Pipeline corrosion bilateral early warning method based on dynamic weight fusion

Abstract

The invention discloses a pipeline corrosion bilateral early warning method based on dynamic weight fusion, and belongs to the technical field of oil and gas pipeline safety intelligent monitoring. According to the method, aiming at the problems of single-model evaluation dimension limitation and insufficient risk factor coupling analysis in the existing corrosion early warning technology, comprehensive early warning is realized by constructing a failure early warning and risk early warning dual-model cooperation mechanism. According to the technical scheme, the method comprises the steps of collecting multi-source corrosion monitoring data and classifying the multi-source corrosion monitoring data into a failure feature set and a risk feature set, establishing a corrosion failure early-warning model based on a residual strength criterion to output a failure early-warning score, adopting a DHGF algorithm to establish a corrosion risk early-warning model to output a risk early-warning score, utilizing an elk optimization algorithm to optimize a dynamic weight coefficient, and obtaining a corrosion risk early-warning result. And obtaining a comprehensive early warning score through weighted fusion. According to the method, the problem of misjudgment caused by single early warning dimension and weight solidification of a traditional method is solved, the accuracy and decision reliability of pipeline corrosion early warning are remarkably improved, and the method is suitable for intelligent safety monitoring of oil and gas pipelines.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent safety monitoring of oil and gas pipelines, and in particular to a bilateral early warning method for pipeline corrosion based on dynamic weight fusion. Background Art

[0002] As the arterial system of modern energy transportation, oil and gas pipelines are exposed to complex and changing service environments for a long time. Their corrosion problems have become a core risk source threatening pipeline safety. Corrosion early warning technology is to detect and prevent behaviors that may cause pipeline damage in advance by monitoring, analyzing and predicting corrosion risks within the system. However, current corrosion early warning technologies mainly rely on a single assessment dimension: one is to predict the risk of structural failure through physical parameters such as wall thickness detection and defect assessment; the other is based on a risk assessment model, which comprehensively considers factors such as environment and operation and maintenance to classify risk levels. However, existing methods have significant defects:

[0003] (1) Fragmentation of assessment dimensions: Existing technologies typically apply failure prediction or risk assessment models independently, lacking a coordinated analysis of physical failure mechanisms and systemic risk factors. For example, residual strength models are unable to reflect the risk of accelerated corrosion caused by environmental mutations, while risk assessment models ignore the dynamic attenuation of the structure's actual bearing capacity, resulting in one-sided warning results.

[0004] (2) Rigid weight allocation: A few methods that attempt to integrate multiple models use a fixed weight strategy, which cannot adapt to the changing characteristics of different sections and different service stages of the pipeline. For example, high-consequence areas need to focus on environmental risk monitoring, while old pipeline sections should increase the failure prediction weight. The fixed weight model leads to the omission of key risks.

[0005] (3) Insufficient use of historical data: Traditional methods have not fully explored the correlation between historical warning data and the actual corrosion status. The adjustment of model parameters relies on manual experience, making it difficult to achieve adaptive optimization of warning strategies.

[0006] These issues lead to existing early warning systems suffering from high false alarm rates and delayed responses to critical risks. The limitations of a single model are particularly pronounced in complex geological environments and variable operating conditions. Therefore, there is an urgent need to develop a bilateral early warning method that can dynamically integrate multi-dimensional assessment results and adaptively optimize weight allocation to achieve accurate early warning of pipeline corrosion risks. Summary of the Invention

[0007] The present invention provides a bilateral early warning method for pipeline corrosion based on dynamic weight fusion, which is characterized by comprising the following steps:

[0008] S1: Collect multi-source corrosion monitoring data of the target oil and gas pipeline, and classify the monitoring data into a failure feature data set and a risk feature data set;

[0009] S2: Establishing a pipeline failure risk warning model based on the pipeline residual strength evaluation criteria, importing the failure feature data set into the failure warning model to perform pipeline failure warning, and outputting a pipeline corrosion failure score F1 according to the failure warning level;

[0010] S3: Using the DHGF algorithm to establish a pipeline corrosion risk warning model, input the risk feature data set into the risk warning model for comprehensive evaluation, and output the pipeline corrosion risk score F2;

[0011] S4: Obtain historical corrosion warning results of oil and gas pipelines, including the historical pipeline corrosion failure score F1 his , Pipeline Corrosion Risk Historical Score and corrosion bilateral early warning history score F his ;

[0012] S5: Establish a dynamic weight optimization model, import historical oil and gas pipeline corrosion warning results into the dynamic weight optimization model, and use the ELK optimization algorithm to solve the model to obtain the bilateral warning dynamic weight, which includes the pipeline failure warning weight α and the pipeline risk warning weight β;

[0013] S6: combining the bilateral warning dynamic weight, performing weighted calculation on the pipeline corrosion failure score and the pipeline corrosion risk score to obtain a pipeline corrosion bilateral warning score, and determining the pipeline corrosion bilateral warning level according to a preset corrosion bilateral warning score threshold;

[0014] The weighted calculation formula is as follows:

[0015] F=αF1+βF2

[0016] Where, F is the pipeline corrosion bilateral warning score; F1 is the pipeline corrosion failure score; F2 is the pipeline corrosion risk score; α is the pipeline failure warning weight; β is the pipeline risk warning weight;

[0017] The corresponding relationship between the corrosion bilateral warning score and the corrosion bilateral warning level is:

[0018] When F∈[0,30), the bilateral corrosion warning level is low; when F∈[30,50), the bilateral corrosion warning level is relatively low; when F∈[50,70), the bilateral corrosion warning level is medium; when F∈[70,85), the bilateral corrosion warning level is relatively high; when F∈[85,100], the bilateral corrosion warning level is high.

