Pipeline fatigue crack propagation behavior prediction method and system based on acoustic emission sensing

Abstract

The invention provides a pipeline fatigue crack propagation behavior prediction method and system based on acoustic emission sensing. The method comprises the following steps: constructing a prediction initial model based on an obtained modeling parameter # imgabs0 # and a fatigue crack propagation rate # imgabs1 #; optimizing model parameters of the prediction initial model by adopting a Bayesian statistical method and a Markov chain Monte Carlo method to obtain optimal model parameters, and adjusting the prediction initial model based on the optimal model parameters to obtain a prediction model; and analyzing and predicting the acquired sample data based on the prediction model to obtain a prediction result. According to the method, the initial prediction model is constructed by integrating the modeling parameters and the fatigue crack growth rate, the initial model is optimized by using the Bayesian statistical method and the Markov chain Monte Carlo method, the prediction model is obtained, the prediction model can effectively predict behavior trends of cracks in different stages, and the prediction accuracy is improved. Potential safety hazards can be found in advance, and the overall safety and reliability of a pipeline system are remarkably improved.

Description

Technical Field

[0001] The invention belongs to the field of fatigue crack acoustic emission monitoring, and in particular to a pipeline fatigue crack propagation behavior prediction method and system based on acoustic emission sensing. Background Art

[0002] Long-distance oil and natural gas pipelines are key infrastructure in the energy transportation system. At present, most of the long-distance oil and gas pipelines that have been built and are under construction in my country use high-grade pipeline steel of grade X65 or above. Among them, X80 pipeline steel has been widely used in my country's oil and gas pipeline projects due to its advantages such as high strength and good corrosion resistance. With the rapid expansion of my country's pipeline network and the gradual aging of pipeline material performance, safety accidents caused by pipeline failure occur frequently.

[0003] After analyzing the pipelines after the safety accident, it was concluded that fatigue cracks are one of the main failure forms of oil and gas pipelines. That is, during the use of oil and gas pipelines, they need to frequently bear alternating loads, which will cause fatigue cracks on the pipelines. The initiation of these cracks will reduce the bearing area of ​​the inner wall of the pipeline. In severe cases, sharp notches will be formed on the inner wall of the pipeline, which will lead to areas with highly concentrated stress inside the pipeline. Under the action of cyclic loads, pressure fluctuations, corrosion, high temperature and impact and other harsh conditions, cracks are prone to occur on the inner wall of the pipeline, leading to serious accidents such as oil and gas pipeline leakage. Therefore, timely grasping the evolution of pipeline fatigue cracks and making timely responses before pipeline failure is a major demand in the field of oil and gas pipeline safety assurance.

[0004] The current mainstream pipeline inspection technologies include eddy current inspection and ultrasonic inspection. Eddy current inspection uses a coil to generate an alternating magnetic field. When there are cracks on the surface of the material, the eddy current path and intensity induced by the magnetic field will change, thereby determining the location of the defect. However, it cannot be used to detect conductive materials, and has poor detection effect on deep cracks or damage. Ultrasonic inspection detects the internal structure or defects of the pipeline through the propagation of sound waves in the pipeline. It cannot detect pipelines with complex shapes or multi-layer structures. At the same time, although the above two methods can detect pipelines, they can only detect whether there are defects in the pipeline, and cannot specifically determine the fatigue crack growth behavior process of the pipeline and the transition points between key expansion stages, thereby inaccurately characterizing the existing defects of the pipeline, resulting in poor failure handling. Summary of the invention

[0005] In order to solve the problem that the existing pipeline detection technology can only detect whether there are defects in the pipeline, but cannot specifically determine the fatigue crack growth behavior process of the pipeline and the transition points between key growth stages, thus leading to inaccurate failure treatment, the present invention proposes a pipeline fatigue crack growth behavior prediction method and system based on acoustic emission sensing. This method can accurately identify the growth behavior of pipeline fatigue cracks and predict them, thereby making pipeline damage treatment more accurate and significantly improving the ability to prevent failures.

