Pipeline corrosion source tracing method and system based on multi-source data fusion

Abstract

The present application relates to the technical field of pipeline corrosion analysis, in particular to a pipeline corrosion source tracing method and system based on multi-source data fusion, which collects flow field time series of each flow field sensing node and electrochemical time series of each electrochemical sensing node, extracts macroscopic turbulent disturbance mode time series at the structural mutation site, performs adaptive filtering on the flow field time series at the structural mutation site to obtain residual flow field time series; when transient disturbance is detected in the residual flow field time series and synchronous transient response is generated in the electrochemical time series, the matching event is identified as a micro-turbulent event induced by corrosion product film breakage; according to the identified micro-turbulent event, the projection of the flow field source space position of the transient disturbance on the pipe wall is calculated to determine the spatial position of the corrosion product film breakage point as the corrosion source position; the micro-turbulent event induced by corrosion product film breakage can be identified, and the spatial position of the corrosion source can be accurately positioned.

Description

Technical Field

[0001] This invention relates to the field of pipeline corrosion analysis technology, specifically to a pipeline corrosion source tracing method and system based on multi-source data fusion. Background Technology

[0002] Pipelines, as core transportation infrastructure in oil and gas, chemical, and other fields, inevitably suffer from corrosion damage during long-term service. Once a corrosion product film (such as oxide, sulfide, or carbonate deposits) forms on the inner wall of the pipeline, its rough surface disturbs the fluid inside, inducing localized microturbulence and generating periodic shear stress at the interface. When this shear stress concentrates in weak areas of the corrosion product film, it causes that area to continuously thin and eventually form a hole. The exposed fresh metal substrate forms a small anodic area, which, together with the surrounding large cathodic area, constitutes a localized corrosion cell, accelerating pitting and other localized corrosion, creating a safety hazard of pipeline perforation, leakage, or even catastrophic accidents.

[0003] Turbulence in pipelines typically occurs at points where the fluid changes direction, such as pipe bends, tees, and diameter changes. These structural abrupt changes can generate turbulent interference in the local flow field, severely impacting the reliability and accuracy of corrosion source location in complex interference environments. Summary of the Invention

[0004] (1) Technical problems to be solved The purpose of this invention is to provide a method and system for tracing the source of pipeline corrosion based on multi-source data fusion, which solves the problem of difficulty in accurately locating the spatial location of corrosion product film damage points in complex flow field interference environments at locations of abrupt changes in pipeline structure.

[0005] (2) Technical solution To achieve the above objectives, on the one hand, the present invention provides a method for tracing the source of pipeline corrosion based on multi-source data fusion, the method comprising: S1. Flow field sensing nodes and electrochemical sensing nodes are set up at the structural change points of the pipeline and at its upstream and downstream locations, respectively, and the flow field time series of each flow field sensing node and the electrochemical time series of each electrochemical sensing node are collected.

[0006] S2. Based on the spatial distribution characteristics of the flow field time series at different sensing nodes, extract the macroscopic turbulence disturbance mode time series at the structural abrupt change locations.

[0007] S3. Using the macroscopic turbulence disturbance mode time series as a reference, perform adaptive filtering on the flow field time series at the structural abrupt change location to obtain the residual flow field time series.

[0008] S4. Perform transient event matching on the residual flow field time series and the electrochemical time series under time synchronization. When a transient disturbance is detected in the residual flow field time series and a synchronous transient response is generated in the electrochemical time series, the matching event is identified as a microturbulent event induced by the damage of the corrosion product film.

[0009] S5. Based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes, the spatial location of the flow field source of the transient disturbance is calculated; the projection of the spatial location of the flow field source on the pipe wall is determined as the spatial location of the corrosion product film rupture point that induces the microturbulent event, and traced back to the location of the corrosion source.

[0010] Furthermore, the method for extracting the macroscopic turbulence disturbance mode time series at locations of structural abrupt changes based on the spatial distribution characteristics of the flow field time series at different sensing nodes includes: The time-frequency transformation of the flow field time series is performed to obtain the time spectrum of each sensing node; the spatial distribution of the time spectrum of different sensing nodes is decomposed into a specific spatial pattern.

