Limited space natural gas pipeline leakage risk assessment method and system

Abstract

The invention discloses a leakage risk assessment method and system for a natural gas pipeline in a limited space, and relates to the technical field of natural gas safety. Various potential risk factors influencing leakage of the natural gas pipeline in the limited space are collected; performing expert opinion consultation and multi-round feedback iteration on the collected various potential risk factors by using a Delphi method, determining a first-level evaluation index and a key risk factor, and constructing a limited space natural gas pipeline leakage risk evaluation index system according to the first-level evaluation index and the key risk factor; and constructing a leakage risk assessment model of the natural gas pipeline in the limited space to research the natural gas pipeline in the limited space. According to the method, the hierarchical Bayesian analysis, the fault tree and the Bayesian network are comprehensively applied, complex dependency and uncertainty in data can be captured and processed, adjustment and optimization can be carried out according to different conditions such as limited space environments, pipeline types and leakage source characteristics, and evaluation requirements in different scenes are met.

Description

Technical Field

[0001] The present invention relates to the field of natural gas safety technology, and more particularly to a method and system for assessing the risk of natural gas pipeline leakage in a confined space. Background Art

[0002] At present, natural gas, as a clean energy, occupies an important position in energy supply. However, the risk of natural gas pipeline leakage in confined spaces (such as underground pipe galleries, tunnels, inside buildings, etc.) is relatively high. Once a leak occurs, it may not only cause serious accidents such as fire and explosion, but also pose a huge threat to human life safety and the environment.

[0003] Existing natural gas leakage risk assessment methods mostly focus on the analysis of a single factor, such as leakage source, leakage rate or meteorological conditions, and lack a systematic assessment method that comprehensively considers multiple factors. In addition, there are few assessment methods for specific environmental conditions in confined spaces, which makes it difficult to meet actual engineering needs.

[0004] Therefore, how to accurately assess the risk of natural gas pipeline leakage in confined spaces and take effective measures is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0005] In view of this, the present invention provides a method and system for assessing the risk of natural gas pipeline leakage in a confined space to solve the problems existing in the background technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A method for assessing the risk of leakage of a natural gas pipeline in a confined space, comprising:

[0008] Collect various potential risk factors affecting natural gas pipeline leakage in confined spaces;

[0009] The Delphi method was used to collect expert opinions and conduct multiple rounds of feedback iterations on various potential risk factors collected, determine the primary evaluation indicators and key risk factors, and build a confined space natural gas pipeline leakage risk evaluation indicator system based on the primary evaluation indicators and key risk factors;

[0010] Vague set and entropy method are used to calculate and determine the weight values ​​of risk assessment indicators respectively, and then the weight values ​​of each risk assessment indicator are processed to obtain the combined weight of the first-level assessment indicators combining subjective and objective factors;

[0011] Based on hierarchical Bayesian analysis and Bayesian network, a confined space natural gas pipeline leakage risk assessment model is constructed;

[0012] The accident precursor data and the combined weights of the first-level evaluation indicators combining subjective and objective factors are input into the confined space natural gas pipeline leakage risk assessment model to conduct a leakage risk assessment on the evolution process of the causes and consequences of natural gas pipeline leakage in the confined space.

[0013] Optionally, the confined space natural gas pipeline leakage risk assessment index system includes four primary assessment indicators: space structure, ventilation conditions, microclimate environment and leakage source.

[0014] Optionally, it also includes using decision-making experiments and evaluation laboratory methods to identify and analyze the interaction between various potential risk factors, conduct risk causal factor correlation analysis for factors with relevant dependencies and influence relationships, deeply explore the internal connections and mutual influence mechanisms between various potential risk factors, conduct coupling analysis around independent intermediate factors, and conduct multi-level and multi-dimensional in-depth analysis of the summarized and collated accident data;

[0015] Based on graph theory and matrix theory, a structured model of the causal relationship between various potential risk factors is constructed to quantify the degree of mutual influence between risk factors, analyze the mutual influence path, intensity and direction between the underlying risk factors, and determine the key risk factors for pipeline leakage risk assessment.