[0006] To achieve the above object, the present invention provides the following technical solutions: The present invention proposes a pipeline fatigue crack growth behavior prediction method based on acoustic emission sensing, which includes the following steps: Based on the obtained modeling parameters and fatigue crack growth rate Construct an initial prediction model; The initial prediction model is optimized by using Bayesian statistical method and Markov chain Monte Carlo method to obtain optimal model parameters, and the initial prediction model is adjusted based on the optimal model parameters to obtain a prediction model; The acquired sample data is analyzed and predicted based on the prediction model to obtain a prediction result.

[0007] Preferably, the modeling parameters The acquisition includes: Obtain the time domain waveform flow signal generated by the natural gas pipeline sample during fatigue test; intercepting the time domain waveform stream signal based on a preset interception condition to obtain an acoustic emission signal; Extracting characteristic parameters of the acoustic emission signal based on the acoustic emission signal; The modeling parameters are calculated based on the characteristic parameters of the acoustic emission signal .

[0008] Preferably, the interception condition is: peak definition time PDT300 , the impact definition time HDT is 600 , the impact lock time HLT is 1000 , the maximum duration is 1000ms and the fixed threshold is 50mv.

[0009] Preferably, extracting characteristic parameters of the acoustic emission signal based on the acoustic emission signal includes: Performing bandpass filtering on the acoustic emission signal to obtain a de-noised emission signal; Feature extraction is performed on the de-noised emission signal to obtain characteristic parameters of the acoustic emission signal.

[0010] Preferably, the modeling parameters are calculated based on the characteristic parameters of the acoustic emission signal. ,include Get the actual number of amplitude values ​​within multiple fixed time intervals; Summing the plurality of amplitude values ​​to obtain a total amplitude value; The average amplitude value is calculated based on the actual number of amplitude values ​​and the total value of the amplitude values ; The number of unit cycle loading cycles is calculated based on the number of cycles in the fixed time interval and the fixed time interval ; Based on the average amplitude value and the number of loading cycles per unit period Calculate the modeling parameters ; The calculation process of the modeling parameters is:

[0011] in, is a fixed time interval, T is the number of cycles in the fixed time interval, represents the sum of the amplitude values ​​of the acoustic emission signals detected in a fixed time interval, and k represents the actual number (in cycles) of the amplitude values ​​of the acoustic emission signals detected in the Nth fixed time interval.

[0012] Preferably, the fatigue crack growth rate The acquisition includes: Obtain the crack length and cycle number of the natural gas pipeline sample during the test; The fatigue crack growth rate is calculated based on the crack length and the number of cycles. .

[0013] Preferably, the Bayesian statistical method and the Markov chain Monte Carlo method are used to optimize the initial prediction model to obtain the optimal model parameters, including: Determining the prior distribution of the model parameters through experimental data and domain knowledge; The prior distribution of the model parameters and the conditional probability distribution of the observed data are constructed using a Bayesian inference method to obtain a posterior distribution; The posterior distribution is sampled and solved based on the Markov Monte Carlo method to obtain the optimal model parameters.

[0014] Preferably, the expression of the posterior distribution is:

[0015] in, is the set of model parameters, x is the observed data obtained after fatigue tests under different loading conditions, is the prior distribution, is the likelihood function; The conditional probability distribution of the observed data is a likelihood function of the observed data constructed based on the observed data obtained after fatigue tests under different loading conditions; The expression of the likelihood function is:

[0016] in, is the first model parameter, is the second model parameter, is the error term The standard deviation of .

[0017] Preferably, the prediction model is:

[0018] in, is the crack growth rate, is the first optimal model parameter, is the second optimal model parameter, is the error term, which follows a normal distribution, that is, , which represents the deviation between the model prediction and the actual observation.