[0011] Based on the time spectrum corresponding to a specific spatial pattern, the flow field time series characterized only by the specific spatial pattern is reconstructed to obtain the macroscopic turbulence disturbance pattern time series at the structural abrupt change location.

[0012] Furthermore, the method for obtaining a specific spatial pattern by pattern decomposing the spatial distribution of the temporal spectrum of different sensing nodes includes: A space-frequency matrix reflecting the distribution of time-frequency energy along the pipeline axis is constructed based on the time spectrum of each sensing node and the spatial location of the sensing nodes.

[0013] Perform eigenorthogonal decomposition on the space-frequency matrix to obtain a set of spatial modes and the energy contribution rate corresponding to each spatial mode.

[0014] From the set of spatial modes, the spatial mode that has the largest energy contribution rate at the structural abrupt change site and shows a monotonically decreasing trend along the downstream direction is selected as a specific spatial mode.

[0015] Furthermore, the method of obtaining the residual flow field time series by performing adaptive filtering on the flow field time series at the location of structural abrupt changes, using the macroscopic turbulence disturbance mode time series as a reference, includes: The time series of the macroscopic turbulent disturbance mode is used as the reference input signal of the adaptive filter, and the time series of the flow field at the structural abrupt change is used as the main input signal of the adaptive filter.

[0016] The adaptive filter iteratively updates its filter weight coefficients based on the correlation between the reference input signal and the main input signal using the minimum mean square error criterion, so as to cancel the signal component characterized by the macroscopic turbulence interference mode from the main input signal, and uses the error signal output by the adaptive filter as the residual flow field time series.

[0017] Furthermore, the method for calculating the spatial location of the transient disturbance source based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes includes: The arrival time and amplitude of transient disturbances received by each downstream flow field sensing node in the same identified microturbulence event are obtained. The arrival time delay difference between two flow field sensing nodes is obtained based on the spatial location of any two flow field sensing nodes and the arrival time. A set of time delay difference equations composed of multiple sets of arrival time delay differences is established.

[0018] Based on the attenuation mapping relationship satisfied by the ratio of the disturbance amplitude of any two flow field sensing nodes and the ratio of the distance from the flow field source of the transient disturbance to the two flow field sensing nodes, an amplitude attenuation gradient constraint equation is established.

[0019] By jointly solving the time delay difference equation set and the amplitude attenuation gradient constraint equation, the spatial location of the transient disturbance source in the fluid medium inside the pipe is obtained.

[0020] Furthermore, the method for establishing the amplitude attenuation gradient constraint equation based on the attenuation mapping relationship satisfied by the ratio of the disturbance amplitudes of any two flow field sensing nodes and the ratio of the distances from the flow field source of the transient disturbance to the two flow field sensing nodes includes: The propagation attenuation law of the transient disturbance in the fluid medium inside the pipe is obtained, and an attenuation model of the disturbance amplitude as a function of propagation distance is established based on the propagation attenuation law.

[0021] Substituting the disturbance amplitudes of any two flow field sensing nodes into the attenuation model yields an expression for the amplitude ratio with the distance from the flow field source to the two flow field sensing nodes as the unknown.

[0022] The amplitude ratio expression is used to establish an equation with the measured ratio of the disturbance amplitudes of the two flow field sensing nodes, which is then transformed into an attenuation mapping relationship corresponding to the ratio of the disturbance amplitudes of the two flow field sensing nodes and the ratio of the distances from the flow field source to the two flow field sensing nodes. Multiple sets of attenuation mapping relationships corresponding to the flow field sensing node pairs are combined to form an amplitude attenuation gradient constraint equation.

