Risk assessment method for overdue service pressure pipeline and terminal equipment
By constructing a fault tree and using the butterfly optimization algorithm to determine the optimal relaxation factor, the problem of inaccurate risk assessment results for pressure pipelines in existing technologies is solved, and a more reliable risk assessment is achieved.
Patent Information
- Application Number
- CN202511602572.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-04
- Publication Date
- 2026-02-06
AI Technical Summary
The inaccuracy of existing risk assessment methods for pressure pipelines that have exceeded their service life is mainly due to the fact that the relaxation factor β is determined subjectively by humans, leading to inaccurate analysis results.
A similarity clustering method based on butterfly optimization algorithm (BOA-SAM) is adopted. By constructing a fault tree, the fuzzy number of expert ratings is obtained to determine the relative consistency of expert ratings. The optimal relaxation factor is determined by using the similarity clustering method, thereby quantifying the failure probability of bottom events and finally determining the failure probability of top events.
It significantly reduces the subjectivity of the assessment results, improves the reliability of risk assessment for pressure pipelines that have exceeded their service life, and provides more accurate risk quantification analysis.
Smart Images
Figure CN121481232A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of industrial facility safety assessment technology, and specifically relates to a risk assessment method and terminal equipment for pressure pipelines that have exceeded their service life based on BOA-SAM-FTA. It is particularly suitable for the quantitative analysis and management of risks of aging pipeline systems in the fields of petrochemicals and energy transmission. Background Technology
[0002] Pressure pipelines in the petrochemical and energy sectors are core facilities in the oil and gas production and processing industry chain. Their high temperature, high pressure, and hazardous media characteristics mean that failure risks are directly linked to public safety and economic operation. With the increasing aging of pipelines, the problem of exceeding their service life is becoming increasingly prominent. Against this backdrop, conducting risk assessments for pressure pipelines exceeding their service life is particularly important.
[0003] In existing technologies, the methods for risk assessment of pressure pipeline failure accidents are mainly based on fault tree analysis, Bayesian network analysis, and hierarchical analysis.
[0004] Current risk assessment of pressure pipelines suffers from the following technical shortcomings: Because it is difficult to obtain actual failure data of pressure pipelines in reality, most studies using fault tree analysis rely on inviting experts to rate the failure probabilities of the base events in the fault tree. After the expert rating, the similarity aggregation method (SAM) is often used to aggregate expert opinions. However, in the traditional similarity aggregation method, the relaxation factor β is usually determined by human subjectivity, which causes the results of fault tree analysis to change with human subjectivity, resulting in inaccurate analysis results. Summary of the Invention
[0005] The purpose of this invention is to provide a risk assessment method and terminal equipment for pressure pipelines that have exceeded their service life, so as to solve the technical problem of inaccurate assessment results in existing risk assessment methods for pressure pipelines that have exceeded their service life.
[0006] The first aspect of the present invention provides a risk assessment method for pressure pipelines that have exceeded their service life, comprising:
[0007] Construct a fault tree with pressure pipeline failure as the top event;
[0008] Obtain the fuzzy numbers corresponding to multiple expert ratings for each bottom event in the fault tree, and determine the relative consistency of the expert ratings based on the fuzzy numbers;
[0009] The fuzzy numbers are clustered using a similarity clustering method, and the optimal relaxation factor is determined based on the clustering results.
[0010] The failure probability of each bottom event is determined based on the relative consistency of expert ratings and the optimal relaxation factor.
[0011] The failure probability of the top event is determined based on the failure probability of each bottom event.
[0012] Preferably, the fuzzy numbers are clustered using a similarity clustering method, and the optimal relaxation factor is determined based on the clustering results, specifically as follows:
[0013] Use any number between [0,1] as the initial relaxation factor, and use the initial relaxation factor as the initial cluster center;
[0014] The fuzzy numbers are clustered using a similarity clustering method, and the optimization algorithm used in the clustering is the butterfly optimization algorithm.