[0019] Furthermore, the failure feature set described in step S1 includes pipeline design pressure, pipeline nominal diameter, pipeline nominal wall thickness, material yield strength, pipeline design coefficient, pipeline weld coefficient, pipeline service time, pipeline monitoring corrosion rate, pipeline corrosion margin and allowable residual strength factor;

[0020] The risk characteristic data set can be divided into numerical indicators and non-numerical indicators; non-numerical indicators can be divided into positive non-numerical indicators and negative non-numerical indicators, and divided into high, relatively high, medium, relatively low and low according to their degree. The risk value of positive non-numerical indicators is positively correlated with the degree, and the risk value of negative non-numerical indicators is negatively correlated with the degree;

[0021] Numerical indicators include pipeline monitoring corrosion rate, pipeline service life, pipeline gas pressure and patrol frequency;

[0022] Positive non-numerical indicators include pipeline medium hazard, construction defect level, pipe material defect level, soil corrosiveness, surrounding work conditions, natural disaster level, and population density around the pipeline;

[0023] Negative non-numerical indicators include the degree of internal corrosion protection, the strength of external corrosion protection, the degree of implementation of the management plan, and the degree of emergency plan formulation.

[0024] Furthermore, the pipeline failure risk warning described in step S2 includes a two-level evaluation process, and the specific steps are as follows:

[0025] S201: Calculate the minimum required wall thickness δ min , the calculation formula is as follows;

[0026]

[0027] Where, δ min is the minimum required wall thickness of the pipeline; ψ is the pipeline design pressure; D is the nominal diameter of the pipeline; ε is the material yield strength; φ is the pipeline design coefficient; λ is the pipeline weld coefficient;

[0028] S202: Calculate the remaining wall thickness of the pipe δ mm , the calculation formula is as follows;

[0029] δ mm =δ-nc r

[0030] Where, δ mm is the remaining wall thickness of the pipeline; δ is the nominal wall thickness of the pipeline; n is the service time of the pipeline; c r Monitor corrosion rates for pipelines;

[0031] S203: Calculate the remaining wall thickness ratio R of the pipe t , the calculation formula is shown as follows;

[0032]

[0033] Where R t is the remaining wall thickness ratio of the pipe;

[0034] S204: When the remaining wall thickness of the pipe is greater than R t Less than the allowable residual strength factor RSF a , the shell parameter Q is calculated as follows; when the remaining wall thickness of the pipe is greater than R t Greater than the allowable residual strength factor RSF a , the shell parameter Q is 50;

[0035]

[0036] Calculate the pipe uniform thickness length L, the calculation formula is as follows:

[0037]

[0038] Where, L is the uniform thickness length of the pipe;

[0039] S205: Execute the first-level warning process. If the first-level corrosion criterion is met, the pipeline corrosion failure warning is in an acceptable state. If the first-level corrosion criterion is not met, enter the second-level warning process. The first-level corrosion criterion expression is:

[0040] δ mm -FCA≥RSF a δ min

[0041] δ mm -FCA≥max[0.5δ min , δ lim ]

[0042] δ mm -FCA≥max[0.5δ min , δ lim ]

[0043] Where, FCA is the pipeline corrosion allowance; δ lim is the pipe limit wall thickness;

[0044] S205: Calculate the pipe's ultimate wall thickness δ lim ,, the calculation formula is as follows:

[0045] δ lim =max[0.2δ,2.5mm]

[0046] Where, δ lim is the pipe limit wall thickness;

[0047] S206: Execute the first-level warning process. If the first-level corrosion criterion is met, the pipeline corrosion failure warning is in an acceptable state. If the first-level corrosion criterion is not met, enter the second-level warning process. The first-level corrosion criterion expression is:

[0048] δ mm -FCA≥RSF a δ min

[0049] δ mm -FCA≥max[0.5δ min , δ lim ]

[0050] δ mm -FCA≥max[0.5δ min , δ lim ]

[0051] Where, FCA is the pipeline corrosion allowance;

[0052] S207: Execute the secondary warning process. If the secondary corrosion criterion is met, the pipeline corrosion failure warning is in a risk state. If the secondary corrosion criterion is not met, the pipeline corrosion failure warning is in an unsafe state. The secondary corrosion criterion expression is:

[0053] R t ≥0.2

[0054]

[0055] S208: Outputting the pipeline corrosion failure score F1 according to the failure warning level. When the pipeline failure warning is in an acceptable state, the pipeline corrosion failure score F1 is 30 points. When the pipeline failure warning is in a risky state, the pipeline corrosion failure score F1 is 60 points. When the pipeline failure warning is in an unsafe state, the pipeline corrosion failure score F1 is 90 points.

[0056] Furthermore, the specific steps of step S3 are as follows:

[0057] S301: Constructing a pipeline corrosion risk early warning indicator system through the Delphi method, wherein the early warning indicator system is composed of a target layer, a criterion layer, and an indicator layer;

[0058] The target layer is pipeline corrosion risk; the criterion layer is pipeline conditions, emergency response capabilities, and environmental conditions; the indicator layer is all indicators in the risk characteristic dataset. Pipeline conditions include pipeline monitoring corrosion rate, pipeline service life, pipeline gas pressure, patrol frequency, pipeline medium hazard, internal corrosion protection level, external corrosion protection strength, and material defect level; emergency response capabilities include the degree of management plan implementation and the degree of emergency plan formulation; and environmental conditions include soil corrosivity, surrounding operating conditions, degree of natural disasters, and population density around the pipeline.

