Method for evaluating failure risk of submarine pipeline based on pythagorean fuzzy empowerment method

Abstract

This invention relates to a comprehensive evaluation method for subsea pipeline failure risk based on the Pythagorean fuzzy weighting method, comprising the following steps: Subsea pipeline risk source identification: Analyzing multiple risk factors, including corrosion, third-party damage, and material defects, for potential risks and accidents that may occur during subsea pipeline operation; constructing a Bayesian network model of subsea pipeline risk accidents; Fuzzifying expert opinions; Weighting expert opinions: Calculating the subjective and objective weights of each expert using the analytic hierarchy process (AHP) and the Dematel method, and combining the subjective and objective weighting method to assign weights to expert opinions; Aggregating the weighted expert opinions based on the PTFEHG operator, calculating the expert opinions under the influence of expert weights and position weights, and obtaining the aggregated result of expert opinions; Bayesian network analysis.

Description

Technical Field

[0001] This invention relates to a comprehensive evaluation method for the failure risk of subsea pipelines based on the Pythagorean fuzzy weighting method, which belongs to the category of subsea pipeline risk rating methods. Background Technology

[0002] Offshore oil and gas resources occupy a vital position in global oil and gas resources. With the rapid development of my country's economy, the demand for oil and gas is constantly increasing, which in turn increases the demand for offshore oil and gas resource development. Submarine pipelines are an important tool for oil and gas transportation, therefore, the risk assessment and analysis of submarine pipelines has received increasing attention from the marine engineering community. Due to the special and complex nature of the marine environment, submarine pipeline risks are characterized by complex risk factors, significant risk coupling effects, and severe consequences from risk accidents. Therefore, the assessment and analysis of submarine pipeline failure risks are of great significance for the safe operation of pipelines and accident prevention.

[0003] Current methods used for risk assessment of subsea pipeline failures, both domestically and internationally, often face challenges such as difficulties in collecting prior information, limited scope of fuzzy expert opinions, and strong subjectivity in expert opinions. Traditional fuzzy weighting methods have significant limitations in the field of subsea pipeline risk assessment. Summary of the Invention

[0004] Based on the above, this invention provides a comprehensive risk assessment method for subsea pipeline failure based on the Pythagorean fuzzy weighting method. It uses Pythagorean fuzzy numbers to quantitatively transform expert opinions, and then combines this with the subjective-objective entropy weighting method to calculate the confidence weights of expert opinions. Finally, it constructs a subsea pipeline risk assessment system based on a Bayesian network. The technical solution is as follows:

[0005] A comprehensive assessment method for failure risk of subsea pipelines based on the Pythagorean fuzzy weighting method includes the following steps:

[0006] Step 1, Identification of Risk Sources for Submarine Pipelines: To address potential risks and accidents that may occur during the operation of submarine pipelines, analyze various risk factors, including corrosion, third-party damage, and material defects; construct a Bayesian network model of submarine pipeline risk accidents.

[0007] Step 2, expert opinion fuzzification: Invite members of the expert group to conduct risk assessments on the risk factors identified in the analysis. Based on Pythagorean fuzzy theory, the qualitative evaluation language given by the experts is transformed into quantitative Pythagorean fuzzy numbers.

[0008] Step 3, Expert Opinion Weighting: The Analytic Hierarchy Process (AHP) and the Dematel method are used to calculate the subjective and objective weights of each expert, and the subjective and objective combined weighting method is used to weight the expert opinions.

[0009] Step 4: Aggregate the weighted expert language based on the PTFEHG operator, calculate the expert opinions under the influence of expert weights and position weights, and obtain the aggregated results of expert opinions;

[0010] Step 5, Bayesian network analysis: The aggregated results of expert opinions are transformed into fuzzy failure rates of the subsea pipeline. The failure probability of the subsea pipeline and the posterior probability of each risk source are calculated and ranked through the causal reasoning and cause diagnosis functions of Bayesian network. The key risk sources with higher rankings are screened out to provide guidance for accident prevention and safety decision-making.

[0011] Furthermore, step 3 shall be performed as follows:

[0012] Step 3.1 Using the Analytic Hierarchy Process (AHP), construct the relative importance judgment matrix A = [a...]. ij ], a ij We will use an importance scale for element i relative to element j and verify its consistency ratio.

