Process pipeline fatigue damage identification and prediction method

Abstract

The invention relates to the technical field of offshore oil safety assessment, in particular to a process pipeline fatigue damage identification and prediction method, which comprises the following steps: step 1, key pipeline primary selection and failure possibility assessment; 2, collecting and preprocessing multi-modal data; 3, performing model construction and optimization on damage identification; 4, fatigue life prediction and failure early warning are carried out; 5, carrying out correlation analysis on the stress spectrum and the service life; 6, multi-parameter dynamic early warning is carried out; and 7, integrating a real-time monitoring system. According to the method, through multi-modal evaluation of fluid-to-vibration, fluid pulsation, high-frequency acoustic vibration and the like, a stress spectrum is generated by utilizing a rain-flow counting method, the crack growth rate and the residual life are predicted in combination with an S-N curve or a Paris formula, and through correlation analysis of the stress spectrum and the life and multi-parameter dynamic early warning, the crack growth rate and the residual life are predicted. Therefore, the method has the advantages that an online system integrating vibration, acoustic emission and stress monitoring is provided, and real-time evaluation and life prediction of fatigue damage are achieved.

Description

Technical Field

[0001] The invention relates to the technical field of offshore oil safety assessment, and in particular to a method for identifying and predicting fatigue damage of process pipelines. Background Art

[0002] The process media of offshore oil platforms are all flammable and explosive hydrocarbons, and the process has a certain pressure, even high pressure. Once a leak occurs, it will cause a large amount of hydrocarbons to leak, causing fire and explosion and other accident risks. Offshore oil platforms that have reached their design lifespan usually still serve beyond their service life, and life extension assessments or regular special inspections do not involve the inspection and evaluation of process pipelines. The strength and structure of key process pipelines will be weakened or damaged under long-term oil and gas corrosion, vibration effects, and time effects, which lays hidden dangers for the failure and leakage of key process pipelines.

[0003] It is difficult to achieve comprehensive coverage of process pipeline fatigue damage by relying on nondestructive testing technology. The main reason is that the pipeline configuration is complex, the position is variable, and it is impossible to combine the influencing loads for predictive analysis. Therefore, a process pipeline fatigue damage identification and prediction method is needed. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the deficiencies in the prior art, the present invention provides a method for identifying and predicting fatigue damage of process pipelines. It has an online system that integrates vibration, acoustic emission, and stress monitoring, and can achieve real-time assessment of fatigue damage and life prediction, thereby solving the above-mentioned problems.

[0006] (II) Technical solution

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The fatigue damage identification and prediction method of process pipeline includes the following steps:

[0009] Step 1: Preliminary selection of key pipelines and failure possibility assessment;

[0010] Step 2: Multimodal data collection and preprocessing;

[0011] Step 3: Model building and optimization of damage identification;

[0012] Step 4: Fatigue life prediction and failure warning;

[0013] Step 5: stress spectrum and life correlation analysis;

[0014] Step 6: Multi-parameter dynamic warning;

[0015] Step 7: Real-time monitoring system integration.

[0016] Preferably, Step 1 includes multi-dimensional failure possibility analysis and identification of dynamic strength critical points. Among them, the multi-dimensional failure possibility analysis: Through multi-modal evaluations such as fluid-to-vibration, fluid pulsation, and high-frequency acoustic vibration, calculate the failure possibility coefficients of each process pipeline, focus on stress concentration areas, parts with high vibration frequencies, and corrosion environment-sensitive parts, and determine the weak links as the monitoring objects in combination with preset criteria; Identification of dynamic strength critical points: Based on the finite element model, perform modal analysis, calculate the stress response in combination with random vibration excitation, and identify the dynamic strength critical points (such as welds, elbows, etc.) of the pipeline structure to provide the target area for subsequent monitoring.