[0059] S302: Calculate the weights of pipeline corrosion risk warning indicators using the analytic hierarchy process; construct a judgment matrix A around the risk warning indicators and calculate the subjective weight vector of the indicators. The calculation formula is as follows:

[0060]

[0061] Where M i is the vector product of each row of the judgment matrix; a ij is the element in the judgment matrix; is the nth square root of the vector product; W i 1 is the subjective weight vector of the indicator;

[0062] Perform a consistency test on the judgment matrix A. If CR < 0.1, the judgment matrix A passes the consistency test. Otherwise, the judgment matrix A fails the consistency test and needs to be reconstructed. The consistency test formula is as follows:

[0063]

[0064] Where a i is the early warning indicator; i is the subjective weight of the indicator; max is the maximum eigenvalue of the judgment matrix; n is the order of the judgment matrix; CI is the consistency index; RI is the consistency ratio;

[0065] The subjective weights of the indicators are objectively corrected through the entropy weight method, and the comprehensive weight matrix W of the early warning indicators is constructed. The correction process is shown in the following formula:

[0066]

[0067] Where r ij is the index proportion; s i is the index entropy value; W i 2 is the indicator information weight; W i is the comprehensive weight of the indicators;

[0068] S303: Construct an indicator sample matrix based on the expert evaluation results, determine the indicator warning levels as high, relatively high, medium, relatively low and low, and the corresponding level threshold set is D = {d n1 ,d n2 ,d n3 ,d n4}={9,7,5,3}, and establish the whitening weight function of each warning level, as shown in the following formula:

[0069]

[0070]

[0071] Where, d ni is the warning level threshold; f1(d ni ) is the measurement function corresponding to the high warning level; f2(d ni ) is the warning level comparison measurement function; f3(d ni ) is the measurement function corresponding to the medium warning level; f4(d ni ) is the measurement function corresponding to the lower warning level; f5(d ni ) is the measurement function corresponding to the low warning level;

[0072] S304: Obtain the membership degree of the early warning indicator under different gray categories according to the whitening weight function, calculate the gray statistical value of the early warning indicator, obtain the gray evaluation weight after normalization, and construct the gray fuzzy weight matrix. The process is shown in the following formula:

[0073]

[0074] Where h jk is the statistical value of the jth indicator belonging to the kth gray class; f k (d ij ) is the membership degree of the jth indicator to the kth gray class; r jk is the evaluation weight of the jth indicator belonging to the kth gray class; R is the gray fuzzy weight matrix;

[0075] S304: Construct a fuzzy comprehensive evaluation matrix P based on the early warning indicator comprehensive weight matrix W and the grey fuzzy weight matrix R. The expression is as follows:

[0076]

[0077] Determine the early warning indicator risk level set as V = {v1, v2, v3, v4, v5} = {10, 8, 6, 4, 2}. Calculate the pipeline corrosion risk score F2 based on the early warning indicator scoring vector. The calculation formula is as follows:

[0078]

[0079] Where p ij is the element in the fuzzy comprehensive evaluation matrix; v i is an element in the fuzzy comprehensive evaluation matrix.

[0080] Furthermore, the specific steps of step S5 are as follows:

[0081] S501: Establish a dynamic weight optimization model, the model objective expression is as follows:

[0082]

[0083] α+β=1

[0084] α>0

[0085] β>0

[0086] Where, F1 his Score the pipeline corrosion failure history, F is the pipeline corrosion risk history score; his is the corrosion bilateral warning history score; α is the pipeline failure warning weight and β is the pipeline risk warning weight;

[0087] S502: Using the model objective function as the fitness function, randomly initialize and generate pipeline failure warning weight α and pipeline risk warning weight β as the initial elk population. The generation process expression is as follows:

[0088] X i =(U i -L i )r+L i

[0089] Where, X i is the initial elk population; U i To solve the upper bound of the space; L i To find the lower bound of the space;

[0090] S503: Set the initial parameters of the algorithm, including the maximum number of iterations T max , male elk proportion Br and population size N;

[0091] S504: Calculate the number B of male elk in the population based on the proportion of male elk. Calculate the fitness of each elk individual in the initial elk population using the fitness function and sort them. Define the first B elk individuals as male elk individuals. The expression is:

[0092] B=Br×N

[0093] S505: A roulette wheel selection mechanism is used to assign mates to male elk individuals, based on the proportion of the male elk individual fitness value in the total fitness value of the population. The expression is:

[0094]

[0095] Where p k The probability of selecting a mate for a male elk individual; f(x k ) is the individual fitness value of male elk;

[0096] S506: The male elk individual that has selected a mate enters the breeding stage and generates new elk individuals. The updating formula is:

[0097] x i (t+1)=x k (t)+α·(x k (t)-x j (t))

[0098] Where x i (t+1) is a newborn elk individual; x k (t) is a male elk individual; x j (t) is a female elk individual; α is the genetic coefficient, which is a random number in the range of [0,1];

[0099] S506: The male elk individuals of the selected female individuals enter the breeding stage, and are updated to generate new male elk individuals and new female elk individuals. The update formula for the new male elk individuals is:

[0100] x k (t+1)=x k (t)+α·(x k (t)-x j (t))

[0101] The individual update formula for newborn female elk is:

[0102] x j (t+1)=x k (t)+β·(x h (t)-x j (t))+γ(x r (t)-x j (t))

[0103] Where x k (t+1) is a newborn male elk individual; x k (t) is a male elk individual; x j (t) is a female elk individual; α is the male genetic coefficient, which is a random number in the range of [0,1]; x h (t) is a random male elk individual; x r (t) is a random female elk individual; β is the male genetic coefficient and γ is a random number in the range [0,1];

[0104] S507: Judgment t≥T max If so, the algorithm is terminated and the optimal pipeline failure warning weight α and pipeline risk warning weight β are output; otherwise, the algorithm returns to S506 to continue iteration. BRIEF DESCRIPTION OF THE DRAWINGS

[0105] Figure 1 This is the flow chart of the method;

[0106] Figure 2 It is a pipeline corrosion risk early warning indicator system. DETAILED DESCRIPTION

[0107] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0108] like Figure 1 As shown, the present invention provides a pipeline corrosion bilateral early warning method based on dynamic weight fusion, which is characterized by comprising the following steps:

[0109] S1: Collect multi-source corrosion monitoring data of the target oil and gas pipeline, and classify the monitoring data into a failure feature data set and a risk feature data set;

[0110] S2: Establishing a pipeline failure risk warning model based on the pipeline residual strength evaluation criteria, importing the failure feature data set into the failure warning model to perform pipeline failure warning, and outputting a pipeline corrosion failure score F1 according to the failure warning level;

[0111] S3: Using the DHGF algorithm to establish a pipeline corrosion risk warning model, input the risk feature data set into the risk warning model for comprehensive evaluation, and output the pipeline corrosion risk score F2;