[0013] Step 3.2 Calculate the pairwise expert E based on the distance measure of the Pythagorean fuzzy number. a E b Distance between evaluation information With consistency S(E) a E b );

[0014] Step 3.3 Use the DEMATEL method to characterize the degree of mutual influence among experts: Construct a normalized direct influence matrix X using expert consensus as an element, and construct a comprehensive influence matrix T, where E is the identity matrix:

[0015] T = [t] ij ] n×n =X(EX) -1

[0016] The influence degree D of the corresponding factor is obtained by summing the elements in the comprehensive influence matrix T row by row. i The degree of influence R of the corresponding factor is obtained by summing the elements in the comprehensive influence matrix T column by column. i ;

[0017] The objective weight of each expert is calculated by combining their influence and the degree to which they are influenced.

[0018]

[0019] In the formula, D represents the objective weight of the experts. i R represents the degree of influence of factor i. i This indicates the degree to which influencing factor i is affected;

[0020] Step 3.4 Integrate subjective expert weights based on the theory of moments estimation. Weighting of objective experts Solve for the optimal value to obtain the expert weights.

[0021] Furthermore, in step 3.1,

[0022] The relative importance judgment matrix is ​​based on the analysis of the hierarchical structure between different factors. It uses high-level elements as the evaluation criteria to make binary comparisons of secondary elements and describes them with a numerical scale to determine the value of each element in the matrix.

[0023] set up:

[0024]

[0025] In the formula, λ max The largest eigenvalue of the judgment matrix is ​​determined by n, where n is the order of the judgment matrix; CI represents the consistency index.

[0026] The consistency randomness ratio (CR) is calculated as CR = CI / RI, where RI is the average randomness consistency index, and its values ​​are shown in the table below:

[0027]

[0028] When CR < 0.1, the judgment matrix satisfies consistency; if CR > 0.1, the judgment matrix needs to be corrected.

[0029] The eigenvector corresponding to the largest eigenvalue is used as the attribute weight. The score of each expert is calculated by weighting the eigenvectors and then normalizing the results to obtain the subjective expert weights. k represents expert k.

[0030] This invention combines expert opinions to assess the operational risks of subsea pipelines, identifies relevant key risk factors, and proposes control measures, providing guidance for subsea pipeline operation. Compared with existing technologies, the beneficial effects of this invention are as follows: (1) Based on Pythagorean fuzzy theory, qualitative expert opinions are transformed into quantitative fuzzy numbers, expanding the fuzzy range of expert opinions and maximizing the retention of information in expert opinions. (2) A combination of subjective and objective weighting methods is used to weight expert opinions, avoiding the influence of the subjectivity of expert opinions on the aggregation results and reducing the data dependence of expert weighting, ensuring the rationality and objectivity of expert weights. (3) The influence of membership degree μ, non-membership degree ν, and hesitation degree π on expert opinions is comprehensively considered, and the PTFEHG operator is used to aggregate expert opinions, expanding the aggregation method of expert opinions. A Bayesian network model of subsea pipeline risk accidents is constructed from multiple aspects, providing guidance for accident prevention and repair strategies. Attached Figure Description

[0031] Figure 1 This is a flowchart of a comprehensive evaluation method for the failure risk of subsea pipelines based on the Pythagorean fuzzy weighting method.

[0032] Figure 2 This is a Bayesian network model diagram of the risk of subsea pipeline failure. Detailed Implementation

[0033] The following describes in detail, with reference to the accompanying drawings and embodiments, a comprehensive evaluation method for the failure risk of subsea pipelines based on the Pythagorean fuzzy weighting method. However, the embodiments of the present invention are not limited thereto.

[0034] Figure 1 The flowchart shows the comprehensive assessment method for failure risk of subsea pipelines based on the Pythagorean fuzzy weighting method. Figure 2 This is a Bayesian network model diagram of the failure risk of the subsea pipeline described in this invention.

[0035] First, a survey was conducted on the risk sources of subsea pipeline failure, and various risk factors, including corrosion, protective defects, and material defects, were analyzed to construct a Bayesian network model of subsea pipeline risk accidents. Marine engineering experts were invited to conduct risk assessments of each risk factor, and their opinions were converted into Pythagorean fuzzy numbers. The Analytic Hierarchy Process (AHP) was used to calculate the attribute scores of each expert and weight them as subjective expert weights. The Dematel method was used to characterize the degree of mutual influence among experts and normalize it into objective expert weights. A combined subjective and objective weighting method was used to assign weights to the fuzzified expert opinions, combining subjective and objective expert weights to construct an optimization model to solve for the final expert weights. The PTFEHG operator was used to aggregate the weighted expert opinions, and the aggregated results were converted into a fuzzy failure rate for subsea pipelines. The fuzzy failure rate was imported into a Bayesian network, and its causal reasoning and cause diagnosis functions were used to calculate the failure probability of the subsea pipeline and the posterior probability of each risk source. Key risk sources with higher probabilities were selected, and corresponding control measures were proposed.