[0017] Preferably, Step 2 includes vibration and acoustic emission signal acquisition, as well as data fusion and noise reduction. Among them, vibration and acoustic emission signal acquisition includes vibration signals: Obtain pipeline vibration data through acceleration sensors, and extract vibration modal parameters (such as natural frequency, vibration mode) based on modal theory; Acoustic emission signals: Use high-frequency sensors to capture the transient elastic waves generated by crack propagation, and extract characteristic parameters (such as energy, frequency, count) in combination with wavelet transform and empirical mode decomposition (EMD); Data fusion and noise reduction: Perform preprocessing such as filtering and downsampling on the original signals to eliminate environmental noise interference and form a standardized input vector (such as vibration modal parameters + acoustic emission feature vectors).

[0018] Preferably, Step 3 includes a multi-modal fusion recognition model and prestress correction and fatigue damage calculation. Among them, the multi-modal fusion recognition model includes vibration signal processing, acoustic emission signal processing, and model fusion. Vibration signal processing: Input into the BP neural network model to locate the damage location and degree; Acoustic emission signal processing: Use the PNN neural network to compare the pattern samples, and determine crack propagation or noise interference in combination with the signal time difference; Model fusion: Take the vibration and acoustic emission fusion vector as the input and output the comprehensive damage recognition result to improve robustness; Prestress correction and fatigue damage calculation: Correct the initial prestress value according to the material yield strength and vibration environment, calculate the peak crossing rate of the random response signal and the stress amplitude probability density function, compile the stress spectrum in combination with the corrected prestress and the rain flow counting method, and use the linear damage accumulation theory (Miner criterion) to evaluate the fatigue damage.

[0019] Preferably, Step 5 obtains the stress level of the key area through a stress-strain measuring device, generates a stress spectrum using the rain flow counting method, and predicts the crack propagation rate and remaining life in combination with the S-N curve or Paris formula.

[0020] Preferably, the sixth step includes acoustic emission parameter threshold warning: when a turning point appears in the cumulative count, energy or amplitude curve, it indicates that the crack enters the rapid propagation stage, triggering a warning and a determination of the damage accumulation threshold: if the corrected fatigue damage value exceeds a preset threshold (such as D>0.9), it is determined that immediate maintenance is required.

[0021] (III) Beneficial effects

[0022] Compared with the prior art, the present invention provides a method for identifying and predicting fatigue damage of process pipelines, having the following beneficial effects:

[0023] 1. Through multi-modal evaluations such as fluid-to-vibration, fluid pulsation, and high-frequency acoustic vibration, the present invention calculates the failure probability coefficients of each process pipeline, focuses on stress concentration areas, high-vibration-frequency areas, and corrosion-environment-sensitive parts, and determines the weak links as the monitoring objects in combination with preset criteria; identification of dynamic strength dangerous points: based on a finite element model for modal analysis, combined with random vibration excitation to calculate the stress response, identify the dynamic strength dangerous points of the pipeline structure (such as welds, elbows, etc.), provide the target area for subsequent monitoring, obtain the pipeline vibration data through an acceleration sensor, and extract vibration modal parameters based on modal theory, such as natural frequency and vibration mode, acoustic emission signal: capture the transient elastic waves generated by crack propagation using a high-frequency sensor, and extract characteristic parameters (such as energy, frequency, count) in combination with wavelet transform and empirical mode decomposition; data fusion and noise reduction: perform preprocessing such as filtering and downsampling on the original signal to eliminate environmental noise interference, form a standardized input vector, and then through a multi-modal fusion recognition model and prestress correction and fatigue damage calculation, correct the initial prestress value according to the material yield strength and vibration environment, calculate the peak crossing rate of the random response signal and the stress amplitude probability density function, compile the stress spectrum in combination with the corrected prestress and the rainflow counting method, evaluate the fatigue damage using the linear damage accumulation theory, obtain the stress level of the key area through a stress-strain measuring device, generate the stress spectrum using the rainflow counting method, predict the crack propagation rate and remaining life in combination with the S-N curve or Paris formula, and through stress spectrum-life correlation analysis and multi-parameter dynamic warning, so that this method has the advantages of an online system integrating vibration, acoustic emission, and stress monitoring, realizing real-time evaluation of fatigue damage and life prediction.

