the future state evaluation values of the indicators. 2.4 Full-Time State Evaluation of Human-Machine-Environment Safety Risks 2.4.1 Full-Time State Safety Risk Coupling Effect Values Under the coupling effect, the occurrence of safety risks in different temporal states can impact other safety risks, ultimately resulting in an overall safety risk that is either greater or smaller than the original safety risk. This means that the original safety risk in different temporal states increases or decreases due to the positive or negative effects of coupling. Given that the coupling of human-machine-environment safety risks in lifting operations increases the likelihood of accidents, this safety risk coupling is considered an amplifying coupling. Therefore, only the positive effects of safety risk coupling are considered. Under the positive effect of coupling, safety risks exhibit an expanding trend. Thus, using the principle of superposition to calculate the temporal coupling value of safety risks, the specific characterization of the temporal coupling effect values of safety risks is as follows: (11) (12) (13) In the formula, Yi(1), Yi(2) and Yi(3) represent the safety risk temporal coupling effect values of lifting operation condition ci in the past, present, and future states, respectively. 2.4.2 Aggregation of Full-Time State Safety Risks During lifting operations, safety risks can be categorized into three attributes: human, machine, and environment, and they also exhibit three temporal states: past, present, and future. The safety risk of a particular lifting operation condition depends not only on the previous condition (past) and the current condition (present), but it also affects the subsequent condition (future). The human-machine-environment safety risks transfer and interact among different conditions. To comprehensively measure the safety risk level of lifting operation conditions, the past, present, and future temporal states of the lifting operation condition cic_ici​ are aggregated. Considering the transfer probabilities of human-machine-environment safety risks between conditions, a past-present-future full-time state analysis model for the human-machine-environment system in lifting operations is established. To mitigate local advantages and embody the “barrel principle”, a nonlinear evaluation method is used to evaluate the full-time state safety risks of lifting operation condition cic_ici​. The specific aggregation function is as follows: (14) In the formula, λm represents the predetermined weight, which is the aggregation weight. 3 Case Analysis 3.1 Case Data Safety management and crane engineering websites were used to collect crane injury accident reports. A total of 107 lifting operation accidents occurring over a six-year period from 2017 to 2022 were selected for lifting operation safety risk analysis. The occurrences of human-machine-environment safety risks in lifting conditions were counted. If an accident in the lifting condition was caused by a safety risk arising from the lifting condition, the transfer frequency of the lifting condition safety risk was increased by one. If it caused an accident in the transporting condition, the transfer frequency of the transporting condition safety risk was increased by one. The transfer probability was obtained by dividing the transfer frequency by the total number of human-machine-environment safety risks in the lifting condition. The temporal transfer of conditions is shown in Figure 3. Fig. 3 Temporal transition diagram of working conditions Qualitative indicator data were obtained using the expert decision scoring method. A panel of 31 experts was formed, with the basic situation of the expert decision group shown in Figure 4. A 0-4 scale was used to indicate the degree of influence between the 25 factors: 0 represents no influence, 1 represents a small influence, 2 represents an influence, 3 represents a significant influence, and 4 represents a very significant influence. After collecting the basic data, the next steps of calculating entropy weights, combined weights, and safety risk evaluation were performed according to the formula. The results are shown in Table 3. Fig. 4 Basic situation of expert decision-making group Tab. 3 Current tense evaluation value of lifting operation conditions Safety Risk Indicators Lifting Conditions Hoisting Conditions Turning Conditions Unlocking Conditions - - - ... - Full-Time State Safety Risk Evaluation Values Because the safety risk of lifting operation conditions can impact the same condition or the immediately subsequent condition, the past state and future state evaluation values are calculated based on the present state and