[0112] S4: Obtain historical corrosion warning results of oil and gas pipelines, including the historical pipeline corrosion failure score F1 his , Pipeline Corrosion Risk Historical Score and corrosion bilateral early warning history score F his ;

[0113] S5: Establish a dynamic weight optimization model, import historical oil and gas pipeline corrosion warning results into the dynamic weight optimization model, and use the ELK optimization algorithm to solve the model to obtain the bilateral warning dynamic weight, which includes the pipeline failure warning weight α and the pipeline risk warning weight β;

[0114] S6: combining the bilateral warning dynamic weight, performing weighted calculation on the pipeline corrosion failure score and the pipeline corrosion risk score to obtain a pipeline corrosion bilateral warning score, and determining the pipeline corrosion bilateral warning level according to a preset corrosion bilateral warning score threshold;

[0115] The weighted calculation formula is as follows:

[0116] F=αF1+βF2

[0117] Where, F is the pipeline corrosion bilateral warning score; F1 is the pipeline corrosion failure score; F2 is the pipeline corrosion risk score; α is the pipeline failure warning weight; β is the pipeline risk warning weight;

[0118] The corresponding relationship between the corrosion bilateral warning score and the corrosion bilateral warning level is:

[0119] When F∈[0,30), the bilateral corrosion warning level is low; when F∈[30,50), the bilateral corrosion warning level is relatively low; when F∈[50,70), the bilateral corrosion warning level is medium; when F∈[70,85), the bilateral corrosion warning level is relatively high; when F∈[85,100], the bilateral corrosion warning level is high.

[0120] In one embodiment, the failure feature set described in step S1 includes pipeline design pressure, pipeline nominal diameter, pipeline nominal wall thickness, material yield strength, pipeline design coefficient, pipeline weld coefficient, pipeline service time, pipeline monitored corrosion rate, pipeline corrosion margin, and allowable residual strength factor;

[0121] The risk characteristic data set can be divided into numerical indicators and non-numerical indicators; non-numerical indicators can be divided into positive non-numerical indicators and negative non-numerical indicators, and divided into high, relatively high, medium, relatively low and low according to their degree. The risk value of positive non-numerical indicators is positively correlated with the degree, and the risk value of negative non-numerical indicators is negatively correlated with the degree;

[0122] Numerical indicators include pipeline monitoring corrosion rate, pipeline service life, pipeline gas pressure and patrol frequency;

[0123] Positive non-numerical indicators include pipeline medium hazard, construction defect level, pipe material defect level, soil corrosiveness, surrounding work conditions, natural disaster level, and population density around the pipeline;

[0124] Negative non-numerical indicators include the degree of internal corrosion protection, the strength of external corrosion protection, the degree of implementation of the management plan, and the degree of emergency plan formulation.

[0125] In one embodiment, the pipeline failure risk warning described in step S2 includes a two-level evaluation process, and the specific steps are as follows:

[0126] S201: Calculate the minimum required wall thickness δ min , the calculation formula is shown as follows;

[0127]

[0128] Where, δ min is the minimum required wall thickness of the pipeline; ψ is the pipeline design pressure; D is the nominal diameter of the pipeline; ε is the material yield strength; φ is the pipeline design coefficient; λ is the pipeline weld coefficient;

[0129] S202: Calculate the remaining wall thickness of the pipe δ mm , the calculation formula is shown as follows;

[0130] δ mm =δ-nc r

[0131] Where, δ mm is the remaining wall thickness of the pipeline; δ is the nominal wall thickness of the pipeline; n is the service time of the pipeline; c r Monitor corrosion rates for pipelines;

[0132] S203: Calculate the remaining wall thickness ratio R of the pipe t , the calculation formula is shown as follows;

[0133]

[0134] Where R t is the remaining wall thickness ratio of the pipe;

[0135] S204: When the remaining wall thickness of the pipe is greater than R t Less than the allowable residual strength factor RSF a , the shell parameter Q is calculated as follows; when the remaining wall thickness of the pipe is greater than R t Greater than the allowable residual strength factor RSF a , the shell parameter Q is 50;

[0136]

[0137] Calculate the pipe uniform thickness length L, the calculation formula is as follows:

[0138]

[0139] Where, L is the uniform thickness length of the pipe;

[0140] S205: Calculate the pipe's ultimate wall thickness δ lim ,, the calculation formula is as follows:

[0141] δ lim =max[0.2δ,2.5mm]

[0142] Where, δ lim is the pipe limit wall thickness;

[0143] S206: Execute the first-level warning process. If the first-level corrosion criterion is met, the pipeline corrosion failure warning is in an acceptable state. If the first-level corrosion criterion is not met, enter the second-level warning process. The first-level corrosion criterion expression is:

[0144] δ mm -FCA≥RSF a δ min

[0145] δ mm -FCA≥max[0.5δ min , δ lim ]

[0146] δ mm -FCA≥max[0.5δ min , δ lim ]

[0147] Where, FCA is the pipeline corrosion allowance;

[0148] S207: Execute the secondary warning process. If the secondary corrosion criterion is met, the pipeline corrosion failure warning is in a risk state. If the secondary corrosion criterion is not met, the pipeline corrosion failure warning is in an unsafe state. The secondary corrosion criterion expression is:

[0149] R t ≥0.2

[0150]

[0151] S208: Outputting the pipeline corrosion failure score F1 according to the failure warning level. When the pipeline failure warning is in an acceptable state, the pipeline corrosion failure score F1 is 30 points. When the pipeline failure warning is in a risky state, the pipeline corrosion failure score F1 is 60 points. When the pipeline failure warning is in an unsafe state, the pipeline corrosion failure score F1 is 90 points.