[0019] The present invention proposes a pipeline fatigue crack growth behavior prediction system based on acoustic emission sensing, which is used to implement the above method, including: The model unit is used to obtain modeling parameters based on and fatigue crack growth rate Construct an initial prediction model; The processing unit is used to optimize the model parameters of the initial prediction model by using the Bayesian statistical method and the Markov chain Monte Carlo method to obtain the optimal model parameters, and adjust the initial prediction model based on the optimal model parameters to obtain the prediction model; The prediction unit is used to analyze and predict the acquired prediction sample data based on the prediction model to obtain a prediction result; The output unit is used to output the prediction results.

[0020] Compared with the prior art, the present invention has the following beneficial technical effects: The present invention discloses a method for predicting pipeline fatigue crack propagation behavior based on acoustic emission sensing. The method constructs an accurate initial prediction model by integrating modeling parameters and fatigue crack propagation rate, makes full use of the advantages of Bayesian statistical method and Markov chain Monte Carlo method, deeply optimizes the model parameters of the initial prediction model, and obtains the optimal model parameters. The optimal parameters not only improve the prediction accuracy of the model, but also enhance its adaptability to complex crack propagation behavior. By adjusting the initial prediction model with the optimal model parameters, a more reliable prediction model is obtained. The prediction model can not only accurately capture the subtle changes in crack propagation, but also effectively predict the behavior trends of cracks at different stages, including the transition points between key propagation stages. Compared with traditional methods, the prediction results of the method are more accurate, which is helpful to discover potential safety hazards in advance and significantly improve the overall safety and reliability of the pipeline system. In addition, the method also has high flexibility and adaptability, and can be widely used in the prediction of fatigue crack propagation behavior of various pipeline systems.

[0021] Furthermore, the method effectively removes noise interference by performing bandpass filtering on the acoustic emission signal, obtains a purer denoised emission signal, and ensures the accuracy and reliability of the characteristic parameters. In the feature extraction stage, the characteristic parameters of the acoustic emission signal are extracted, that is, the amplitude value is proposed, and the actual number of amplitude values ​​in multiple fixed time intervals is obtained, and the total amplitude value is obtained by summing these amplitude values, and the average amplitude value is calculated. Combined with the number of cycles in the fixed time interval and the fixed time interval itself, the number of loading cycles per unit period is obtained, and finally the modeling parameters are obtained, so that the initial prediction model constructed by the modeling parameters can predict the fatigue crack growth behavior of the pipeline with high precision, enhance the model's description ability of complex crack growth behavior, help reduce the risk of pipeline failure, and improve overall safety.

[0022] Furthermore, this method obtains the crack length and number of cycles generated in the natural gas pipeline sample during the test, and uses these data to calculate the fatigue crack growth rate, thereby ensuring the accuracy and reliability of the crack growth rate. When optimizing the initial prediction model, the prior distribution of the model parameters is determined by the test data, and the prior information is fully utilized to improve the initial estimation accuracy of the model parameters. The Bayesian inference method is used to combine the prior distribution of the model parameters with the conditional probability distribution of the observed data to construct the posterior distribution of the model parameters, thereby realizing the extraction of information from the observed data, updating and optimizing the model parameters, and sampling and solving the posterior distribution based on the Markov Monte Carlo method to obtain the optimal model parameters, thereby improving the efficiency of solving the model parameters and ensuring the accuracy and reliability of the model parameters. Through the above steps, a more accurate prediction model is obtained, which can more accurately describe the crack growth behavior, making the predicted pipeline fatigue crack growth behavior more accurate and reliable.