[0016] Optionally, Vague set and entropy method are used for calculation respectively to determine the weight values ​​of risk assessment indicators respectively, thereby obtaining the risk value of the first-level assessment indicator, and the mutual influence degree between the risk factors obtained in the decision-making test and the evaluation laboratory method analysis is normalized to obtain the weight value of the first-level assessment indicator;

[0017] The coupling degree model is used to calculate the coupling risk value of the first-level evaluation index, determine the coupling risk value under various coupling scenarios, and determine the impact of multiple factors coupling on the system risk status by comparing and analyzing the coupling risk values;

[0018] A judgment matrix is ​​established and the total impact matrix is ​​calculated. The subjective weights are determined after consistency test, and the EW method is introduced to correct the subjective weights, thus obtaining the combined weights of the first-level evaluation indicators that combine subjective and objective factors.

[0019] Optionally, building a confined space natural gas pipeline leakage risk assessment model specifically includes:

[0020] Using the Bayesian network-based fault tree analysis method to identify the root causes of natural gas pipeline leakage accidents in confined spaces and their potential chain reactions;

[0021] By constructing a Bayesian fault tree model, the logical chain and causal relationship of the accident are revealed;

[0022] The top event node is used as the hub node, connected to the cause node and the result node, and the fault tree is equivalently converted into nodes and edges in the Bayesian network, and the fault tree model is converted into a Bayesian network model.

[0023] Optionally, the judgment matrix is ​​in the form of:

[0024]

[0025] In the formula, a ij It represents the influence of factor i on j, and n is the number of factors.

[0026] Optionally, the formula for calculating the total impact matrix is:

[0027] T=X(IX) -1

[0028] In the formula, X = s·A, I is the identity matrix.

[0029] A confined space natural gas pipeline leakage risk assessment system, comprising:

[0030] Data collection module, which collects various potential risk factors affecting natural gas pipeline leakage in confined spaces;

[0031] The indicator system construction module uses the Delphi method to collect expert opinions and multiple rounds of feedback iterations on various potential risk factors collected, determine the first-level evaluation indicators and key risk factors, and construct a confined space natural gas pipeline leakage risk evaluation indicator system based on the first-level evaluation indicators and key risk factors;

[0032] The weight calculation module uses Vague set and entropy method to calculate and determine the weight values ​​of risk assessment indicators respectively, and then processes the weight values ​​of each risk assessment indicator to obtain the combined weight of the first-level assessment indicators that combines subjective and objective factors;

[0033] Model building module, based on hierarchical Bayesian analysis and Bayesian network, builds a confined space natural gas pipeline leakage risk assessment model;

[0034] The risk assessment module inputs the accident precursor data and the combined weights of the subjective and objective first-level evaluation indicators into the confined space natural gas pipeline leakage risk assessment model, and conducts leakage risk assessment on the evolution process of the causes and consequences of natural gas pipeline leakage in the confined space.

[0035] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method and system for assessing the risk of leakage of a natural gas pipeline in a confined space, which has the following beneficial effects:

[0036] 1. The present invention comprehensively considers multiple dimensions of natural gas pipeline leakage in confined spaces, including leakage source characteristics, environmental conditions, accident scenario evolution, and consequential losses, and constructs a comprehensive and systematic evaluation framework. This comprehensiveness and systematicness ensures the accuracy and reliability of the evaluation results.

[0037] 2. By adopting advanced data processing and analysis technologies such as hierarchical Bayesian analysis (HBA) and Bayesian network (BN), the present invention can accurately reflect the real characteristics of the pipeline and the uncertainty of the operating environment in real time, thereby improving the accuracy of risk prediction. At the same time, the data-driven model can be continuously optimized with the addition of new data, maintaining the real-time and dynamic nature of the assessment results.

[0038] 3. The present invention can be adjusted and optimized according to different confined space environments, pipeline types, leakage source characteristics and other conditions to meet the assessment needs in different scenarios. In addition, by introducing advanced theories such as fuzzy loss rate theory, this method can effectively handle the ambiguity and uncertainty in loss data and improve the applicability of the assessment results. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0040] Figure 1 The present invention provides a flow chart of the method. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] The embodiment of the present invention discloses a method for assessing the risk of leakage of a natural gas pipeline in a confined space. Figure 1 As shown, including:

[0043] Collect various potential risk factors affecting natural gas pipeline leakage in confined spaces;

[0044] The Delphi method was used to collect expert opinions and conduct multiple rounds of feedback iterations on various potential risk factors collected, determine the primary evaluation indicators and key risk factors, and build a confined space natural gas pipeline leakage risk evaluation indicator system based on the primary evaluation indicators and key risk factors;

[0045] Vague set and entropy method are used to calculate and determine the weight values ​​of risk assessment indicators respectively, and then the weight values ​​of each risk assessment indicator are processed to obtain the combined weight of the first-level assessment indicators combining subjective and objective factors;

[0046] Based on hierarchical Bayesian analysis and Bayesian network, a confined space natural gas pipeline leakage risk assessment model is constructed;

[0047] The accident precursor data and the combined weights of the first-level evaluation indicators combining subjective and objective factors are input into the confined space natural gas pipeline leakage risk assessment model to conduct a leakage risk assessment on the evolution process of the causes and consequences of natural gas pipeline leakage in the confined space.