[0152] In one embodiment, the specific steps of step S3 are as follows:

[0153] S301: Constructing a pipeline corrosion risk early warning indicator system through the Delphi method, wherein the early warning indicator system is composed of a target layer, a criterion layer, and an indicator layer;

[0154] The target layer is pipeline corrosion risk; the criterion layer is pipeline conditions, emergency response capabilities, and environmental conditions; the indicator layer is all indicators in the risk characteristic dataset. Pipeline conditions include pipeline monitoring corrosion rate, pipeline service life, pipeline gas pressure, patrol frequency, pipeline medium hazard, internal corrosion protection level, external corrosion protection strength, and material defect level; emergency response capabilities include the degree of management plan implementation and the degree of emergency plan formulation; and environmental conditions include soil corrosivity, surrounding operating conditions, degree of natural disasters, and population density around the pipeline.

[0155] S302: Calculate the weights of pipeline corrosion risk warning indicators using the analytic hierarchy process; construct a judgment matrix A around the risk warning indicators and calculate the subjective weight vector of the indicators. The calculation formula is as follows:

[0156]

[0157] Where M i is the vector product of each row of the judgment matrix; a ij is the element in the judgment matrix; is the nth square root of the vector product; W i 1 is the subjective weight vector of the indicator;

[0158] Perform a consistency test on the judgment matrix A. If CR < 0.1, the judgment matrix A passes the consistency test. Otherwise, the judgment matrix A fails the consistency test and needs to be reconstructed. The consistency test formula is as follows:

[0159]

[0160] Where a i is the early warning indicator; i is the subjective weight of the indicator; max is the maximum eigenvalue of the judgment matrix; n is the order of the judgment matrix; CI is the consistency index; RI is the consistency ratio;

[0161] The subjective weights of the indicators are objectively corrected through the entropy weight method, and the comprehensive weight matrix W of the early warning indicators is constructed. The correction process is shown in the following formula:

[0162]

[0163]

[0164] Where r ij is the index proportion; s i is the index entropy value; W i 2 is the indicator information weight; W i is the comprehensive weight of the indicators;

[0165] S303: Construct an indicator sample matrix based on the expert evaluation results, determine the indicator warning levels as high, relatively high, medium, relatively low and low, and the corresponding level threshold set is D = {d n1 ,d n2 ,d n3 ,d n4}={9,7,5,3}, and establish the whitening weight function of each warning level, as shown in the following formula:

[0166]

[0167]

[0168] Where, d ni is the warning level threshold; f1(d ni ) is the measurement function corresponding to the high warning level; f2(d ni ) is the warning level comparison measurement function; f3(d ni ) is the measurement function corresponding to the medium warning level; f4(d ni ) is the measurement function corresponding to the lower warning level; f5(d ni ) is the measurement function corresponding to the low warning level;

[0169] S304: Obtain the membership degree of the early warning indicator under different gray categories according to the whitening weight function, calculate the gray statistical value of the early warning indicator, obtain the gray evaluation weight after normalization, and construct the gray fuzzy weight matrix. The process is shown in the following formula:

[0170]

[0171] Where h jk is the statistical value of the jth indicator belonging to the kth gray class; f k (d ij ) is the membership degree of the jth indicator to the kth gray class; r jk is the evaluation weight of the jth indicator belonging to the kth gray class; R is the gray fuzzy weight matrix;

[0172] S304: Construct a fuzzy comprehensive evaluation matrix P based on the early warning indicator comprehensive weight matrix W and the grey fuzzy weight matrix R. The expression is as follows:

[0173]

[0174] Determine the early warning indicator risk level set as V = {v1, v2, v3, v4, v5} = {10, 8, 6, 4, 2}. Calculate the pipeline corrosion risk score F2 based on the early warning indicator scoring vector. The calculation formula is as follows:

[0175]

[0176] Where p ij is the element in the fuzzy comprehensive evaluation matrix; v i is an element in the fuzzy comprehensive evaluation matrix.

[0177] In one embodiment, the specific steps of step S5 are as follows:

[0178] S501: Establish a dynamic weight optimization model, the model objective expression is as follows:

[0179]

[0180] α+β=1

[0181] α>0

[0182] β>0

[0183] Where, F1 his Score the pipeline corrosion failure history, F is the pipeline corrosion risk history score; his is the corrosion bilateral warning history score; α is the pipeline failure warning weight and β is the pipeline risk warning weight;

[0184] S502: Using the model objective function as the fitness function, randomly initialize and generate pipeline failure warning weight α and pipeline risk warning weight β as the initial elk population. The generation process expression is as follows:

[0185] X i =(U i -L i )r+L i

[0186] Where, X i is the initial elk population; U i To solve the upper bound of the space; L i To solve the lower bound of space;

[0187] S503: Set the initial parameters of the algorithm, including the maximum number of iterations T max , male elk proportion Br and population size N;

[0188] S504: Calculate the number B of male elk in the population based on the proportion of male elk. Calculate the fitness of each elk individual in the initial elk population using the fitness function and sort them. Define the first B elk individuals as male elk individuals. The expression is:

[0189] B=Br×N

[0190] S505: A roulette wheel selection mechanism is used to assign mates to male elk individuals, based on the proportion of the male elk individual fitness value in the total fitness value of the population. The expression is:

[0191]

[0192] Where p k The probability of selecting a mate for a male elk individual; f(x k ) is the individual fitness value of male elk;

[0193] S506: The male elk individual that has selected a mate enters the breeding stage and generates new elk individuals. The updating formula is:

[0194] x i (t+1)=x k (t)+α·(x k (t)-x j (t))

[0195] Where x i (t+1) is a newborn elk individual; x k (t) is a male elk individual; x j (t) is a female elk individual; α is the genetic coefficient, which is a random number in the range of [0,1];

[0196] S506: The male elk individuals of the selected female individuals enter the breeding stage, and are updated to generate new male elk individuals and new female elk individuals. The update formula for the new male elk individuals is:

[0197] x k (t+1)=x k (t)+α·(x k (t)-x j (t))

[0198] The individual update formula for newborn female elk is:

[0199] x j (t+1)=x k (t)+β·(x h (t)-x j (t))+γ(x r (t)-x j (t))

[0200] Where x k (t+1) is a newborn male elk individual; x k (t) is a male elk individual; x j (t) is a female elk individual; α is the male genetic coefficient, which is a random number in the range of [0,1]; x h(t) is a random male elk individual; x r (t) is a random female elk individual; β is the male genetic coefficient and γ is a random number in the range [0,1];

[0201] S507: Judgment t≥T max If so, the algorithm is terminated and the optimal pipeline failure warning weight α and pipeline risk warning weight β are output; otherwise, the algorithm returns to S506 to continue iteration.