[0015] Determine the membership function of the fuzzy numbers in each iteration of the butterfly optimization algorithm, and determine the uncertainty value based on the membership function of the fuzzy numbers in each iteration;
[0016] The optimal relaxation factor is determined based on the uncertainty value.
[0017] Preferably, the uncertainty value is determined based on the membership function of the fuzzy number at each iteration, as follows:
[0018]
[0019] In the formula, Let be the uncertainty value of the rating by the i-th expert. Let f(i) be the membership function of the fuzzy number corresponding to the rating of the i-th expert.
[0020] Preferably, the relative consistency of expert ratings is determined based on the fuzzy number, specifically as follows:
[0021] The similarity of expert ratings is determined based on the fuzzy number;
[0022] The average degree of consistency of expert ratings is determined based on the similarity of the expert ratings.
[0023] The relative consistency of expert ratings is determined based on the average consistency of expert ratings.
[0024] Preferably, the failure probability of each bottom event is determined based on the relative consistency of expert ratings and the optimal relaxation factor, specifically as follows:
[0025] The aggregation weight of each expert for each bottom event is determined based on the relative consistency of expert ratings and the optimal relaxation factor.
[0026] The expert summary opinion is determined based on the aggregate weight and the fuzzy number corresponding to the expert rating.
[0027] The expert consensus opinions are defuzzified, and the failure probability of each bottom event is determined based on the defuzzified expert consensus opinions.
[0028] Preferably, the aggregation weight of each expert for each bottom event is determined based on the relative consistency of expert ratings and the optimal relaxation factor, as follows:
[0029] W i =βφ i +(1-β)R i ;
[0030] In the formula, W i Let β be the aggregate weight of the i-th expert, β be the optimal relaxation factor, and φ be the aggregate weight of the i-th expert. i For the weight allocation of the i-th expert, R i The degree of relative consistency in the ratings of the i-th expert.
[0031] Preferably, the expert summary opinion is determined based on the aggregation weight and the fuzzy number corresponding to the expert rating, as follows:
[0032]
[0033] In the formula, To compile expert opinions, W i Let i be the aggregate weight of the i-th expert. The rating of the i-th expert for the bottom event corresponds to the fuzzy number, where N is the total number of experts.
[0034] Preferably, the expert opinions are de-blurred, specifically as follows:
[0035] The centroid method was used to de-fuzzify the expert opinions.
[0036] Preferably, after determining the failure probability of the top event based on the failure probability of each bottom event, the method further includes:
[0037] The importance of each bottom event to the top event is determined based on the failure probability of each bottom event.
[0038] A second aspect of the present invention provides a terminal device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described risk assessment method for pressure pipelines exceeding their service life.
[0039] The risk assessment method and terminal equipment for pressure pipelines exceeding their service life according to the present invention have the following advantages compared with the prior art:
[0040] This invention combines fuzzy numbers from expert ratings to determine the optimal relaxation factor and improves the relaxation factor, thereby significantly reducing the influence of subjectivity and greatly improving the reliability of risk assessment for pressure pipelines that have exceeded their service life. Attached Figure Description
[0041] Figure 1 This is a flowchart of a risk assessment method for pressure pipelines that have exceeded their service life, according to an embodiment of the present invention.
[0042] Figure 2 This is a model diagram of the fault tree in an embodiment of the present invention.
[0043] Figure 3 This is a natural language fuzzy graph judged by experts in an embodiment of the present invention.
[0044] Figure 4 This is a ranking chart of the importance of events in an embodiment of the present invention. Detailed Implementation
[0045] The present invention provides a method for risk assessment of pressure pipelines exceeding their service life, in conjunction with the accompanying drawings and embodiments. However, the present invention is not limited to the examples described herein; these embodiments are provided to disclose the invention more thoroughly and to fully convey its scope to those skilled in the art. The terminology used in the exemplary embodiments illustrated in the drawings is not intended to limit the invention. Unless otherwise stated, the terminology used herein (including technical terms) has its common meaning as understood by those skilled in the art. Furthermore, it is understood that terms defined in commonly used dictionaries should be understood to have a meaning consistent with the context of their relevant field and should not be construed as having an idealized or overly formal meaning.