[0023] The present invention discloses a pipeline fatigue crack growth behavior prediction system based on acoustic emission sensing. The system constructs a prediction initial model through a model unit, uses a processing unit to combine Bayesian statistics and Markov chain Monte Carlo method for deep optimization, obtains optimal model parameters, and adjusts the prediction model accordingly. The prediction unit then efficiently analyzes sample data based on the model and outputs accurate prediction results, thereby improving prediction accuracy, significantly enhancing prediction efficiency and practicality, and helping to reduce the risk of pipeline failure and ensure overall safety. The system has a clear structure, is easy to operate and maintain, has good expansibility and adaptability, and can be widely used in the prediction of fatigue crack growth behavior of various pipeline systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A flow chart of a method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing is proposed for the present invention; Figure 2 The present invention proposes a linear correlation diagram between crack length and loading cycle number under different stress ratios in a pipeline fatigue crack growth behavior prediction method based on acoustic emission sensing; Figure 3 A linear correlation diagram of fatigue crack growth rate (FCPR) and stress intensity factor range under different stress ratios in the embodiment of the present invention; Figure 4 A linear correlation diagram between modeling parameters and fatigue crack growth rate in an embodiment of the present invention; Figure 5 A result comparison diagram of the fatigue crack growth rate prediction value and the actual value under different stress ratios in the embodiment proposed by the present invention; Figure 6This is a result comparison diagram of the predicted value and the actual value of fatigue crack length under different stress ratios in the embodiment proposed by the present invention. DETAILED DESCRIPTION

[0025] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the present invention. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0026] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0027] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0028] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, an electrical connection, or a communication; it can be a direct connection, or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0029] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may include that the first and second features are in direct contact, or may include that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, a first feature being "above", "above" and "above" a second feature includes that the first feature is directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below", "below" and "below" a second feature includes that the first feature is directly above and obliquely above the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0030] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0031] See also Figure 1 The present invention proposes a pipeline fatigue crack growth behavior prediction method based on acoustic emission sensing, which includes the following steps: Based on the obtained modeling parameters and fatigue crack growth rate The initial prediction model is constructed. Specifically, the modeling parameters The acquisition includes: Sampling is performed from a natural gas pipeline to obtain multiple samples, different stresses are set for the multiple samples, and two acoustic emission sensors are used to perform dual-channel acquisition tests on the multiple samples; The time-domain waveform flow signals generated by the three-point bending test of the specimen under different stress ratios are obtained, and the time-domain waveform flow signals under different loading conditions are obtained; In this embodiment, a dual sensor method using time series cross-correlation is adopted, and only the time domain waveform flow signals of two channels existing simultaneously are retained as valid data; that is, two sensors with stable performance and high sensitivity are selected to ensure that they can capture the target acoustic emission signal. The sensors should be installed at a position where the target signal can be received, and the interference of environmental noise is minimized as much as possible. The two sensors are connected to the data acquisition system respectively to ensure the accuracy and synchronization of data acquisition, and the parameters of the data acquisition system, such as sampling frequency, data storage format, etc., are configured to meet the subsequent data processing requirements. At the same time, the data acquisition functions of the two sensors are started to ensure that they can capture the target signal at the same time. For each collected data point, its corresponding timestamp is recorded, and the collected signals are pre-processed to ensure that the signals of the two channels are aligned in time. The cross-correlation algorithm is used to calculate the cross-correlation function between the signals of the two channels. By finding the maximum value of the cross-correlation function, the time delay between the two signals can be determined. According to the result of the cross-correlation function, only the acoustic emission signals of the two channels existing simultaneously are retained as valid data. The dual sensor method using time series cross-correlation can effectively reduce the influence of external noise and retain only the time domain waveform flow signals of the two channels existing simultaneously as valid data. Through strict operating steps and data processing procedures, the accuracy and reliability of this method in practical applications can be ensured.