[0023] Furthermore, the method for obtaining the spatial location of the transient disturbance source in the fluid medium within the pipe by jointly solving the time delay difference equation set and the amplitude attenuation gradient constraint equation includes: By combining the time delay difference equations and the amplitude attenuation gradient constraint equations, a joint solution model is constructed with the spatial location of the flow field source as the unknown variable.

[0024] Using the distance ratio from the flow field source to each flow field sensing node determined by the amplitude attenuation gradient constraint equation as the constraint condition, the time delay difference equation set is constrained to obtain the constrained time delay difference equation set.

[0025] An iterative optimization algorithm is used to solve the constrained time delay difference equations. In each iteration, the theoretical arrival time delay to each flow field sensing node is calculated based on the current estimated spatial location of the flow field source. The residual between the theoretical arrival time delay and the measured arrival time is used to update the current estimated spatial location of the flow field source until the preset convergence condition is met, thereby obtaining the spatial location of the flow field source of the transient disturbance in the fluid medium inside the pipe.

[0026] Furthermore, the method for determining the spatial location of the flow field source's spatial position on the pipe wall as the spatial location of the corrosion product film rupture point that induces microturbulence events includes: Obtain a three-dimensional geometric model of the inner wall of the pipeline at the site of structural abrupt change, as well as the layout coordinates of the electrochemical sensing nodes that generate synchronous transient responses.

[0027] Based on the three-dimensional geometric model of the inner wall, the spatial location of the transient disturbance source is calculated and projected perpendicularly onto the inner wall of the pipe to obtain the set of projection points of the spatial location of the flow field source on the inner wall of the pipe.

[0028] Based on the correlation between each projection point in the projection point set and the electrochemical sensing node that generates a synchronous transient response, the spatial location of the corrosion product film damage point that induces the microturbulent event is determined from the projection point set.

[0029] Furthermore, the method for determining the spatial location of the corrosion product film damage point that induces the microturbulent event from the set of projection points based on the correlation between each projection point in the set of projection points and the electrochemical sensing node that generates a synchronous transient response includes: The deployment coordinates of the electrochemical sensing nodes that generate synchronous transient responses and the electrochemical transient characteristics of the synchronous transient responses received by the electrochemical sensing nodes are obtained.

[0030] The spatial distance between each projection point in the projection point set and the layout coordinates of the electrochemical sensing node is calculated, and the electrochemical-spatial correlation confidence of each projection point is constructed based on the spatial distance and electrochemical transient characteristics.

[0031] The projection point corresponding to the maximum value of the electrochemical-spatial correlation confidence in the projection point set is determined as the spatial location of the corrosion product film damage point that induces microturbulence events.

[0032] On the other hand, based on the same inventive concept, this invention also provides a pipeline corrosion source tracing system based on multi-source data fusion, the system comprising: The sensing node data acquisition module is used to deploy flow field sensing nodes and electrochemical sensing nodes at the structural abrupt changes in the pipeline and at its upstream and downstream locations, respectively, to collect the flow field time series of each flow field sensing node and the electrochemical time series of each electrochemical sensing node.

[0033] The macroscopic turbulence disturbance mode time series extraction module is used to extract the macroscopic turbulence disturbance mode time series at locations of structural abrupt changes based on the spatial distribution characteristics of the flow field time series at different sensing nodes.

[0034] The residual flow field time series acquisition module is used to perform adaptive filtering on the flow field time series at the structural abrupt change location, with the macroscopic turbulence disturbance mode time series as a reference, to obtain the residual flow field time series.

[0035] The microturbulence event identification module is used to match transient events in time synchronization between the residual flow field time series and the electrochemical time series. When a transient disturbance is detected in the residual flow field time series and a synchronous transient response is generated in the electrochemical time series, the matching event is identified as a microturbulence event induced by the damage of the corrosion product film.

[0036] The corrosion source location determination module is used to calculate the spatial location of the transient disturbance source based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes; the projection of the spatial location of the flow field source on the pipe wall is determined as the spatial location of the corrosion product film rupture point that induces the microturbulent event, and traced back to the corrosion source location.