[0046] The first aspect of this invention provides a risk assessment method for pressure pipelines that have exceeded their service life, such as... Figures 1 to 4 As shown, it includes:
[0047] Step 1: Construct a fault tree (FTA) with pressure pipeline failure as the top event.
[0048] This invention uses pressure pipeline failure as the top event, corrosion-related failures, material degradation-related failures, mechanical damage, and connection failures as intermediate events, and stress corrosion cracking, hydrogen embrittlement, and corrosion fatigue as bottom events in the fault tree. A fault tree and a fault tree event description table are constructed based on this. The constructed fault tree is as follows: Figure 2 As shown in Table 1, the event table is as follows.
[0049] Table 1 Fault Tree Events
[0050] serial number Event Name serial number Event Name serial number Event Name T Pressure pipeline failure M42 Thread failure X12 The inner coating thickness has decreased or peeled off. M1 Corrosion-related failures M43 Welding defects X13 Third-party sabotage M2 Material degradation-related failures X1 High transport velocity X14 external shock M3 Mechanical damage X2 <![CDATA[High H2S content in the medium]]> X15 Construction defects M4 Connection failure X3 <![CDATA[High CO2 content in the medium]]> X16 weld cracking M11 Internal corrosion X4 Stress corrosion cracking X17 Gasket deterioration M12 External corrosion X5 External chemical corrosion X18 Insufficient bolt warning force M13 Special corrosion X6 Stray current corrosion X19 Thread corrosion leakage M21 Temperature effect X7 Hydrogen embrittlement X20 Over-tightening strips threads M22 Coating failure X8 Corrosion fatigue X21 Incomplete root penetration M31 External damage X9 High temperature aging X22 Heat-affected zone cracks M32 Internal defects X10 Low-temperature brittleness M41 Flange leakage X11 Water accumulation under the anti-corrosion layer
[0051] Step 2: Obtain the fuzzy numbers corresponding to multiple expert ratings for each bottom event in the fault tree, and determine the relative consistency of the expert ratings based on the fuzzy numbers. Specifically:
[0052] Step 2.1: Obtain the fuzzy number corresponding to the multiple expert ratings for each bottom event in the fault tree.
[0053] This invention employs an approach where industry experts are invited to rate the events at the bottom of the fault tree. The experts assess the probability of each event occurring based on their professional knowledge and experience. In semantic evaluation methods, events are typically categorized into different levels of probability, such as... Figure 3 As shown.
[0054] In this embodiment of the invention, seven industry experts were invited to rate the 22 bottom events of the fault tree, and the rating results are shown in Table 2.
[0055] Table 2 Expert Ratings
[0056] expert 1 2 3 4 5 6 7 expert 1 2 3 4 5 6 7 X1 L H H L L ML ML X12 VL M M VL M L VL X2 VH H VH H M H VH X13 MH ML MH MH H MH H X3 L L ML ML L M ML X14 ML ML MH MH L MH H X4 M MH H MH MH MH H X15 M ML M M ML ML ML X5 ML ML H H M MH MH X16 MH M MH M MH H VH X6 MH L MH MH H ML ML X17 M ML M ML ML ML ML X7 ML H L H L ML M X18 ML ML M ML ML ML ML X8 VL H VL H VL MH MH X19 L ML ML ML L ML ML X9 ML L ML L L ML ML X20 L ML M ML L ML ML X10 VL L VL VL L ML ML X21 ML ML MH ML L MH H X11 VL VL VL VL L VL M X22 ML ML VL VL VL ML ML
[0057] Step 2.2: Determine the relative consistency of expert ratings based on fuzzy numbers, specifically as follows:
[0058] Step 2.2.1: Determine the similarity of expert ratings based on fuzzy numbers, as shown in formula (1):
[0059]
[0060] In the formula, Let be the similarity between the ratings of the i-th expert and the j-th expert for the same event. and These are the ratings of the i-th expert and the j-th expert for the same basic event, respectively, and are trapezoidal fuzzy numbers. and They are and The membership function. In measuring all S... ij (Right now After that, a protocol matrix (AM) can be constructed, which represents the degree of consensus among experts.