[0048] In some embodiments, the confined space natural gas pipeline leakage risk assessment index system is constructed through the following steps:

[0049] Extensively collect and analyze data on natural gas pipeline accidents that occurred in confined spaces, and extract the key factors and specific problems that affect natural gas pipeline leakage in confined spaces through detailed accident case analysis;

[0050] Exemplarily, the embodiment of the present invention collects information related to the target natural gas pipeline to be analyzed as needed, including pipeline design, construction, operation, environmental conditions, etc., and summarizes and classifies the collected information. At the same time, data on multiple natural gas pipeline accidents that occurred in similar environments were widely collected and deeply analyzed. Through detailed case analysis, the key factors and specific problems affecting pipeline leakage were accurately extracted, and the deep-seated risk causes of natural gas pipelines were further identified and analyzed. In order to further enhance the scientificity and accuracy of the assessment, authoritative literature in related fields was studied in depth, and the Delphi method was used to invite multiple experts to conduct multiple rounds of anonymous feedback and statistical summaries, thereby preliminarily screening and establishing a series of potential risk factors affecting confined space natural gas pipeline leakage.

[0051] Based on these selected risk factors, and referring to the latest research results at home and abroad and the valuable opinions of senior industry experts, these risk factors are systematically classified into four core dimensions (primary evaluation indicators): spatial structure, ventilation conditions, microclimate environment and leakage source. This classification framework not only comprehensively covers the key factors affecting pipeline leakage risks, but also lays a solid foundation for subsequent risk assessment.

[0052] The Decision Experiment and Evaluation Laboratory (DEMATEL) method is used to identify and analyze the interactions between various potential risk factors, and a correlation analysis of risk causal factors is conducted on factors with relevant dependencies and influence relationships. The inherent connections and mutual influence mechanisms between various potential risk factors are deeply explored, and a coupling analysis is conducted around mutually independent intermediate layer factors. An in-depth multi-level and multi-dimensional analysis of the summarized and collated accident data is conducted.

[0053] According to the comparison of the influence degree between risk-causing factors, the direct influence matrix A of risk-causing factors is constructed. The form of the judgment matrix is:

[0054]

[0055] In the formula, a ij It indicates the influence of the i-th factor on the j-th factor, and n is the number of factors.

[0056] The formula for calculating the total impact matrix is:

[0057] T=X(IX) -1

[0058] In the formula, X = s·A, I is the identity matrix.

[0059] Compute the sum of all rows in the total influence matrix:

[0060]

[0061] In the formula, r i It represents the total impact of factor i on other factors.

[0062] Compute the sum of all columns in the total influence matrix:

[0063]

[0064] Based on the calculated sequence (r+c,rc), a causal relationship diagram of risk causal factors is generated.

[0065] Based on graph theory and matrix theory, by constructing a structured model of the causal relationship between various potential risk factors, quantifying the degree of mutual influence between risk factors, analyzing the mutual influence path, intensity and direction between the underlying risk factors, and identifying the key factors that have a significant impact on pipeline leakage risk. This process not only deepens the understanding of the internal evolution mechanism of pipeline leakage risk, but also provides an important basis for formulating targeted risk prevention and control strategies.

[0066] Based on the above four core dimensions and the identified key risk factors, combined with the specific properties and characteristics of each factor, a confined space natural gas pipeline leakage risk assessment index system is designed and constructed, as shown in Table 1.

[0067] Table 1 Confined space natural gas pipeline leakage risk assessment index system

[0068]

[0069] Vague sets are used to determine subjective weights, and the entropy method (EW) is introduced to correct the subjective weights. Game theory is used to process the two weight values, thus obtaining the combined weights of the first-level evaluation indicators that combine subjective and objective factors.