[0037] (3) Beneficial effects Compared with the prior art, the beneficial effects of the present invention are: 1. By deploying flow field sensing nodes and electrochemical sensing nodes in the pipeline at the site of structural abrupt change and upstream and downstream, and introducing macroscopic turbulence interference mode extraction and adaptive filtering mechanism, the background turbulence interference caused by structural abrupt change is effectively suppressed. The micro-turbulence transient signal induced by corrosion product film damage is accurately separated from the complex flow field noise, realizing high-reliability identification of weak corrosion signals and effectively capturing corrosion events.

[0038] 2. Further integrating the time delay difference information of transient disturbances reaching downstream nodes with amplitude attenuation gradient constraints, the three-dimensional spatial location of the flow field source is accurately located through the joint solution model. Combined with the synchronous transient response of electrochemical sensing nodes and the confidence assessment of electrochemical-spatial correlation, the location of the flow field source is accurately projected onto the inner wall of the pipeline, realizing the accurate tracing of corrosion product film damage points and significantly improving the proactive control and precise maintenance capabilities of pipeline corrosion. Attached Figure Description

[0039] Figure 1 This is a flowchart of the pipeline corrosion source tracing method based on multi-source data fusion of the present invention.

[0040] Figure 2 This is a schematic diagram of the module composition of the pipeline corrosion source tracing system based on multi-source data fusion of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1: As Figure 1 As shown, this embodiment provides a method for tracing the source of pipeline corrosion based on multi-source data fusion. The method includes: S1. Flow field sensing nodes and electrochemical sensing nodes are set up at the structural change points of the pipeline and at its upstream and downstream locations, respectively, and the flow field time series of each flow field sensing node and the electrochemical time series of each electrochemical sensing node are collected. For example, flow field sensing nodes are used to collect flow field time series and can be implemented using pressure pulsation sensors or flow velocity sensors. Electrochemical sensing nodes are used to collect electrochemical time series and can be implemented using embedded electrochemical impedance sensors. These nodes are arranged in an array along the axial and circumferential directions of the pipe to achieve spatial coverage of different areas of the pipe's inner wall. Taking a pipe bend as an example, two flow field sensing nodes can be installed as reference nodes on the upstream straight pipe section, two flow field sensing nodes can be installed on the bend's arc section body, and at least three flow field sensing nodes can be sequentially installed as downstream receiving nodes on the downstream straight pipe section. Simultaneously, at least four electrochemical sensing nodes are evenly distributed circumferentially along the inner wall of the bend's arc section. Each sensing node records its accurate three-dimensional coordinates after installation. Each flow field sensing node synchronously and continuously samples the pressure pulsations of the flow field inside the pipe at a uniform sampling frequency. The selection of the sampling frequency should satisfy the Nyquist sampling theorem and cover the characteristic frequency band of microturbulence induced by corrosion product film damage. Based on engineering experience with pipe diameter and flow velocity, the characteristic frequency band of microturbulence induced by corrosion product film rupture is typically located in the range of several hertz to several hundred hertz, and the sampling frequency is generally no less than twice the upper limit of the characteristic frequency band. Each electrochemical sensing node synchronously acquires the electrochemical response signal of the pipe wall, and the sampling frequency is consistent with that of the flow field sensing node to ensure the effectiveness of subsequent time synchronization matching.

[0043] S2. Based on the spatial distribution characteristics of the flow field time series at different sensing nodes, extract the macroscopic turbulence disturbance mode time series at the structural abrupt change locations.

[0044] The method for extracting the macroscopic turbulence disturbance mode time series at locations of structural abrupt changes based on the spatial distribution characteristics of the flow field time series at different sensing nodes includes: The time-frequency transformation of the flow field time series is performed to obtain the time spectrum of each sensing node; the spatial distribution of the time spectrum of different sensing nodes is decomposed into a specific spatial pattern.