[0061]
[0062] Step 2.2.2: Determine the average consistency of expert ratings based on the similarity of the expert ratings, as shown in formula (3):
[0063]
[0064] In the formula, B i Let be the average degree of consistency in the ratings of the i-th expert, and N be the total number of experts.
[0065] Step 2.2.3: Determine the relative consistency of expert ratings based on the average consistency of expert ratings, as shown in formula (4):
[0066]
[0067] In the formula, R i The degree of relative consistency in the ratings of the i-th expert.
[0068] The purpose of determining the relative consistency of the rating of the i-th expert in this embodiment of the invention is to determine the aggregation weight W of the i-th expert. i :
[0069] W i =βφ i +(1-β)R i (5)
[0070] In the formula, β∈[0,1] is the relaxation factor, reflecting whether individual opinions or public opinions are valued; φ i The weights assigned to the experts are shown in Table 3.
[0071] Table 3 Expert Weight Allocation Table
[0072] job title Education On-site work experience Rank weight Rank weight Rank weight Senior Rank 5 PhD 5 ≥2510 Subtropical high 4 Master's 4 15≤q＜258 Middle 3 Undergraduate 3 8≤q＜156 Grade 2 Associate Degree 2 2≤q＜84 Technology 1 Secondary vocational school 1 q＜22
[0073] In the prior art, the relaxation factor is often determined objectively by humans, usually set to 0.5, which has the problem of strong subjectivity. In order to overcome this problem, the embodiments of the present invention use the subsequent step 3 to determine the optimal relaxation factor.
[0074] Although experts were invited to rate the events at the bottom of the fault tree, these semantic values alone cannot quantify the probability of failure of the events. Therefore, this embodiment aggregates expert opinions through similarity aggregation and then uses fuzzy set theory to transform the semantic values provided by the experts into probability values, thereby quantifying the probability of failure of the events, as detailed in subsequent steps 3 and 4.
[0075] Step 3: Cluster the fuzzy numbers using the Similarity Aggregation Method (SAM), and determine the optimal relaxation factor based on the clustering results. Specifically:
[0076] Step 3.1: Use any number between [0,1] as the initial relaxation factor β, and use the initial relaxation factor β as the initial cluster center.
[0077] Step 3.2: Cluster the fuzzy numbers using the similarity clustering method. The optimization algorithm used in the clustering is the butterfly optimization algorithm.
[0078] The similarity clustering method in the embodiments of this invention can be cosine similarity clustering method, hash clustering method, etc.
[0079] Step 3.3: Determine the membership function of the fuzzy number in each iteration of the Butterfly Optimization Algorithm (BOA), and determine the uncertainty value based on the membership function of the fuzzy number in each iteration, as shown in formula (6):
[0080]
[0081] In the formula, Let be the uncertainty value of the rating of the i-th expert, and let be the objective function of the butterfly optimization algorithm. Let be the membership function of the fuzzy number of the rating of the i-th expert.
[0082] Step 3.4: Determine the optimal relaxation factor based on the uncertainty value.
[0083] In this embodiment of the invention, after obtaining the uncertainty value at each iteration, the optimal relaxation factor corresponding to the uncertainty value can be obtained.
[0084] Because the search process of the BOA algorithm has a certain degree of randomness, the optimal value obtained each time is different. Therefore, in this embodiment, the BOA algorithm is run 500 times, and the average value is taken as the optimal relaxation factor. Finally, the optimal relaxation factor is found to be 0.97.