[0202] Taking some gas production pipelines, gas gathering pipelines and gathering and transportation pipelines in a gas field gathering and transportation system as research examples, a bilateral early warning of pipeline corrosion based on dynamic weight fusion is conducted to verify the practicality of the method proposed in this application. The pipeline failure feature dataset is shown in Table 1, where the future corrosion margin FCA is 2.00 mm and the permissible corrosion intensity factor RSF is 2.00 mm. a Take 0.9, the design pressure ψ is 6.3MPa, and the design coefficient φ is 0.7.

[0203] Table 1 Pipeline failure feature dataset

[0204]

[0205] Taking the gas gathering pipeline G as an example, the pipeline failure warning process is introduced in detail. After calculation, the minimum wall thickness δ min is 4.2478 mm, and the remaining wall thickness of the pipe δ mm is 3.3708mm, and the remaining wall thickness is R t is 0.7935, the average thickness length L is 58.299 mm, and the pipe limit wall thickness δ lim 2.5mm.

[0206] S201: Calculate the required minimum wall thickness δ min

[0207]

[0208] S202: Calculate the remaining wall thickness of the pipe δ mm

[0209] δ mm =δ-nc r =6-12×0.2191=3.3708mm

[0210] S203: Calculate the remaining wall thickness ratio R of the pipe t

[0211]

[0212] S204: Calculate the uniform thickness length L

[0213] Since Rt <RSF a , the shell parameter Q is calculated using the following formula;

[0214]

[0215] Calculate the uniform thickness length of the pipe based on the shell parameter Q:

[0216]

[0217] S205: Calculate the ultimate pipe wall thickness

[0218] δ lim =max[0.2δ,2.5mm]=max[0.2×8mm,2.5mm]=2.5mm

[0219] S206: Executing the first-level warning process, the result shows that the gas gathering pipeline G does not meet the first-level corrosion criterion and needs to enter the second-level warning process;

[0220] δ mm -FCA=3.3708-2=1.3708≤0.9×4.42478=3.8231Not satisfied

[0221] δ mm -FCA=1.3708≤max[0.5δ min , δ lim ]=2.5mm does not meet

[0222] S207: Executing the secondary warning process, the result shows that the pipeline does not meet the secondary corrosion criterion, and the corrosion failure warning of the gas gathering pipeline G is in an unsafe state;

[0223] R t =0.7935≥0.2 satisfies

[0224]

[0225] S208: Outputting a pipeline corrosion failure score F1 according to the failure warning level. The failure warning of the gas gathering pipeline G is in an unsafe state, and the pipeline corrosion failure score F1 is 90 points.

[0226] Table 2 shows the corrosion failure warning results of some gas production pipelines, gas gathering pipelines, and gathering and transportation pipelines in a gas field’s gathering and transportation system.

[0227] Table 2 Pipeline corrosion failure warning results

[0228]

[0229] The pipeline condition data in the pipeline risk characteristic dataset are shown in Table 3, and the emergency response capabilities and environmental conditions in the pipeline risk characteristic dataset are shown in Table 4.

[0230] Table 3 Pipeline data

[0231]

[0232]

[0233] Table 4 Pipeline emergency capacity and environmental conditions data

[0234]

[0235] The pipeline corrosion risk early warning index system was constructed by the Delphi method, and the pipeline corrosion risk early warning index weights were calculated using the hierarchical analysis method. The judgment matrix composition and weight calculation results of each layer of the pipeline corrosion risk early warning index system are shown in Tables 5 to 8.

[0236] Table 5. Composition of the judgment matrix and weight calculation results of the criterion layer

[0237]

[0238] Table 6 Index layer - pipeline condition judgment matrix composition and weight calculation results

[0239]

[0240]

[0241] Table 7 Index layer - emergency capability judgment matrix composition and weight calculation results

[0242]

[0243] Table 8 Index layer-environmental condition judgment matrix composition and weight calculation results

[0244]

[0245] The pipeline corrosion risk warning score F2 calculated based on the DHGF integrated algorithm is shown in Table 9. As shown in Table 9, the corrosion risk warning scores of most pipelines are below 30, which is within a relatively safe range; the gas gathering pipeline D has the highest corrosion risk warning score, which is 70.474.

[0246] Table 9 Calculation results of pipeline corrosion risk warning score F2

[0247]

[0248] The dynamic weight optimization model was solved using the ELK optimization algorithm. The pipeline failure warning weight α was determined to be 0.753, and the pipeline risk warning weight β was determined to be 0.247. Combined with the bilateral warning dynamic weights, the pipeline corrosion failure score and pipeline corrosion risk score were weightedly calculated. The calculation results are shown in Table 10. The bilateral warning level of pipeline corrosion was determined based on the preset warning level interval threshold. Among them, the bilateral warning level of pipeline corrosion of gas gathering pipeline D was medium, and the bilateral warning level of pipeline corrosion of gathering and transmission pipeline G was high.

[0249] Table 10 Calculation results of pipeline bilateral corrosion warning score F

[0250]

[0251] The foregoing is merely an example of the embodiments of this specification and is not intended to limit the embodiments of this specification. For those skilled in the art, various modifications and variations of the embodiments of this specification are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of this specification shall be included within the scope of the claims of the embodiments of this specification.