formulas (5) and (6). The detailed values are shown in Table 4. Tab. 4 Comprehensive-temporal evaluation value of lifting operation conditions Safety Risk Indicators Lifting Conditions Hoisting Conditions Turning Conditions Unlocking Conditions Past Present Future Past Present Future Past Present Future Past Present Future - - - ... - Calculation of Safety Risk Coupling Effect The average value of the impact and the influenced degrees of each safety risk indicator, as scored by the 31 experts, is used to determine the safety risk factor values for each indicator, denoted as Xij. The safety risk impact evaluation range is [0,4][0,4][0,4], thus Aij=4 and Bij=0. The safety risk coupling effect coefficient is calculated according to formula (1). Using the previously obtained combined weights and the safety risk coupling effect coefficients, the full-time state coupling effect values for the three subsystems (human, machine, environment) of the 25 safety risk indicators are calculated according to formulas (11), (12), and (13). The specific results are shown in Table 5. Tab. 5 Temporal effect value of safety risk coupling in lifting operation conditions Safety Risk Effect Coefficient Lifting Conditions Hoisting Conditions Turning Conditions Unlocking Conditions Past Present Future Past Present Future Past Present Future Past Present Future - - - ... - 3.4 Calculation of Aggregated Full-Time State Safety Risk Values Using the safety risk formulas for various temporal states and considering the transfer of safety risks between conditions, the comprehensive safety risk for the four lifting operation conditions across the past, present, and future states is calculated and ranked. Following the principle of focusing on the present while considering both the past and future, weights are assigned as follows: λ1=0.3, λ2=0.5 and λ3=0.2. A comprehensive analysis is conducted for the four lifting operation conditions. The aggregated full-time state safety risk evaluation and ranking are shown in Table 6. The full-time state safety risk levels for different lifting operation conditions are illustrated in Figure 5. Tab. 6 Comprehensive-temporal safety risk ranking of working conditions Working Conditions Past Safety Risk Present Safety Risk Future Safety Risk Full-time State Safety Risk Ranking Ranking Ranking Ranking Lifting - Hoisting - Turning - Unlocking - Fig. 5 The magnitude of safety risks in comprehensive-temporal of lifting operation conditions Based on the comprehensive analysis and comparison illustrated in Figure 5, the safety risk factors within the human-machine-environment subsystems that significantly impact lifting operations across various temporal states are as follows: Unsafe Human Behaviors: violating operation procedures r7, lack of communication r10, weak safety awareness r11, unauthorized command r6, lack of education and training r4, failure to inspect and rectify hazards r8. Unsafe Machine Conditions: lack of maintenance r18, unqualified lifting gear r14, insufficient component strength r15, failure of safety devices r16. Unsafe Environmental Factors: poor visibility in the work area r23, no warning signs in the work area r25. From a management perspective, a comprehensive analysis of the human-machine-environment subsystems reveals that human factors, as the main component of lifting operations, have a significant impact on the safe and smooth execution of operations. Among these, violating operation procedures and lack of communication are the most significant safety risk factors. The safety risks associated with mechanical equipment are the next most impactful on lifting operations. Environmental factors have the least noticeable impact on lifting operations within the overall human-machine-environment system. In summary, human factors represent the greatest safety risk in lifting operations. Efforts should be focused on improving the professional skills and operational proficiency of lifting operators. Additionally, the impact of mechanical equipment on lifting operations is also substantial. 