[0045] The method for obtaining a specific spatial pattern by pattern decomposing the spatial distribution of the temporal spectrum of different sensing nodes includes: A space-frequency matrix reflecting the distribution of time-frequency energy along the pipeline axis is constructed based on the time spectrum of each sensing node and the spatial location of the sensing nodes.

[0046] Perform eigenorthogonal decomposition on the space-frequency matrix to obtain a set of spatial modes and the energy contribution rate corresponding to each spatial mode.

[0047] From the set of spatial modes, the spatial mode that has the largest energy contribution rate at the structural abrupt change site and shows a monotonically decreasing trend along the downstream direction is selected as a specific spatial mode.

[0048] For example, time-frequency transformation is performed on the flow field time series collected by each flow field sensing node to obtain the time spectrum of each sensing node. The time-frequency transformation can employ a short-time Fourier transform. Acquired time-domain signals The short-time Fourier transform is defined as: ;in, For window functions (such as the Hanning window). Centered on time Frequency. Time spectrum. This reflects the energy distribution of the signal at different times and frequencies. In the macroscopic turbulence interference mode extraction stage, to capture the overall spatial characteristics of macroscopic turbulence, a representative time window is selected. This representative time window should be a stable flow period containing typical macroscopic turbulence characteristics; a continuous window with the largest signal variance within the entire acquisition period can be chosen. Within this time window, the average energy of each frequency component is taken to obtain the frequency-energy distribution vector of each sensing node. Let the first value be denoted as _____. Each sensing node at frequency The average energy at that location is Arrange the frequency-energy distribution vectors of all sensing nodes row by row according to their spatial location (along the pipe axis) to construct a space-frequency matrix. ,matrix The number of rows equals the total number of sensing nodes. The number of columns equals the number of discrete frequency points. Matrix elements The selection of discrete frequency points can be based on the Nyquist sampling theorem and actual requirements to determine the frequency range. Discrete frequency points are then selected uniformly within this range, thereby determining the number of discrete frequency points. .

[0049] For space-frequency matrix Perform eigenorthogonal decomposition to transform the matrix Perform singular value decomposition: ;in, It is a singular value diagonal matrix; It is a right singular vector matrix; The column vectors are the spatial modes, the th... spatial modes The energy contribution rate is defined as its corresponding singular value. The ratio of the square of to the sum of the squares of all singular values: ;in Let be the rank of the matrix. From all spatial modes, select the spatial mode that has the largest energy contribution rate of the corresponding component at the structural abrupt change site and whose components show a monotonically decreasing trend along the downstream direction, and denote it as a specific spatial mode. This is because macroscopic turbulence generated at abrupt structural changes is the main source of flow field disturbance. Its energy is strongest at these changes and gradually decreases as the fluid propagates downstream due to viscous dissipation and flow mixing. Therefore, a monotonically decreasing energy level along the downstream direction is a typical characteristic of the spatial distribution of macroscopic turbulence. If multiple spatial modes simultaneously satisfy the above conditions, the one with the highest energy contribution rate is selected as the specific spatial mode.

[0050] Based on the time-frequency spectrum corresponding to a specific spatial model, the flow field time series characterized solely by the specific spatial model is reconstructed, yielding the macroscopic turbulence disturbance mode time series at locations of structural abrupt changes. The specific spatial model is then projected back into the time-frequency domain to reconstruct the flow field time series for each sensing node, characterized solely by the specific spatial model. Specifically, the reconstruction matrix corresponding to the specific spatial model... The frequency-energy distribution of nodes corresponding to structural abrupt changes in the middle section is combined with time-spectrum phase information to perform a short-time inverse Fourier transform, reconstructing the time series of macroscopic turbulent disturbance modes at the structural abrupt changes, denoted as […]. .