[0070] Use Vague sets to calculate the subjective weights of indicators and construct Vague set expert evaluation matrix V nm , which is of the form:

[0071]

[0072] Calculate the fuzzy value of the indicator:

[0073]

[0074]

[0075] Where, t ij represents the support of the jth expert for indicator i, f ij It represents the degree of opposition of the jth expert to indicator i.

[0076] According to the Vague set utility function, the subjective weight of the indicator is defined as:

[0077]

[0078] In the formula,

[0079] The objective weights of the evaluation indicators are calculated based on the entropy method (EW), and an evaluation database is established based on the constructed indicators and actual data. Its form is:

[0080] X=[x ij ] nxm

[0081] Where n is the number of scoring samples, m is the number of evaluation indicators, and x is ij The risk level score of the i-th indicator factor is given by the j-th evaluator.

[0082] Normalize each indicator and calculate the proportion of the i-th indicator in the j-th sample. The formula is:

[0083]

[0084] In the formula, i=1,2,…,m; j=1,2,…,n.

[0085] Calculate the information entropy value of each indicator:

[0086]

[0087] In the formula, is a constant.

[0088] Calculate the coefficient of variation:

[0089] D j =1-E j

[0090] Determine the objective weight of the evaluation indicators:

[0091]

[0092] Analyze the optimal decision or weight allocation scheme through game theory, optimize the indicator weights, and construct the weight set:

[0093] w k =(w k1 ,w k2 ,…w kn )

[0094] Where k = 1, 2, ..., L, w k is the weight set vector determined by the kth method, n is the number of evaluation indicators, L is the number of weight determination methods selected, and the value of L in this method is 2. Then the formula for any linear combination of these vectors in this method is:

[0095]

[0096] In the formula, w c is a weight set, c1 and c2 are weight combination coefficients.

[0097] According to the idea of ​​the game theory aggregation model, the Nash equilibrium point is selected, and its objective function is:

[0098] min||w c -w k ||2

[0099] Thus, the appropriate weight combination coefficient can be obtained and normalized to obtain:

[0100]

[0101]

[0102] The optimal weight based on game theory combination weighting is obtained as follows:

[0103]

[0104] In a specific embodiment, the embodiment of the present invention fully integrates rich information from multiple data sources under the solid framework of hierarchical Bayesian analysis (HBA), and carefully constructs a highly informative prior probability distribution for key parameters of interest in natural gas pipeline leakage risk assessment. This construction process not only deeply considers the complex dependencies between parameters, but also fully captures uncertainty factors, ensuring that the model accurately maps to the real world.

[0105] By introducing diffusion methods or adopting non-informative prior distributions as starting points, a sophisticated sampling strategy is implemented for the hyperparameters in the model, so as to deeply embed the uncertainty of data samples at the prior distribution level, allowing the model to more realistically reflect the physical characteristics of natural gas pipelines, the complex changes in the operating environment, and the potential risk dynamics.

[0106] The fault tree (BT) analysis method before the Bayesian network (BN) is used to identify the root cause of the natural gas pipeline leakage accident in the confined space and its potential chain consequences. Through the systematic BT model construction, the logical chain and causal relationship of the accident are revealed. Then, the top event node is used as the hub node, connected with the cause node and the result node, and the fault tree is equivalently converted into nodes and edges in the Bayesian network, completing the conversion from the BT model to the BN model, so as to carry out subsequent evolution simulation and probability calculation of the accident scene.

[0107] To ensure the comprehensiveness and accuracy of the model construction, the system collects relevant information on all basic events, consequences, etc. in the model, including but not limited to: (1) Detailed statistics on the actual occurrence or failure frequency of basic events, consequences, etc. on an annual basis to provide a solid data basis for probability assessment; (2) By extensively consulting accident databases or authoritative reliability manuals, the baseline probability of each basic event is verified and determined to ensure the objectivity of the assessment; (3) For specific accident scenarios, on-site investigations and assessments are organized to obtain more accurate actual loss data.

[0108] The Poisson likelihood combined with the gamma prior method is used to predict the number of events in the future time period. This method can effectively deal with the uncertainty under small sample data and improve the stability and reliability of the prediction. For the probability assessment of rare events, a combination of Bernoulli likelihood and beta prior is used. At the same time, combined with expert opinions or on-site statistical data, the hyper-prior parameters are carefully set to further enhance the accuracy and applicability of the prediction.