4 Conclusion (1) This research analyzes the human-machine-environment (HME) elements involved in lifting operation conditions. Based on the theory of human-machine safety engineering, a coarse-grained HME safety risk subsystem for lifting operations is derived, and by further refining this subsystem, 25 safety risk factors are identified. (2) A full-time evaluation model for safety risks in lifting operation conditions is constructed. By analyzing the temporal transfer process of safety risks in lifting conditions, the past, present, and future temporal values of safety risks in lifting operation conditions are determined. The entropy weight method is adopted to derive the safety risk weight coefficients, the full-time risk coupling effect value is measured, and a full-time analysis model spanning the past, present, and future of the HME system in lifting operations is established. (3) Empirical cases of lifting accidents are analyzed. A total of 107 lifting injury accident cases from 2017 to 2022 are selected, and the direct and indirect causes of these accidents are examined. The analysis reveals that the lifting condition is the most hazardous among the four lifting operation conditions, and human unsafe behaviors have the greatest impact on lifting operations, with the most significant coupling effect among human unsafe behavior risk factors. In the subsequent stages, the focus will be on digital twin technology for full-time condition risk analysis in lifting operations, establishing an experimental research paradigm that combines ergonomics and virtual simulation to achieve forward reasoning analysis for safety risk management and control. Ethics statement An ethical statement is not applicable to this work. Disclosure statement No potential conflict of interest was reported by the author(s). References [1] Ali M, Mohamad M I. Crane failure and accident in construction[J]. Faculty of Civil Engineering, University Teknology Malaysia, Malaysia, 2016:136-149. [2] Han Y, Yin Z Z, Liu J L, et al. Characteristics and Causes of Hazard Cognitive Bias of Construction Workers[J]. Journal of Civil Engineering and Management, 2019,36(05):56-61. [3] Xu X J, Wang L Q, Fang X M, et al. Interactive visual simulation of offshore hoisting operationand numerical modelling[J]. Journal of Harbin Institute of Technology, 2018,50(01):160-168. [4] Song X, Xie Z. Application of man-machine-environment system engineering in coal mines safety management[J]. Procedia Engineering, 2014,84:87-92. [5] Shapira A, Lyachin B. Identification and analysis of factors affecting safety on construction sites with tower cranes[J]. Journal of Construction Engineering and Management, 2009,135(1):24-33. [6] Sun S M. Research on Unsafe Action Law of Front-line Operators inBuilding Construction Crane[J]. Safety ＆ Security, 2020,41(05):7-13. [7] Fu G, Cao J, Zhou L, et al. Comparative study of HFACS and the 24Model accident causation models[J]. Petroleum Science, 2017,14(3):570-578. [8] Fu G, Xie X, Jia Q, et al. The development history of accident causation models in the past 100 years: 24Model, a more modern accident causation model[J]. Process Safety and Environmental Protection, 2020,134:47-82. [9] Jin L H, Wang L C, Chen Y, et al. Response characteristics of lifting operation driven byvisual information[J]. China Safety Science Journal, 2022,32(04):198-204. [10] Jin Lianghai, Li Jiaxin, Zheng Xiazhong, et al. Trajectory planning model for lifting operation considering spatial conflict[J]. China Safety Science Journal, 2020,30(03): 47-52. [11] Wang Yan. Establishment of a risk assessment system on social security for CNPC’s international businesses based on improved risk matrix[J]. Journal of Safety Science and Technology, 2019,15(S2):5-10. [12] Aminbakhsh S, Gunduz M, Sonmez R. Safety risk assessment using analytic hierarchy process (AHP) during planning and budgeting of construction projects[J]. Journal of safety research, 2013,46:99-105. [13] Huang G, Sun S, Zhang D. Safety evaluation of construction based on the improved AHP-grey model[J]. Wireless Personal Communications, 2018,103(1):209-219. [14] Lin Dayong, Lei Mingfeng, Cao Haorong, et al. Study on construction risk of shield tunnel under-passing existing railway by fuzzy comprehensive evaluation[J]. Journal of Railway Science and Engineering, 2018,15(05):-. [15] Zhang Gan, Guo Shengyu, Zhou Xiaojie, et al. Research and application of intelligent device for collaborative perception of safety risk in lifting operations[J]. China Safety Science Journal, 2023,33(06):122-127. [16] Jin Kuiguang. Safety assessment of container ship navigating in heavy seabased on cloud model[J]. Journal of Shanghai Maritime University 2019,40(01):32-36. [17] Cheng Yi, Miao Zhuoqian, Feng Qinrong. Granular Computation Based on Fuzzy Rough Sets[J]. Computer science 2007(07):142-145. [18] Song Dongming, Zhu Yaoqin, Wu Huizhong. Study on Problem Decomposition of Problem Solving Process in Hall for Workshop of Metasynthetic Engineering[J]. Journal of System Simulation, 2008(18):-. [19] Paltrinieri N, Khan F, Cozzani V. Coupling of advanced techniques for dynamic risk management[J]. Journal of Risk Research, 2015, 18(7):910-930. [20] Zhang J, Xu K, You G, et al. Causation analysis of risk coupling of gas explosion accident in Chinese underground coal mines[J]. Risk analysis, 2019, 39(7):-.