[0085] The implementation of step 3 above in the Matlab program is as follows: Input the expert's basic information and the fuzzy number of the expert rating into the Matlab program. The relaxation factor is an arbitrary number between [0,1]. Then, use the SAM method to aggregate these data to obtain the results. Finally, calculate the uncertainty value of the aggregated results. This operation will iterate hundreds or thousands of times in the BOA program, obtaining the relaxation factor corresponding to each uncertainty value, and finally taking the average value as the optimal relaxation factor β.
[0086] Step 4: Determine the failure probability of each bottom event based on the relative consistency of expert ratings and the optimal relaxation factor β, specifically as follows:
[0087] Step 4.1: Based on the relative consistency R of the expert ratings i The optimal relaxation factor β determines the aggregate weight W for each expert on each bottom event. i Specifically, the aggregation weight of the experts is determined using the formula (5) above.
[0088] Step 4.2: Based on the aggregation weight W i The fuzzy numbers corresponding to the expert ratings are used to determine the expert summary opinions, as shown in formula (7):
[0089]
[0090] In the formula, To compile expert opinions, W i Let i be the aggregate weight of the i-th expert. The fuzzy number corresponding to the rating of the i-th expert for the bottom event. The expert summary opinion for each bottom event can be obtained according to formula (7).
[0091] Step 4.3: Defuzzify the expert summary opinions and determine the failure probability of each bottom event based on the defuzzified expert summary opinions.
[0092] Since the expert opinions obtained in the above process are fuzzy numbers, they need to be defuzzified. In this embodiment of the invention, the centroid method is used for defuzzification, as shown in formula (8):
[0093]
[0094] In the formula, P represents the expert summary opinion after defuzzification, and a, b, c, and d represent the four parameters of the trapezoidal fuzzy number, respectively.
[0095] After deblurring, the deblurred expert consensus P for each base event is converted into a failure probability using the Onisawa formula. The Onisawa formula is shown below:
[0096]
[0097] In the formula, FP is the failure probability, and k is as shown in formula (10):
[0098]
[0099] The failure probabilities of the 22 bottom events obtained in the embodiments of the present invention, as well as the uncertainty values of the traditional SAM method and the improved BOA-SAM method, are shown in Table 4.
[0100] Table 4 Comparison of Results
[0101]
[0102] Step 5: Determine the failure probability of the top event based on the failure probability of each bottom event.
[0103] In the fault tree, each bottom event is connected to an intermediate event via an AND gate and an OR gate, and they are independent of each other. The formulas for calculating the AND and OR gates are as follows:
[0104]
[0105] In the formula, n is the total number of base events, FP(X) i Let P(t) be the failure probability of the i-th bottom event, and P(t) be the failure probability of the top event.
[0106] The embodiment of the present invention further includes the following after step 5:
[0107] Based on the failure probability of each bottom event, the importance of each bottom event to the top event is determined. The formulas for calculating the logical AND and OR gates are as follows:
[0108]
[0109] The failure probability P(T) of the top event in the fault tree was calculated to be 0.117641. The importance ranking of the bottom events is as follows: Figure 4 As shown.
[0110] pass Figure 4 Importance analysis reveals that stress corrosion cracking, third-party damage, weld cracking, and high H2S content in the medium are key factors leading to pressure pipeline failure.
[0111] This invention is essentially a risk assessment method for pressure pipelines that have exceeded their service life based on BOA-SAM-FTA. By improving the relaxation factor, the subjectivity of the assessment results can be significantly reduced, thereby improving the reliability of the assessment results.
[0112] The foregoing has shown and described the basic principles, main features, and advantages of this invention. Through practical application, the above methods provide a reference for enterprises and relevant practitioners to take timely and accurate measures to assess the risks of pressure pipelines exceeding their service life.
[0113] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.