[0109] Through the above process, the precise occurrence probability of each node in the BN model is calculated, the dynamic evolution analysis of the risk of natural gas pipeline leakage accidents is realized, and the probability prediction results of pipeline leakage accidents are obtained.

[0110] In order to more comprehensively assess the economic losses and social impacts that may be caused by accidents, the fuzzy loss rate theory was introduced, and a detailed loss assessment was conducted for each potential consequence scenario to obtain the fuzzy loss rate assessment results. This assessment process fully considers the fuzziness and uncertainty in the loss data, making the assessment results closer to the actual situation.

[0111] Finally, the probability prediction results of pipeline leakage accidents are combined with the fuzzy loss rate assessment results to obtain the specific risk value of each consequence scenario. There are 5 accident consequence levels in total, and the median of the loss ratio range is:

[0112] R=[0.5 0.05 0.005 0.0005 0.00005]

[0113] Define the probability of loss ratio determined by expert judgment as vector q d , calculate the vector q d The value of the ath element in is:

[0114] q d =[q d1 q d2 q d3 q d4 q d5 ]

[0115]

[0116] Where d is the accident consequence scenario, q d1 is the probability that the loss ratio of consequence d reaches level 1, and so on, q da-c is the matrix q of expert c d The judgment of the ath element in, b is the number of experts, w c is the weight of expert c.

[0117] Calculate the fuzzy loss rate of the consequence state d:

[0118] g d =q d ·R T

[0119]

[0120] Where W is the total investment of the project, W d is the loss caused by the accident consequence d, T d is the loss ratio of accident consequence d, g dis the fuzzy loss rate of accident consequence d.

[0121] Calculate the total risk of pipeline leakage:

[0122]

[0123] Where H is the total risk value of pipeline leakage, g d is the fuzzy loss rate of accident consequence state d, p gd is the probability of occurrence of consequence state d.

[0124] These quantitative risk data provide valuable decision-making support for risk managers, allowing them to formulate more effective and targeted risk management strategies and emergency response measures.

[0125] The confined space natural gas pipeline leakage risk assessment method provided by the embodiment of the present invention studies the natural gas pipeline in the confined space, and comprehensively uses the hierarchical Bayesian analysis (HBA), the method combining the fault tree (BT) and the Bayesian network (BN), statistical tools such as Poisson likelihood and gamma prior, Bernoulli likelihood and beta prior, and fuzzy loss rate theory. The effective combination of these methods can capture and process the complex dependencies and uncertainties in the data, and assess the risk of natural gas pipeline leakage accidents.

[0126] A confined space natural gas pipeline leakage risk assessment system, comprising:

[0127] Data collection module, which collects various potential risk factors affecting natural gas pipeline leakage in confined spaces;

[0128] The indicator system construction module uses the Delphi method to collect expert opinions and multiple rounds of feedback iterations on various potential risk factors collected, determine the first-level evaluation indicators and key risk factors, and construct a confined space natural gas pipeline leakage risk evaluation indicator system based on the first-level evaluation indicators and key risk factors;

[0129] The weight calculation module uses Vague set and entropy method to calculate and determine the weight values ​​of risk assessment indicators respectively, and then processes the weight values ​​of each risk assessment indicator to obtain the combined weight of the first-level assessment indicators that combines subjective and objective factors;

[0130] Model building module, based on hierarchical Bayesian analysis and Bayesian network, builds a confined space natural gas pipeline leakage risk assessment model;

[0131] The risk assessment module inputs the accident precursor data and the combined weights of the subjective and objective first-level evaluation indicators into the confined space natural gas pipeline leakage risk assessment model, and conducts leakage risk assessment on the evolution process of the causes and consequences of natural gas pipeline leakage in the confined space.

[0132] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0133] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for assessing the risk of leakage of a confined space natural gas pipeline, characterized in that: include: Collect various potential risk factors affecting natural gas pipeline leakage in confined spaces; The Delphi method was used to collect expert opinions and conduct multiple rounds of feedback iterations on various potential risk factors collected, determine the primary evaluation indicators and key risk factors, and build a confined space natural gas pipeline leakage risk evaluation indicator system based on the primary evaluation indicators and key risk factors; Vague set and entropy method are used to calculate and determine the weight values ​​of risk assessment indicators respectively, and then the weight values ​​of each risk assessment indicator are processed to obtain the combined weight of the first-level assessment indicators combining subjective and objective factors; Based on hierarchical Bayesian analysis and Bayesian network, a confined space natural gas pipeline leakage risk assessment model is constructed; The accident precursor data and the combined weights of the first-level evaluation indicators combining subjective and objective factors are input into the confined space natural gas pipeline leakage risk assessment model to conduct a leakage risk assessment on the evolution process of the causes and consequences of natural gas pipeline leakage in the confined space.