Claims

1. A bilateral early warning method for pipeline corrosion based on dynamic weight fusion, characterized in that: The following steps are involved: S1: Collect multi-source corrosion monitoring data of the target oil and gas pipeline, and classify the monitoring data into a failure feature data set and a risk feature data set; S2: Establishing a pipeline failure risk warning model based on the pipeline residual strength evaluation criteria, importing the failure feature data set into the failure warning model to perform pipeline failure warning, and outputting a pipeline corrosion failure score F1 according to the failure warning level; S3: Using the DHGF algorithm to establish a pipeline corrosion risk warning model, input the risk feature data set into the risk warning model for comprehensive evaluation, and output the pipeline corrosion risk score F2; S4: Obtain historical corrosion warning results of oil and gas pipelines, including the historical pipeline corrosion failure score F1 his , Pipeline corrosion risk history score F2 his and corrosion bilateral early warning history score F his ; S5: Establish a dynamic weight optimization model, import historical oil and gas pipeline corrosion warning results into the dynamic weight optimization model, and use the ELK optimization algorithm to solve the model to obtain the bilateral warning dynamic weight, which includes the pipeline failure warning weight α and the pipeline risk warning weight β; S6: combining the bilateral warning dynamic weight, performing weighted calculation on the pipeline corrosion failure score and the pipeline corrosion risk score to obtain a pipeline corrosion bilateral warning score, and determining the pipeline corrosion bilateral warning level according to a preset corrosion bilateral warning score threshold; The weighted calculation formula is as follows: F=αF1+βF2 Where, F is the pipeline corrosion bilateral warning score; F1 is the pipeline corrosion failure score; F2 is the pipeline corrosion risk score; α is the pipeline failure warning weight; β is the pipeline risk warning weight; The corresponding relationship between the corrosion bilateral warning score and the corrosion bilateral warning level is: When F∈[0,30), the bilateral corrosion warning level is low; when F∈[30,50), the bilateral corrosion warning level is relatively low; when F∈[50,70), the bilateral corrosion warning level is medium; when F∈[70,85), the bilateral corrosion warning level is relatively high; when F∈[85,100], the bilateral corrosion warning level is high.

2. A bilateral early warning method for pipeline corrosion based on dynamic weight fusion according to claim 1, characterized in that: The failure feature set described in step S1 includes pipeline design pressure, pipeline nominal diameter, pipeline nominal wall thickness, material yield strength, pipeline design coefficient, pipeline weld coefficient, pipeline service time, pipeline monitoring corrosion rate, pipeline corrosion margin and allowable residual strength factor; The risk characteristic data set can be divided into numerical indicators and non-numerical indicators; non-numerical indicators can be divided into positive non-numerical indicators and negative non-numerical indicators, and divided into high, relatively high, medium, relatively low and low according to their degree. The risk value of positive non-numerical indicators is positively correlated with the degree, and the risk value of negative non-numerical indicators is negatively correlated with the degree; Numerical indicators include pipeline monitoring corrosion rate, pipeline service life, pipeline gas pressure and patrol frequency; Positive non-numerical indicators include pipeline medium hazard, construction defect level, pipe material defect level, soil corrosiveness, surrounding work conditions, natural disaster level, and population density around the pipeline; Negative non-numerical indicators include the degree of internal corrosion protection, the strength of external corrosion protection, the degree of implementation of the management plan, and the degree of emergency plan formulation.

3. The bilateral early warning method for pipeline corrosion based on dynamic weight fusion according to claim 1 is characterized in that: The pipeline failure risk warning described in step S2 includes a two-level evaluation process, and the specific steps are as follows: S201: Calculate the minimum required wall thickness δ min , the calculation formula is shown as follows; Where, δ min is the minimum required wall thickness of the pipeline; ψ is the pipeline design pressure; D is the nominal diameter of the pipeline; ε is the material yield strength; φ is the pipeline design coefficient; λ is the pipeline weld coefficient; S202: Calculate the remaining wall thickness of the pipe δ mm , the calculation formula is as follows; d mm =δ-nc r Where, δ mm is the remaining wall thickness of the pipeline; δ is the nominal wall thickness of the pipeline; n is the service time of the pipeline; c r Monitor corrosion rates for pipelines; S203: Calculate the remaining wall thickness ratio R of the pipe t , the calculation formula is shown as follows; Where R t is the remaining wall thickness ratio of the pipe; S204: When the remaining wall thickness of the pipe is greater than R t Less than the allowable residual strength factor RSF a , the shell parameter Q is calculated as follows; when the remaining wall thickness of the pipe is greater than R t Greater than the allowable residual strength factor RSF a , the shell parameter Q is 50; Calculate the pipe uniform thickness length L, the calculation formula is as follows: Where, L is the uniform thickness length of the pipe; S205: Calculate the pipe's ultimate wall thickness δ lim ,, the calculation formula is as follows: d lim =max[0.2δ,2.5mm] Where, δ lim is the pipe limit wall thickness; S206: Execute the first-level warning process. If the first-level corrosion criterion is met, the pipeline corrosion failure warning is in an acceptable state. If the first-level corrosion criterion is not met, enter the second-level warning process. The first-level corrosion criterion expression is: d mm -FCA≥RSF a d min d mm -FCA≥max[0.5δ min ,d lim ] d mm -FCA≥max[0.5δ min ,d lim ] Where, FCA is the pipeline corrosion allowance; S207: Execute the secondary warning process. If the secondary corrosion criterion is met, the pipeline corrosion failure warning is in a risk state. If the secondary corrosion criterion is not met, the pipeline corrosion failure warning is in an unsafe state. The secondary corrosion criterion expression is: R t ≥0.2 S208: Outputting the pipeline corrosion failure score F1 according to the failure warning level. When the pipeline failure warning is in an acceptable state, the pipeline corrosion failure score F1 is 30 points. When the pipeline failure warning is in a risky state, the pipeline corrosion failure score F1 is 60 points. When the pipeline failure warning is in an unsafe state, the pipeline corrosion failure score F1 is 90 points.