[0036] Step 1: Identification of Risk Sources for Submarine Pipelines. To address potential risks and accidents during the operation of submarine pipelines, this step analyzes various risk factors, including corrosion, protective defects, and material defects, and constructs a Bayesian network topology for submarine pipeline risk accidents.

[0037] Step 2: Expert Opinion Fuzzification. Invite members of the expert panel to conduct risk assessments of the identified risk factors. Based on Pythagorean fuzzy theory, convert the qualitative evaluation language provided by the experts into corresponding Pythagorean fuzzy numbers. Where p i q ir i s i The numerical scale representing the trapezoidal fuzzy number, and the fuzzy interval representing the fuzzy number. These represent the corresponding membership degree and non-membership degree, respectively, and their membership function and non-membership function are defined as follows:

[0038]

[0039] The rules for converting semantic sets and fuzzy numbers are shown in Table 1:

[0040] Table 1 Expert Semantic Transformation Rules

[0041]

[0042] Step 3: Weighting Expert Opinions. The Analytic Hierarchy Process (AHP) and the Demetal method are used to calculate the subjective and objective weights for each expert, and then a combined subjective and objective weighting method is used to assign weights to expert opinions.

[0043] Step 3.1 Using the Analytic Hierarchy Process (AHP), construct a relative importance judgment matrix A = [a...]. ij The consistency ratio of the data is verified. The judgment matrix is ​​based on the analysis of the hierarchical structure between different factors. It uses high-level factors as evaluation criteria to perform binary comparisons on secondary factors and describes them with a numerical scale to determine the value of each element in the matrix.

[0044]

[0045] In the formula, a ij λ is the importance scale for element i relative to element j, and the scaling criteria are shown in Table 2. max To determine the largest eigenvalue of the matrix, n is the order of the matrix.

[0046] Table 2. Judgment Matrix Numerical Scale and Its Meaning

[0047]

[0048]

[0049] We introduce the consistency random ratio CR, i.e., CR = CI / RI. RI is the average random consistency index, and its values ​​are shown in Table 3.

[0050] Table 3 Consistency Index Values

[0051]

[0052] When CR < 0.1, the judgment matrix satisfies consistency; if CR > 0.1, the judgment matrix needs to be corrected.

[0053] The score of each expert is calculated by weighting the eigenvector corresponding to the largest eigenvalue as the attribute weight, and then normalized to obtain the subjective expert weight.

[0054] Step 3.2 Based on the distance measure of the Pythagorean fuzzy number, as shown in equations (2) to (3), calculate the pairwise expert E. a E b Distance between evaluation information With consistency S(E) a E b ).

[0055]

[0056]

[0057] Step 3.3 Use the DEMATEL method to characterize the degree of mutual influence among experts. Construct a normalized direct influence matrix X using expert consensus as an element. Construct a comprehensive influence matrix T using equation (4), where E is the identity matrix.

[0058] T = [t] ij ] n×n =X(EX) -1 (4)

[0059] The influence degree D of the corresponding factor is obtained by summing the elements of matrix T row by row. i The degree of influence R of the corresponding factor is obtained by summing the elements in matrix T column by column. i The calculation formulas are as follows (5) to (6).

[0060]

[0061]

[0062] By combining the influence and the degree of influence of each expert, the objective weight of the expert can be calculated.

[0063]

[0064] In the formula, D represents the objective weight of the experts. i R represents the degree of influence of factor i. i This indicates the degree to which influencing factor i is affected.

[0065] Step 3.4 Integrate subjective expert weights based on the theory of moments estimation. The optimal weight is determined by combining the objective expert weights with the objective expert weights. The calculation method is shown in equation (8):

[0066]

[0067] In the formula, ω k This represents the optimal weight for expert k. This represents the subjective weight of expert k. Represents the objective weight of expert k

[0068] Step 4: Aggregation of expert opinions. Based on the PTFEHG operator, the weighted expert opinions are aggregated to obtain the aggregated results. First, equation (9) is used to calculate the expert opinion β under the influence of expert weights. j :

[0069]

[0070] In the formula, α j Let ω be the Pythagorean trapezoidal fuzzy number representing the opinion of the j-th expert. j Let be the optimal expert weight for the j-th expert, and n be the balance coefficient.