[0024] 2. In the present invention, vibration signal processing: input into a BP neural network model to locate the damage position and degree; acoustic emission signal processing: use a PNN neural network to compare the pattern samples, and combine the signal time difference to determine crack propagation or noise interference; model fusion: use the vibration and acoustic emission fusion vector as the input, and output the comprehensive damage recognition result to improve the robustness. Specific embodiments

[0025] The embodiments of the present invention will be described below, and the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] The present invention relates to a method for identifying and predicting fatigue damage of process pipelines, including the following steps:

[0027] Step 1: Preliminary selection of key pipelines and assessment of failure possibility;

[0028] Step 1 includes multi-dimensional failure possibility analysis and identification of dynamic strength critical points. Among them, multi-dimensional failure possibility analysis: Through multi-modal evaluations such as fluid-to-vibration, fluid pulsation, and high-frequency acoustic vibration, calculate the failure possibility coefficients of each process pipeline, focus on stress concentration areas, parts with high vibration frequencies, and corrosion environment-sensitive parts, and determine the weak links as the monitoring objects in combination with preset criteria; Identification of dynamic strength critical points: Based on the finite element model, perform modal analysis, calculate the stress response in combination with random vibration excitation, and identify the dynamic strength critical points (such as welds, elbows, etc.) of the pipeline structure to provide the target area for subsequent monitoring;

[0029] Step 2: Multi-modal data acquisition and preprocessing;

[0030] Step 2 includes vibration and acoustic emission signal acquisition, as well as data fusion and noise reduction. Among them, vibration and acoustic emission signal acquisition includes vibration signals: Obtain pipeline vibration data through acceleration sensors, and extract vibration modal parameters (such as natural frequencies, vibration modes) based on modal theory. Acoustic emission signals: Use high-frequency sensors to capture transient elastic waves generated by crack propagation, and extract characteristic parameters (such as energy, frequency, count) in combination with wavelet transform and empirical mode decomposition (EMD); Data fusion and noise reduction: Perform preprocessing such as filtering and downsampling on the original signals to eliminate environmental noise interference and form a standardized input vector (such as vibration modal parameters + acoustic emission feature vectors);

[0031] Step 3: Construct and optimize the model for damage identification;

[0032] Step 3 includes a multi-modal fusion recognition model and prestress correction and fatigue damage calculation. The multi-modal fusion recognition model includes vibration signal processing, acoustic emission signal processing, and model fusion. Vibration signal processing: Input into the BP neural network model to locate the damage position and degree. Acoustic emission signal processing: Use the PNN neural network to compare the pattern samples and determine crack propagation or noise interference in combination with the signal time difference. Model fusion: Use the vibration and acoustic emission fusion vector as the input and output the comprehensive damage recognition result to improve robustness. Prestress correction and fatigue damage calculation: Correct the initial prestress value according to the material yield strength and vibration environment, calculate the peak crossing rate of the random response signal and the stress amplitude probability density function, compile the stress spectrum in combination with the corrected prestress and the rain flow counting method, and evaluate the fatigue damage using the linear damage accumulation theory (Miner criterion).

[0033] Step 4: Fatigue life prediction and failure warning;

[0034] Step 5: Stress spectrum and life correlation analysis. Obtain the stress level in the key area through the stress-strain measurement device, generate the stress spectrum using the rain flow counting method, and predict the crack propagation rate and remaining life in combination with the S-N curve or Paris formula;

[0035] Step 6: Multi-parameter dynamic warning, including acoustic emission parameter threshold warning: When turning points appear in the cumulative count, energy, or amplitude curve, it indicates that the crack enters the rapid propagation stage and triggers a warning, and damage accumulation threshold determination: If the corrected fatigue damage value exceeds the preset threshold (such as D>0.9), it is determined that immediate maintenance is required;

[0036] Step 7: Real-time monitoring system integration.