4. The bilateral early warning method for pipeline corrosion based on dynamic weight fusion according to claim 1 is characterized in that: The specific steps of step S3 are as follows: S301: Constructing a pipeline corrosion risk early warning indicator system through the Delphi method, wherein the early warning indicator system is composed of a target layer, a criterion layer, and an indicator layer; The target layer is pipeline corrosion risk; the criterion layer is pipeline conditions, emergency response capabilities, and environmental conditions; the indicator layer is all indicators in the risk characteristic dataset. Pipeline conditions include pipeline monitoring corrosion rate, pipeline service life, pipeline gas pressure, patrol frequency, pipeline medium hazard, internal corrosion protection level, external corrosion protection strength, and material defect level; emergency response capabilities include the degree of management plan implementation and the degree of emergency plan formulation; and environmental conditions include soil corrosivity, surrounding operating conditions, degree of natural disasters, and population density around the pipeline. S302: Calculate the weights of pipeline corrosion risk warning indicators using the analytic hierarchy process; construct a judgment matrix A around the risk warning indicators and calculate the subjective weight vector of the indicators. The calculation formula is as follows: Where M i is the vector product of each row of the judgment matrix; a ij is the element in the judgment matrix; is the nth square root of the vector product; W i 1 is the subjective weight vector of the indicator; Perform a consistency test on the judgment matrix A. If CR < 0.1, the judgment matrix A passes the consistency test. Otherwise, the judgment matrix A fails the consistency test and needs to be reconstructed. The consistency test formula is as follows: Where a i is the early warning indicator; i is the subjective weight of the indicator; max is the maximum eigenvalue of the judgment matrix; n is the order of the judgment matrix; CI is the consistency index; RI is the consistency ratio; The subjective weights of the indicators are objectively corrected through the entropy weight method, and the comprehensive weight matrix W of the early warning indicators is constructed. The correction process is shown in the following formula: Where r ij is the index proportion; s i is the index entropy value; W i 2 is the indicator information weight; W i is the comprehensive weight of the indicators; S303: Construct an indicator sample matrix based on the expert evaluation results, determine the indicator warning levels as high, relatively high, medium, relatively low and low, and the corresponding level threshold set is D = {d n1 ,d n2 ,d n3 ,d n4 }={9,7,5,3}, and establish the whitening weight function of each warning level, as shown in the following formula: Where, d ni is the warning level threshold; f1(d ni ) is the measurement function corresponding to the high warning level; f2(d ni ) is the warning level comparison measurement function; f3(d ni ) is the measurement function corresponding to the medium warning level; f4(d ni ) is the measurement function corresponding to the lower warning level; f5(d ni ) is the measurement function corresponding to the low warning level; S304: Obtain the membership degree of the early warning indicator under different gray categories according to the whitening weight function, calculate the gray statistical value of the early warning indicator, obtain the gray evaluation weight after normalization, and construct the gray fuzzy weight matrix. The process is shown in the following formula: Where h jk is the statistical value of the jth indicator belonging to the kth gray class; f k (d ij ) is the membership degree of the jth indicator to the kth gray class; r jk is the evaluation weight of the jth indicator belonging to the kth gray class; R is the gray fuzzy weight matrix; S304: Construct a fuzzy comprehensive evaluation matrix P based on the early warning indicator comprehensive weight matrix W and the grey fuzzy weight matrix R. The expression is as follows: Determine the early warning indicator risk level set as V = {v1, v2, v3, v4, v5} = {10, 8, 6, 4, 2}. Calculate the pipeline corrosion risk score F2 based on the early warning indicator scoring vector. The calculation formula is as follows: Where p ij is the element in the fuzzy comprehensive evaluation matrix; v i is an element in the fuzzy comprehensive evaluation matrix.

5. The bilateral early warning method for pipeline corrosion based on dynamic weight fusion according to claim 1 is characterized in that: The specific steps of step S5 are as follows: S501: Establish a dynamic weight optimization model, the model objective expression is as follows: Where, F1 his Pipeline corrosion failure history score, F2 his F is the pipeline corrosion risk history score; his is the corrosion bilateral warning history score; α is the pipeline failure warning weight and β is the pipeline risk warning weight; S502: Using the model objective function as the fitness function, randomly initialize and generate pipeline failure warning weight α and pipeline risk warning weight β as the initial elk population. The generation process expression is as follows: X i =(U i -L i )r+L i Where, X i is the initial elk population; U i To solve the upper bound of the space; L i To find the lower bound of the space; S503: Set the initial parameters of the algorithm, including the maximum number of iterations T max , male elk proportion Br and population size N; S504: Calculate the number B of male elk in the population based on the proportion of male elk. Calculate the fitness of each elk individual in the initial elk population using the fitness function and sort them. Define the first B elk individuals as male elk individuals. The expression is: B=Br×N S505: A roulette wheel selection mechanism is used to assign mates to male elk individuals, based on the proportion of the male elk individual fitness value in the total fitness value of the population. The expression is: Where p k The probability of selecting a mate for a male elk individual; f(x k ) is the individual fitness value of male elk; S506: The male elk individual that has selected a mate enters the breeding stage and generates new elk individuals. The updating formula is: x i (t+1)=x k (t)+α·(x k (t)-x j (t)) Where x i (t+1) is a newborn elk individual; x k (t) is a male elk individual; x j (t) is a female elk individual; α is the genetic coefficient, which is a random number in the range of [0,1]; S506: The male elk individuals of the selected female individuals enter the breeding stage, and are updated to generate new male elk individuals and new female elk individuals. The update formula for the new male elk individuals is: x k (t+1)=x k (t)+α·(x k (t)-x j (t)) The individual update formula for newborn female elk is: x j (t+1)=x k (t)+β·(x h (t)-x j (t))+γ(x r (t)-x j (t)) Where x k (t+1) is a newborn male elk individual; x k (t) is a male elk individual; x j (t) is a female elk individual; α is the male genetic coefficient, which is a random number in the range of [0,1]; x h (t) is a random male elk individual; x r (t) is a random female elk individual; β is the male genetic coefficient and γ is a random number in the range [0,1]; S507: Judgment t≥T max If so, the algorithm is terminated and the optimal pipeline failure warning weight α and pipeline risk warning weight β are output; otherwise, the algorithm returns to S506 to continue iteration.

Images