[0032] Fatigue crack growth rate The acquisition includes: The crack opening displacement measuring instrument is used to measure multiple samples respectively, and the lengths and cycle times of multiple cracks generated during fatigue tests are obtained; that is, the lengths and cycle times of cracks generated during fatigue tests of natural gas pipeline samples are obtained; Calculate the fatigue crack growth rate based on crack length and cycle number ; Specifically, the relationship between crack length and loading cycle number under different stress ratios is first determined, the crack extension stage under different loading cycles is identified, the change of stress intensity factor ∆K, crack length a and number of cycles (i.e. fatigue cycle number N) in Stage II are obtained, and the fatigue crack growth rate is calculated by the change of stress intensity factor ∆K, crack length a and number of cycles (i.e. fatigue cycle number N). ; The calculation process is:

[0033]

[0034] Where a is the crack length, N is the number of fatigue cycles, and n are material parameters, is the change of stress intensity factor; is the maximum stress intensity factor, is the minimum stress intensity factor.

[0035] The time domain waveform flow signal is intercepted based on the preset interception conditions to obtain the acoustic emission signal; The preset interception condition is the peak definition time PDT is 300 , the impact definition time HDT is 600 , the impact lock time HLT is 1000 , the maximum duration is 1000ms, the fixed threshold is 50mv; under the preset interception conditions, the dual-channel acoustic emission signal is intercepted; Acoustic emission signal characteristic parameters are obtained based on acoustic emission signal extraction; The acoustic emission signal is subjected to bandpass filtering to reduce the influence of noise in the acoustic emission signal to obtain a denoised emission signal, and feature extraction is performed on the denoised emission signal to obtain a characteristic parameter of the acoustic emission signal; wherein the characteristic parameter of the acoustic emission signal used in the present invention is an amplitude value (A).

[0036] In this embodiment, the voltage value in the noise-removed transmission signal is extracted to obtain a reference voltage. The reference voltage is determined based on the signal noise floor, and the value range is 0.01-0.03 mV. The amplitude value is calculated based on the voltage value and the signal noise floor value. The calculation process is: Amplitude value (dB) =

[0037] Among them, V is the voltage value and 0.01mV is the reference voltage.

[0038] The modeling parameters are calculated based on the characteristic parameters of the acoustic emission signal ; Specifically, the actual number of amplitude values ​​within multiple fixed time intervals is obtained, the obtained amplitude values ​​are summed to obtain the total value of the multiple amplitude values, and the average amplitude value is calculated based on the actual number of amplitude values ​​and the total value of the amplitude values. ; Each fixed time interval contains T cycles; Average amplitude The calculation process is:

[0039] Where k represents the actual number of amplitude values ​​of the acoustic emission signal detected in the Nth fixed time interval (in cycles), It represents the sum of the amplitude values ​​of the acoustic emission signals detected within a fixed time interval; The number of unit cycle loading cycles is calculated based on the number of cycles in the fixed time interval and the fixed time interval ; As shown below:

[0040] in, is a fixed time interval, and T is the number of cycles in the fixed time interval.

[0041] The average amplitude value calculated by and the number of loading cycles per unit period Calculate the modeling parameters , as follows:

[0042] Based on the obtained modeling parameters and fatigue crack growth rate Construct an initial prediction model; The initial prediction model is:

[0043] in, is the crack growth rate (FCPR), are model parameters, is the error term; In order to facilitate the solution of the model, a logarithmic transformation was performed and the following linear form model was obtained:

[0044] in, is the crack growth rate (FCPR), is the first model parameter, is the second model parameter, is the error term, which represents the deviation between the model prediction and the actual observed value.

[0045] The Bayesian statistical method and the Markov chain Monte Carlo method are used to optimize the initial prediction model to obtain the optimal model parameters; In this embodiment, the model parameters are determined by the Bayesian inference method to determine the posterior distribution of the model parameters, where the first model parameter and the second model parameter Treated as a random variable, get the first model parameter and the second model parameter Under the Bayesian framework, the posterior distribution is used to obtain the conditional probability distribution of the observed data obtained from fatigue tests under different stresses. and the second model parameter The prior distribution of the model and the conditional probability distribution of the observed data are constructed through the Bayesian inference method to obtain the posterior distribution. The Markov Monte Carlo (MCMC) method is used to effectively approximate the posterior distribution through random sampling, which can generate a large number of samples to help estimate the distribution characteristics of the model parameters. In the Bayesian model of fatigue crack growth, the parameters may have complex correlations and nonlinear relationships. The MCMC method can handle these complexities and finally update the parameter estimates through an iterative process, so that the model can better fit the actual observed data and obtain the first optimal model parameters. , the second optimal model parameters and the error term ; The expression of the posterior distribution is:

[0046] in, is the set of model parameters, x is the observed data obtained after fatigue tests under different loading conditions, represents the prior distribution, is the likelihood function; Among them, the error term It obeys normal distribution, that is ; The conditional probability distribution of the observed data is a likelihood function of the observed data constructed based on the observed data obtained after fatigue tests under different loading conditions; The expression of the likelihood function is:

[0047] in, is the first model parameter, is the second model parameter, is the error term The standard deviation of .

[0048] Based on the optimal model parameters , and The linear form model was adjusted to obtain the prediction model;

[0049] in, is the crack growth rate (FCPR), is the first optimal model parameter, is the second optimal model parameter, is the error term, which follows a normal distribution, that is, ; represents the deviation between the model prediction value and the actual observation value.

[0050] The obtained prediction sample data is predicted based on the prediction model to obtain the prediction result.

[0051] Specifically, multiple samples are used as test sets in turn and are respectively input into the prediction model for cross-training to test the prediction model; The initial crack size obtained from the fatigue test of the predicted sample is obtained, and an arbitrary time is set for prediction to obtain the crack extension length.

[0052] The following set of embodiments further explains this method; Six single-edge notch (SEN) specimens were prepared from a section of natural gas long-distance pipeline. The diameter of the pipeline is 1422 mm, the length is 265 mm, and the wall thickness is 22 mm. Before preparing the specimens, an X-ray scan was performed on the sampling area to exclude obvious defects inside the material. Each specimen is 40 mm wide, 180 mm long, and 15 mm thick. A 9 mm notch was precisely cut in the center of the specimen to induce fatigue cracks and control their extension along the center. The sensor uses the Fuji AE144SA40 resonant acoustic sensor from Japan. The sensor has an operating frequency range of 50-400 kHz, a resonant frequency of 140 kHz, a preamplifier gain of 40 dB, a sampling frequency of 2 MHz, and room temperature fatigue loading at a constant loading frequency of 15 Hz. A crack opening displacement (COD) extensometer is placed at the notch of each specimen to measure the fatigue crack length in real time. Before the fatigue test, an 11 mm crack was prefabricated at the notch tip of each specimen. The fatigue machine controls the crack growth rate to increase from 10-5mm / cycle until the specimen breaks.

[0053] Fatigue tests were conducted on 6 specimens under different stress ratios, i.e., stress ratios of R=0.1, R=0.3, and R=0.5. Two Fuji AE144SA40 resonant acoustic sensors were used to collect time domain waveform flow signals from the start of the test to the final fracture of the specimen in each test. Finally, 6 data sets (S1-S6) were obtained. The six data sets S1-S6 were intercepted to obtain the acoustic emission signals. The interception condition was that the peak definition time PDT was 300 , the impact definition time HDT is 600 , the impact lock time HLT is 1000 , the maximum duration is 1000ms, and the fixed threshold is 50mv to intercept the time domain waveform flow signal and obtain the acoustic emission signal; The amplitude value is obtained based on the acoustic emission signal extraction; the modeling parameters are calculated based on the amplitude value ; Obtain the crack length and cycle number of the retired natural gas pipeline specimens during fatigue testing, and determine the relationship between the crack length and the number of loading cycles at different stress ratios, such as Figure 2 As shown in the figure, it can be seen that as the stress ratio increases, the number of loading cycles required for the fatigue crack to reach the same length will also increase. The change in stress intensity factor at different stress ratios is The crack growth rate under different loading cycles is basically the same, and the crack growth rate is not affected by the stress ratio. The curves of each sample basically overlap. The crack growth stage under different loading cycles is identified, and the change of stress intensity factor ∆K, crack length a and number of cycles (i.e. fatigue cycle number N) in Stage II are obtained. The fatigue crack growth rate is calculated by the change of stress intensity factor ∆K, crack length a and number of cycles (i.e. fatigue cycle number N). ; The fatigue crack growth rate is calculated based on the crack length and the number of cycles , through the fatigue crack growth rate The change curve of the stress intensity factor ∆K is established in the double logarithmic coordinate system, as shown in Figure 3 As shown, the fatigue crack growth rate can be verified and optimized based on the change curve , evaluate the fatigue crack growth rate The accuracy and applicability of fatigue crack growth rate and modeling parameters An initial prediction model is constructed and optimized to obtain the prediction model; an initial crack size of 11 mm is input into the prediction model for prediction to obtain prediction results, wherein the prediction results include the predicted crack growth rate and the fatigue crack length prediction value.