Claims
1. A risk assessment method for pressure pipelines exceeding their service life, characterized in that, include: Construct a fault tree with pressure pipeline failure as the top event; Obtain the fuzzy numbers corresponding to multiple expert ratings for each bottom event in the fault tree, and determine the relative consistency of the expert ratings based on the fuzzy numbers; The fuzzy numbers are clustered using a similarity clustering method, and the optimal relaxation factor is determined based on the clustering results. The failure probability of each bottom event is determined based on the relative consistency of expert ratings and the optimal relaxation factor. The failure probability of the top event is determined based on the failure probability of each bottom event.
2. The risk assessment method for pressure pipelines exceeding their service life according to claim 1, characterized in that, The fuzzy numbers are clustered using a similarity clustering method, and the optimal relaxation factor is determined based on the clustering results, specifically as follows: Use any number between [0,1] as the initial relaxation factor, and use the initial relaxation factor as the initial cluster center; The fuzzy numbers are clustered using a similarity clustering method, and the optimization algorithm used in the clustering is the butterfly optimization algorithm. Determine the membership function of the fuzzy numbers in each iteration of the butterfly optimization algorithm, and determine the uncertainty value based on the membership function of the fuzzy numbers in each iteration; The optimal relaxation factor is determined based on the uncertainty value.
3. The risk assessment method for pressure pipelines exceeding their service life according to claim 2, characterized in that, The uncertainty value is determined based on the membership function of the fuzzy number at each iteration, as follows: In the formula, Let be the uncertainty value of the rating by the i-th expert. Let f(i) be the membership function of the fuzzy number corresponding to the rating of the i-th expert.
4. The risk assessment method for pressure pipelines exceeding their service life according to claim 1, characterized in that, The relative consistency of expert ratings is determined based on the fuzzy number, specifically as follows: The similarity of expert ratings is determined based on the fuzzy number; The average degree of consistency of expert ratings is determined based on the similarity of the expert ratings. The relative consistency of expert ratings is determined based on the average consistency of expert ratings.
5. The risk assessment method for pressure pipelines exceeding their service life according to claim 1, characterized in that, The failure probability of each bottom event is determined based on the relative consistency of expert ratings and the optimal relaxation factor, specifically as follows: The aggregation weight of each expert for each bottom event is determined based on the relative consistency of expert ratings and the optimal relaxation factor. The expert summary opinion is determined based on the aggregate weight and the fuzzy number corresponding to the expert rating. The expert consensus opinions are defuzzified, and the failure probability of each bottom event is determined based on the defuzzified expert consensus opinions.
6. The risk assessment method for pressure pipelines exceeding their service life according to claim 5, characterized in that, The aggregation weight of each expert for each bottom event is determined based on the relative consistency of expert ratings and the optimal relaxation factor, as follows: W i =bf i +(1-β)R i ; In the formula, W i Let β be the aggregate weight of the i-th expert, β be the optimal relaxation factor, and φ be the aggregate weight of the i-th expert. i For the weight allocation of the i-th expert, R i The degree of relative consistency in the ratings of the i-th expert.
7. The risk assessment method for pressure pipelines exceeding their service life according to claim 5, characterized in that, The expert summary opinion is determined based on the aggregated weights and the fuzzy numbers corresponding to the expert ratings, as follows: In the formula, To compile expert opinions, W i Let i be the aggregate weight of the i-th expert. The rating of the i-th expert for the bottom event corresponds to the fuzzy number, where N is the total number of experts.
8. The risk assessment method for pressure pipelines exceeding their service life according to claim 5, characterized in that, The aforementioned expert opinions were de-fuzzified, specifically as follows: The centroid method was used to defuzzify the expert opinions.
9. The risk assessment method for pressure pipelines exceeding their service life according to claim 1, characterized in that, After determining the failure probability of the top event based on the failure probability of each bottom event, the method further includes: The importance of each bottom event to the top event is determined based on the failure probability of each bottom event.
10. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the risk assessment method for pressure pipelines exceeding their service life as described in any one of claims 1 to 9.