[0037] The beneficial effects of the present invention are as follows: Through multi-modal evaluations such as fluid-to-vibration, fluid pulsation, and high-frequency acoustic vibration, the present invention calculates the failure probability coefficients of each process pipeline, focuses on stress concentration areas, high-vibration-frequency areas, and corrosion-environment-sensitive parts, and determines the weak links as the monitoring objects in combination with preset criteria; Identification of dynamic strength dangerous points: Based on the finite element model, modal analysis is carried out, and the stress response is calculated in combination with random vibration excitation to identify the dynamic strength dangerous points (such as welds, elbows, etc.) of the pipeline structure, providing the target area for subsequent monitoring. The pipeline vibration data is obtained through acceleration sensors, and vibration modal parameters such as natural frequency and vibration mode are extracted based on modal theory; Acoustic emission signal: Use high-frequency sensors to capture the transient elastic waves generated by crack propagation, and extract characteristic parameters (such as energy, frequency, count) in combination with wavelet transform and empirical mode decomposition; Data fusion and noise reduction: The original signal is preprocessed such as filtering and downsampling to eliminate environmental noise interference, forming a standardized input vector. Then, through the multi-modal fusion recognition model and prestress correction and fatigue damage calculation, the initial prestress value is corrected according to the material yield strength and vibration environment, the peak crossing rate of the random response signal and the stress amplitude probability density function are calculated, and the stress spectrum is compiled in combination with the corrected prestress and the rain flow counting method. The linear damage accumulation theory is used to evaluate the fatigue damage, the stress level of the key area is obtained through the stress-strain measuring device, the stress spectrum is generated using the rain flow counting method, and the crack propagation rate and remaining life are predicted in combination with the S-N curve or Paris formula. And through the stress spectrum-life correlation analysis and multi-parameter dynamic early warning.

[0038] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for identifying and predicting fatigue damage of process pipelines, characterized in that, The following steps are involved: Step 1: Preliminary selection of key pipelines and failure possibility assessment; Step 2: Multimodal data collection and preprocessing; Step 3: Model building and optimization of damage identification; Step 4: Fatigue life prediction and failure warning; Step 5: stress spectrum and life correlation analysis; Step 6: Multi-parameter dynamic warning; Step 7: Real-time monitoring system integration.

2. The method for identifying and predicting fatigue damage of process pipelines according to claim 1, wherein: The step one includes multi-dimensional failure possibility analysis and dynamic strength danger point identification.

3. The method for identifying and predicting fatigue damage of process pipelines according to claim 1, wherein: The step 2 includes vibration and acoustic emission signal acquisition and data fusion and noise reduction, wherein the vibration and acoustic emission signal acquisition includes vibration signal: obtaining pipeline vibration data through an acceleration sensor, extracting vibration modal parameters (such as natural frequency, vibration type) based on modal theory, and acoustic emission signal: using a high-frequency sensor to capture transient elastic waves generated by crack expansion, and combining wavelet transform and empirical mode decomposition to extract characteristic parameters (such as energy, frequency, count).

4. The method for identifying and predicting the fatigue damage of a process pipeline according to claim 1, wherein: The step three includes a multimodal fusion identification model and prestress correction and fatigue damage calculation, wherein the multimodal fusion identification model includes vibration signal processing, acoustic emission signal processing and model fusion; prestress correction and fatigue damage calculation: correct the initial prestress value according to the material yield strength and the vibration environment, calculate the peak penetration rate and stress amplitude probability density function of the random response signal, compile the stress spectrum in combination with the corrected prestress and rain flow counting method, and use the linear damage accumulation theory to evaluate fatigue damage.

5. The process pipeline fatigue damage identification and prediction method according to claim 1, characterized in that: The fifth step is to obtain the stress level of the key area by using a stress-strain measuring device, generate a stress spectrum using a rain flow counting method, and predict the crack growth rate and remaining life in combination with the SN curve or the Paris formula.

6. The method for identifying and predicting fatigue damage of process pipelines according to claim 1, characterized in that: The step six includes an acoustic emission parameter threshold warning: when a turning point appears in the cumulative count, energy or amplitude curve, it indicates that the crack has entered a rapid expansion stage, triggering a warning and damage accumulation threshold determination: if the corrected fatigue damage value exceeds the preset threshold, it is determined that immediate maintenance is required.