[0054] See also Figure 4 , the figure shows the modeling parameters The linear correlation diagram with fatigue crack growth rate. Through the linear relationship shown in the figure, it can be understood that this method can more accurately reflect the dynamic process of cracks.

[0055] See also Figure 5 The figure shows a comparison between the test value and the predicted value of the fatigue crack growth rate of the sample at different stress ratios (R=0.1, 0.3, 0.5). From the figure, it can be seen that the predicted crack growth rate is in good agreement with the test value, and thus the method proposed in the present invention can accurately predict the fatigue crack growth behavior of the pipeline.

[0056] See also Figure 6The figure shows the comparison between the predicted and actual values ​​of fatigue crack length under the initial crack condition (11mm) and at different stress ratios (R=0.1, 0.3, 0.5). Although there is a difference between the predicted and actual values ​​of fatigue crack length, the predicted crack length evolution trend is close to the actual measured value, which shows that this method can predict the fatigue crack propagation behavior of the pipeline.

[0057] The present invention proposes a pipeline fatigue crack propagation behavior prediction system based on acoustic emission sensing, which is used to implement the above-mentioned method, including an acquisition unit, a model unit, a processing unit, a prediction unit and an output unit; the acquisition unit is communicatively connected to the model unit, the model unit is communicatively connected to the processing unit, the processing unit is communicatively connected to the prediction unit, and the prediction unit is communicatively connected to the output unit and the acquisition unit.

[0058] Among them, the acquisition unit is used to obtain the modeling parameters and fatigue crack growth rate ; The model unit is used to obtain modeling parameters based on and fatigue crack growth rate Construct an initial prediction model; The processing unit is used to optimize the model parameters of the initial prediction model by using the Bayesian statistical method and the Markov chain Monte Carlo method to obtain the optimal model parameters, and adjust the initial prediction model based on the optimal model parameters to obtain the prediction model; The prediction unit is used to predict the acquired prediction sample data based on the prediction model to obtain a prediction result; The output unit is used to output the prediction results.

[0059] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved.

[0060] In addition, it should be understood that although this specification is described in accordance with the implementation methods, not every implementation method contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should take the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art. The above content is only to illustrate the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A pipeline fatigue crack growth behavior prediction method based on acoustic emission sensing, characterized in that: The following steps are involved: Based on the obtained modeling parameters and fatigue crack growth rate Construct an initial prediction model; The model parameters of the initial prediction model are optimized by using Bayesian statistical method and Markov chain Monte Carlo method to obtain optimal model parameters, and the initial prediction model is adjusted based on the optimal model parameters to obtain a prediction model; The acquired sample data is analyzed and predicted based on the prediction model to obtain a prediction result.