[0051] S3. Using the macroscopic turbulence disturbance mode time series as a reference, perform adaptive filtering on the flow field time series at the structural abrupt change location to obtain the residual flow field time series. The method for obtaining the residual flow field time series by performing adaptive filtering on the flow field time series at locations of structural abrupt changes, using the macroscopic turbulence disturbance mode time series as a reference, includes: The time series of the macroscopic turbulence disturbance mode is used as the reference input signal of the adaptive filter, and the time series of the flow field at the structural abrupt change is used as the main input signal of the adaptive filter. The adaptive filter iteratively updates its filter weight coefficients based on the correlation between the reference input signal and the main input signal using the minimum mean square error criterion, so as to cancel the signal component characterized by the macroscopic turbulence interference mode from the main input signal, and uses the error signal output by the adaptive filter as the residual flow field time series.

[0052] For example, the reconstructed macroscopic turbulence disturbance mode time series The reference input signal for the adaptive filter is the actual flow field time series collected at the location of structural abrupt changes. The reference input signal serves as the main input signal for the adaptive filter. The core mechanism of an adaptive filter lies in: the reference input signal... With the main input signal The macroscopic turbulence components in the signal are highly correlated, while the microturbulence signal is uncorrelated with the macroscopic turbulence interference mode. Therefore, the adaptive filter can automatically estimate and eliminate the macroscopic turbulence components in the main input signal during the iteration process, while retaining the microturbulence signal.

[0053] Let the filter weight vector of the adaptive filter be... ,in Let the filter order be . This is the current iteration step. The filter output estimate is: ;in This is the current and historical value vector of the reference input signal. The error signal (i.e., the point-by-point values ​​of the residual flow field time series) is: Based on the minimum mean square error criterion, the minimum mean square algorithm is used to iteratively update the filter weight vector: ;in, This is the step size factor, used to control the iteration convergence speed and stability. Because... and The dimensions of all of them are consistent with the reference input signal (i.e., they have the dimensions of signal amplitude), while the filter weight vector It is a dimensionless vector, therefore The dimension of this is the reciprocal of the signal power (i.e., the reciprocal of the variance of the reference input signal). The range of values ​​for the step size factor is... ,in Reference input signal The power estimate (with dimensions consistent with the square of the signal amplitude) can be obtained through statistical calculation of historical data of the reference input signal. An excessively large step size factor can cause the algorithm to diverge, while an excessively small step size will result in slow convergence. In practical applications, a suitable value can be selected within the above range based on debugging results; for example, a step size of [missing value]. Filter order The selection of the relevant length needs to cover macroscopic turbulence disturbances, which can usually be achieved by calculating the autocorrelation function of the reference input signal and taking the autocorrelation function attenuated to its peak value. The number of time delay samples corresponding to a location is used as a reference for the filter order.

[0054] The iterative updates continue until the mean square value of the error signal converges to a stable value. The error signal sequence at each time point after stable convergence is then processed. This forms a complete residual flow field time series. The main components of macroscopic turbulence interference have been eliminated from the residual flow field time series, while the micro-turbulence transient signal components that are unrelated to macroscopic turbulence are retained.

[0055] S4. Perform transient event matching on the residual flow field time series and the electrochemical time series under time synchronization. When a transient disturbance is detected in the residual flow field time series and a synchronous transient response is generated in the electrochemical time series, the matching event is identified as a microturbulent event induced by the damage of the corrosion product film. For example, transient event matching is performed on the residual flow field time series and the electrochemical time series collected by each electrochemical sensing node under time synchronization. In the residual flow field time series, a fixed threshold or adaptive threshold method is used to detect transient disturbance events. The adaptive threshold can be defined as a multiple of the standard deviation of the background noise of the residual flow field time series, for example, 3 to 5 times the standard deviation. The standard deviation of the background noise is obtained by statistically calculating the residual flow field time series during the silent period. The silent period can be the time period when the pipe is not flowing or when the flow is normal and stable at the initial stage of acquisition, or it can be automatically identified by the time period where the minimum variance of the sliding window is located. When the signal amplitude at a certain moment in the residual flow field time series exceeds the set threshold, and the width of the time window that continuously exceeds the threshold is within a reasonable range (e.g., no more than 100 milliseconds), a transient disturbance is determined to have been detected. Simultaneously, the electrochemical time series of each electrochemical sensing node is monitored synchronously, and the electrochemical transient response is also detected using a threshold judgment method. The threshold selection method is consistent with that on the flow field side, and is determined based on the baseline signal statistics of each electrochemical sensing node during the silent period. When a transient disturbance is detected in the residual flow field time series, and any electrochemical sensing node generates a synchronous electrochemical transient response within a set time synchronization window (the width of the time synchronization window can be determined by combining the maximum possible propagation delay from the flow field source to the farthest electrochemical sensing node, for example, 50 milliseconds to 200 milliseconds), then this event is identified as a microturbulent event induced by corrosion product film damage. This is because when the corrosion product film is damaged, the fresh metal substrate is exposed to the fluid, simultaneously generating microturbulent disturbances in the flow field and an electrochemical impedance response. These two are synchronous in time, while the macroscopic turbulent background noise caused by abrupt changes in the pipe structure does not synchronously trigger the transient response of the electrochemical sensing node, thus enabling effective differentiation.