[0071] The fuzzy number β of expert opinions under the influence of expert weights is calculated according to equation (10). i The score function is sorted in descending order to obtain... Represents the fuzzy numbers (β1, β2, ..., β) n A permutation of ) makes

[0072]

[0073] In the formula, s represents the score function of the fuzzy number, μ represents the membership degree of the fuzzy number, and ν represents the non-membership degree of the fuzzy number. This indicates the degree of hesitation of the fuzzy number.

[0074] Based on the trapezoidal Pythagorean fuzzy Einstein hybrid geometric operator (PTFEHG), the fuzzy number β of expert opinions is analyzed. j Aggregation is performed to obtain the aggregated expert opinions under the influence of position weights. The PTFEHG mathematical model is as follows:

[0075]

[0076] In the formula, w = (w1, w2, ..., w n ) T This represents the location-related normal distribution weights. When n=5, the location weights are taken as w=(0.1117, 0.2365, 0.3036, 0.2365, 0.1117) according to the normal distribution method. T Pythagorean trapezoidal fuzzy number

[0077] The centroid method is used to perform defuzzification to obtain the fuzzy probability score (FPS), and then the equations (13) to (14) are used to convert it into the fuzzy failure rate (FFR).

[0078]

[0079]

[0080]

[0081] Step 5 constructs a Bayesian network conditional probability table based on fault tree and OR gate theory. The FFR of each basic node is imported into the model as the prior probability to calculate the failure probability of the top accident and the posterior probability of each node. This identifies the highest-ranking critical risk sources, providing guidance for accident prevention and safety decision-making.

[0082] Example

[0083] This embodiment evaluates the failure risk of an in-service gas pipeline in Liwan District, South China Sea.

[0084] Step 1 analyzes the various risk factors that may occur during the operation of subsea pipelines, and constructs a Bayesian network topology for subsea pipeline risk accidents, such as... Figure 2 As shown in Table 4, the meaning of each risk source is as follows.

[0085] Table 4 Risk Sources of Submarine Pipelines

[0086]

[0087] Step 2: Invite members of the expert panel to conduct a risk assessment of the identified risk factors. Transform the qualitative assessments provided by the experts into corresponding Pythagorean fuzzy numbers according to the rules in Table 2.

[0088] Step 3 involves combining subjective and objective weighting methods to empower expert opinions.

[0089] Step 3.1 Using the Analytic Hierarchy Process (AHP), construct a relative importance judgment matrix A for four aspects: expert title, education, length of service, and age, and verify its consistency ratio:

[0090]

[0091] Solving for λ, we get λ max =4.1725, and the corresponding eigenvector is ω′ = [0.8711, 0.4547, 0.167, 0.0807]. T

[0092] The judgment matrix satisfies the consistency test.

[0093] The score for each expert is calculated by weighting the feature vectors as attribute weights, and then normalized to obtain the subjective expert weight ω. s =[0.243, 0.2196, 0.2243, 0.1495, 0.1636] T .

[0094] Step 3.2 Calculate pairwise expert E a E b Consistency S(E) between a E b ), and construct a normalized matrix X. Calculate the objective expert weight ω using formulas (4) to (7). o = [0.2009, 0.2009, 0.2014, 0.1995, 0.1974] T

[0095] Step 3.3 Construct an optimization model to solve for the optimal weights. The calculation method is shown in Equation (8), and the optimal expert weights ω=[0.2446,0.2181,0.2237,0.1529,0.1608] are obtained. T

[0096] Step 4 aggregates the weighted expert language based on the PTFEHG operator. Taking X2 (outer coating damage) as an example, the expert evaluation of X2 is {H, L, L, H, H}, which is converted into a Pythagorean fuzzy number under the influence of expert weights:

[0097]

[0098]

[0099]

[0100]

[0101]

[0102] The score function of the fuzzy number is calculated using equation (10), and then sorted in descending order to obtain the following sorting results.

[0103]

[0104] The fuzzy numbers of ranked expert opinions are aggregated based on the PTFEHG operator, and the positional weights are taken as w = (0.1117, 0.2365, 0.3036, 0.2365, 0.1117) according to the normal distribution method. TThe aggregated expert opinions under the influence of position weights were obtained as follows: α2 = ([0.3757, 0.5329, 0.6367, 0.8957]; 0.7568, 0.2700, 0.5953), which were then defuzzified using the centroid method and converted to FFR. The FFRs for all risk sources were calculated using the above method, as shown in Table 5.