2. A confined space natural gas pipeline leakage risk assessment method according to claim 1, characterized in that: The confined space natural gas pipeline leakage risk assessment index system includes four primary assessment indicators: space structure, ventilation conditions, microclimate environment and leakage source.

3. A confined space natural gas pipeline leakage risk assessment method according to claim 1, characterized in that: It also includes using decision-making experiments and evaluation laboratory methods to identify and analyze the interaction between various potential risk factors, conducting risk causal factor correlation analysis on factors with relevant dependencies and influence relationships, deeply exploring the internal connections and mutual influence mechanisms between various potential risk factors, conducting coupling analysis around independent intermediate factors, and conducting multi-level and multi-dimensional in-depth analysis of the summarized and collated accident data; Based on graph theory and matrix theory, a structured model of the causal relationship between various potential risk factors is constructed to quantify the degree of mutual influence between risk factors, analyze the mutual influence path, intensity and direction between the underlying risk factors, and determine the key risk factors for pipeline leakage risk assessment.

4. A confined space natural gas pipeline leakage risk assessment method according to claim 3, characterized in that: Vague set and entropy method are used to calculate and determine the weight values ​​of risk assessment indicators respectively, so as to obtain the risk value of the first-level evaluation indicator. The mutual influence degree between risk factors obtained in the process of decision-making test and evaluation laboratory method analysis is normalized to obtain the weight value of the first-level evaluation indicator. The coupling degree model is used to calculate the coupling risk value of the first-level evaluation index, determine the coupling risk value under various coupling scenarios, and determine the impact of multiple factors coupling on the system risk status by comparing and analyzing the coupling risk values; A judgment matrix is ​​established and the total impact matrix is ​​calculated. The subjective weights are determined after consistency test, and the EW method is introduced to correct the subjective weights, thus obtaining the combined weights of the first-level evaluation indicators that combine subjective and objective factors.

5. The method for assessing the risk of leakage of a confined space natural gas pipeline according to claim 1, characterized in that: The construction of a confined space natural gas pipeline leakage risk assessment model specifically includes: Using the Bayesian network-based fault tree analysis method to identify the root causes of natural gas pipeline leakage accidents in confined spaces and their potential chain reactions; By constructing a Bayesian fault tree model, the logical chain and causal relationship of the accident are revealed; The top event node is used as the hub node, connected to the cause node and the result node, and the fault tree is equivalently converted into nodes and edges in the Bayesian network, and the fault tree model is converted into a Bayesian network model.

6. A confined space natural gas pipeline leakage risk assessment method according to claim 4, characterized in that: The form of the judgment matrix is: In the formula, a ij It represents the influence of factor i on j, and n is the number of factors.

7. A confined space natural gas pipeline leakage risk assessment method according to claim 4, characterized in that: The formula for calculating the total impact matrix is: T=X(I-X) -1 In the formula, X = s·A, I is the identity matrix.

8. A confined space natural gas pipeline leakage risk assessment system, using a confined space natural gas pipeline leakage risk assessment method according to any one of claims 1 to 7, characterized in that: include: Data collection module, which collects various potential risk factors affecting natural gas pipeline leakage in confined spaces; The indicator system construction module uses the Delphi method to collect expert opinions and multiple rounds of feedback iterations on various potential risk factors collected, determine the first-level evaluation indicators and key risk factors, and construct a confined space natural gas pipeline leakage risk evaluation indicator system based on the first-level evaluation indicators and key risk factors; The weight calculation module uses Vague set and entropy method to calculate and determine the weight values ​​of risk assessment indicators respectively, and then processes the weight values ​​of each risk assessment indicator to obtain the combined weight of the first-level assessment indicators that combines subjective and objective factors; Model building module, based on hierarchical Bayesian analysis and Bayesian network, builds a confined space natural gas pipeline leakage risk assessment model; The risk assessment module inputs the accident precursor data and the combined weights of the subjective and objective first-level evaluation indicators into the confined space natural gas pipeline leakage risk assessment model, and conducts leakage risk assessment on the evolution process of the causes and consequences of natural gas pipeline leakage in the confined space.

Images