[0056] S5. Based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes, the spatial location of the flow field source of the transient disturbance is calculated. The method for calculating the spatial location of the transient disturbance source based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes includes: The arrival time and amplitude of transient disturbances received by each downstream flow field sensing node in the same identified microturbulence event are obtained. The arrival time delay difference between two flow field sensing nodes is obtained based on the spatial location of any two flow field sensing nodes and the arrival time. A time delay difference equation set consisting of multiple sets of arrival time delay differences is established. Based on the attenuation mapping relationship satisfied by the ratio of the disturbance amplitude of any two flow field sensing nodes and the ratio of the distance from the flow field source of the transient disturbance to the two flow field sensing nodes, an amplitude attenuation gradient constraint equation is established. The method for establishing the amplitude attenuation gradient constraint equation based on the attenuation mapping relationship satisfied by the ratio of the disturbance amplitude of any two flow field sensing nodes and the ratio of the distances from the flow field source of the transient disturbance to the two flow field sensing nodes includes: Obtain the propagation attenuation law of the transient disturbance in the fluid medium inside the pipe, and establish an attenuation model of the disturbance amplitude as a function of propagation distance based on the propagation attenuation law; Substituting the disturbance amplitudes of any two flow field sensing nodes into the attenuation model yields an amplitude ratio expression with the distance from the flow field source to the two flow field sensing nodes as the unknown. The amplitude ratio expression is used to establish an equation with the measured ratio of the disturbance amplitudes of the two flow field sensing nodes, which is then transformed into an attenuation mapping relationship corresponding to the ratio of the disturbance amplitudes of the two flow field sensing nodes and the ratio of the distances from the flow field source to the two flow field sensing nodes. Multiple sets of attenuation mapping relationships corresponding to the flow field sensing node pairs are combined to form an amplitude attenuation gradient constraint equation.

[0057] For example, after identifying a microturbulent event, the arrival time and amplitude of the transient disturbance received by each downstream flow field sensing node are extracted. The arrival time is determined as follows: in the residual flow field time series of each downstream flow field sensing node, the time of maximum value of the transient disturbance envelope is taken as the arrival time, denoted as the i-th time. The arrival times of the downstream flow field sensing nodes are The disturbance amplitude is .