[0105] Table 5 Probabilities of Each Node

[0106]

[0107]

[0108] Step 5 constructs a Bayesian network conditional probability table based on fault tree and OR gate theory, and imports the FFR of each basic node as prior probability into the model, such as... Figure 2 As shown in Table 5, the failure probability of the top accident, P = 3.64E⁻², and the posterior probabilities of each node are calculated. It can be seen that the posterior probabilities and FFRs of nodes X2, X3, and X8 are much greater than those of the other nodes, requiring focused prevention and control.

Claims

1. A comprehensive assessment method for failure risk of subsea pipelines based on the Pythagorean fuzzy weighting method, comprising the following steps: Step 1, Identification of Risk Sources for Submarine Pipelines: To address potential risks and accidents that may occur during the operation of submarine pipelines, analyze various risk factors, including corrosion, third-party damage, and material defects; construct a Bayesian network model of submarine pipeline risk accidents. Step 2, expert opinion fuzzification: Invite members of the expert group to conduct risk assessments on the risk factors identified in the analysis. Based on Pythagorean fuzzy theory, the qualitative evaluation language given by the experts is transformed into quantitative Pythagorean fuzzy numbers. Step 3, Expert Opinion Weighting: The Analytic Hierarchy Process (AHP) and the Dematel method are used to calculate the subjective and objective weights of each expert, and the subjective and objective combined weighting method is used to weight the expert opinions. Step 4: Aggregate the weighted expert language based on the PTFEHG operator, calculate the expert opinions under the influence of expert weights and position weights, and obtain the aggregated results of expert opinions; Step 5, Bayesian network analysis: The aggregated results of expert opinions are transformed into fuzzy failure rates of the subsea pipeline. The failure probability of the subsea pipeline and the posterior probability of each risk source are calculated and ranked through the causal reasoning and cause diagnosis functions of Bayesian network. The key risk sources with higher rankings are screened out to provide guidance for accident prevention and safety decision-making.

2. The comprehensive assessment method for failure risk of subsea pipelines according to claim 1, characterized in that, Step 3 should be performed as follows: Step 3.1 Using the Analytic Hierarchy Process (AHP), construct the relative importance judgment matrix A = [a...]. ij ], a ij We will use an importance scale for element i relative to element j and verify its consistency ratio. Step 3.2 Calculate the pairwise expert E based on the distance measure of the Pythagorean fuzzy number. a E b Distance between evaluation information With consistency S(E) a E b ); Step 3.3 Use the DEMATEL method to characterize the degree of mutual influence among experts: Construct a normalized direct influence matrix X using expert consensus as an element, and construct a comprehensive influence matrix T, where E is the identity matrix: T=[t ij ] n×n =X(E-X) -1 The influence degree D of the corresponding factor is obtained by summing the elements in the comprehensive influence matrix T row by row. i The degree of influence R of the corresponding factor is obtained by summing the elements in the comprehensive influence matrix T column by column. i ; The objective weight of each expert is calculated by combining their influence and the degree to which they are influenced. In the formula, D represents the objective weight of the experts. i R represents the degree of influence of factor i. i This indicates the degree to which influencing factor i is affected; Step 3.4 Integrate subjective expert weights based on the theory of moments estimation. Weighting of objective experts Solve for the optimal value to obtain the expert weights.

3. The comprehensive evaluation method for failure risk of subsea pipelines according to claim 2, characterized in that, In step 3.1, The relative importance judgment matrix is ​​based on the analysis of the hierarchical structure between different factors. It uses high-level elements as the evaluation criteria to make binary comparisons of secondary elements and describes them with a numerical scale to determine the value of each element in the matrix. set up: In the formula, λ max The largest eigenvalue of the judgment matrix is ​​determined by n, where n is the order of the judgment matrix; CI represents the consistency index. The consistency randomness ratio (CR) is calculated as CR = CI / RI, where RI is the average randomness consistency index, and its values ​​are shown in the table below: When CR < 0.1, the judgment matrix satisfies consistency; if CR > 0.1, the judgment matrix needs to be corrected. The eigenvector corresponding to the largest eigenvalue is used as the attribute weight. The score of each expert is calculated by weighting the eigenvectors and then normalizing the results to obtain the subjective expert weights. k represents expert k.

Images