2. The method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing according to claim 1 is characterized in that: The modeling parameters The acquisition includes: Obtain the time domain waveform flow signal generated by the natural gas pipeline sample during fatigue test; intercepting the time domain waveform stream signal based on a preset interception condition to obtain an acoustic emission signal; Extracting characteristic parameters of the acoustic emission signal based on the acoustic emission signal; The modeling parameters are calculated based on the characteristic parameters of the acoustic emission signal .

3. The method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing according to claim 2 is characterized in that: The interception conditions are: peak definition time PDT300 , the impact definition time HDT is 600 , the impact lock time HLT is 1000 , the maximum duration is 1000ms and the fixed threshold is 50mv.

4. The method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing according to claim 2 is characterized in that: Acoustic emission signal characteristic parameters are obtained based on the acoustic emission signal extraction, including: Performing bandpass filtering on the acoustic emission signal to obtain a de-noised emission signal; Feature extraction is performed on the de-noised emission signal to obtain characteristic parameters of the acoustic emission signal.

5. The method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing according to claim 2 is characterized in that: The modeling parameters are calculated based on the characteristic parameters of the acoustic emission signal ,include Get the actual number of amplitude values ​​within multiple fixed time intervals; Summing the plurality of amplitude values ​​to obtain a total amplitude value; The average amplitude value is calculated based on the actual number of amplitude values ​​and the total value of the amplitude values ; The number of unit cycle loading cycles is calculated based on the number of cycles in the fixed time interval and the fixed time interval ; Based on the average amplitude value and the number of loading cycles per unit period Calculate the modeling parameters ; The calculation process of the modeling parameters is: in, is a fixed time interval, T is the number of cycles in the fixed time interval, represents the sum of the amplitude values ​​of the acoustic emission signals detected in a fixed time interval, and k represents the actual number (in cycles) of the amplitude values ​​of the acoustic emission signals detected in the Nth fixed time interval.

6. The method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing according to claim 1 is characterized in that: The fatigue crack growth rate The acquisition includes: Obtain the crack length and cycle number of natural gas pipeline specimens during fatigue testing; The fatigue crack growth rate is calculated based on the crack length and the number of cycles. .

7. The method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing according to claim 1 is characterized in that: The Bayesian statistical method and the Markov chain Monte Carlo method are used to optimize the initial prediction model to obtain the optimal model parameters, including: Determining the prior distribution of the model parameters through experimental data and domain knowledge; The prior distribution of the model parameters and the conditional probability distribution of the observed data are constructed using the Bayesian inference method to obtain a posterior distribution; The posterior distribution is sampled and solved based on the Markov Monte Carlo method to obtain the optimal model parameters.

8. The method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing according to claim 1 is characterized in that: The expression of the posterior distribution is: in, is the set of model parameters, x is the observed data obtained after fatigue tests under different loading conditions, represents the prior distribution, is the likelihood function; The conditional probability distribution of the observed data is a likelihood function of the observed data constructed based on the observed data obtained after fatigue tests under different loading conditions; The expression of the likelihood function is: in, is the first model parameter, is the second model parameter, is the error term The standard deviation of .

9. The method for predicting pipeline fatigue crack growth behavior based on acoustic emission sensing according to claim 1, characterized in that: The prediction model is: in, is the crack growth rate, is the first optimal model parameter, is the second optimal model parameter, is the error term, It obeys normal distribution, that is , which represents the deviation between the model prediction and the actual observation.

10. A pipeline fatigue crack growth behavior prediction system based on acoustic emission sensing, used to implement the method described in any one of claims 1 to 9, characterized in that: include: The model unit is used to obtain modeling parameters based on and fatigue crack growth rate Construct an initial prediction model; The processing unit is used to optimize the initial prediction model by using Bayesian statistical method and Markov chain Monte Carlo method to obtain optimal model parameters, and adjust the initial prediction model based on the optimal model parameters to obtain a prediction model; The prediction unit is used to analyze and predict the acquired prediction sample data based on the prediction model to obtain a prediction result; The output unit is used to output the prediction results.

Images