[0058] For any two downstream flow field sensing nodes and By utilizing their respective arrival times and the spatial locations of the sensing nodes, an arrival time delay difference equation is established. Let the three-dimensional coordinates of the flow field source be... , No. The three-dimensional coordinates of the downstream flow field sensing nodes are: Flow field source to the first The distance between each sensing node along the fluid propagation path is The propagation speed of flow field disturbance in the fluid medium inside the pipe is (The propagation speed can be determined based on the sound velocity of the fluid medium inside the pipe combined with the average flow velocity), then: The above equations are established for all possible node pairs, forming multiple sets of arrival time delay difference equations, thus constituting a time delay difference equation set. Downstream deployment... When there are multiple flow field sensing nodes, it is possible to establish A set of equations can be used, but solving several linearly independent sets simultaneously is sufficient for the solution. The remaining equations can be used to improve the redundancy and robustness of the solution. Regarding the establishment of the amplitude attenuation gradient constraint equations, the propagation and attenuation law of transient disturbances in the fluid medium within the pipe is determined, using a reference distance... The disturbance amplitude at the distance from the known intensity excitation source during the calibration experiment (i.e., the measurement reference distance). Based on this, the variation of the disturbance amplitude with propagation distance is modeled as follows: ;in, For transmission distance, The absorption attenuation coefficient of the fluid medium. This represents the geometric diffusion attenuation exponent. In the above expression, Dimensions and Consistent (in terms of signal amplitude units such as Pa or m / s). and All dimensions are in meters (m). For pressure pulsation propagation in a closed pipe, geometric diffusion attenuation is less constrained by the pipe wall. The value of is in the range of 0 to 1, where it can be approximated as 1 for plane wave propagation. In actual engineering and It is recommended to conduct calibration experiments, measuring amplitude attenuation data at different distances from an excitation source of known intensity, and then determining the attenuation model through least-squares fitting. The first... The and the first Substituting the disturbance amplitude received by each downstream flow field sensing node into the attenuation model, we obtain the expression for the ratio of amplitudes: ; compare the measured amplitude with Substitute, establish and The equation represents the attenuation mapping relationship of unknown quantities.

[0059] By jointly solving the time delay difference equation set and the amplitude attenuation gradient constraint equation, the spatial location of the transient disturbance source in the fluid medium inside the pipe is obtained.

[0060] The method for obtaining the spatial location of the transient disturbance source in the fluid medium within the pipe by jointly solving the time delay difference equation set and the amplitude attenuation gradient constraint equation includes: By combining the time delay difference equations and the amplitude attenuation gradient constraint equations, a joint solution model is constructed with the spatial location of the flow field source as the unknown variable.

[0061] Using the distance ratio from the flow field source to each flow field sensing node determined by the amplitude attenuation gradient constraint equation as the constraint condition, the time delay difference equation set is constrained to obtain the constrained time delay difference equation set.

[0062] An iterative optimization algorithm is used to solve the constrained time delay difference equations. In each iteration, the theoretical arrival time delay to each flow field sensing node is calculated based on the current estimated spatial location of the flow field source. The residual between the theoretical arrival time delay and the measured arrival time is used to update the current estimated spatial location of the flow field source until the preset convergence condition is met, thereby obtaining the spatial location of the flow field source of the transient disturbance in the fluid medium inside the pipe.

[0063] For example, by simultaneously solving the time delay difference equations and the amplitude attenuation gradient constraint equations, a flow field source spatial coordinate system can be constructed. This is a joint solution model for unknown variables. During the solution process, the distance ratios from the flow field source to each downstream flow field sensing node are determined using the amplitude attenuation gradient constraint equations. These distance ratios are then substituted into the time delay difference equations as constraints, transforming the unconstrained time delay difference equations into a constrained set. This reduces the degrees of freedom of the unknown variables and improves the stability of the solution.

[0064] Iterative optimization algorithms (such as the Gauss-Newton method or the Levenberg-Marquardt algorithm) are used to solve the constrained time delay difference equations. In each iteration, the estimated spatial location of the current flow field source is used. Calculate the theoretical distance from the flow field source to each downstream flow field sensing node. Then calculate the theoretical arrival delay. The objective function is constructed using the residual between the theoretical arrival time delay and the measured arrival time, and the estimated spatial location of the flow field source is updated using gradient information. The iteration termination condition (i.e., the preset convergence condition) is set as follows: the change in the estimated spatial location of the flow field source between two adjacent iterations is less than a set threshold (e.g., 0.001m), or the number of iterations reaches a preset maximum number of iterations; the iteration terminates when either condition is met. The estimated spatial location of the flow field source output after iteration convergence is the spatial location of the transient disturbance in the fluid medium within the pipe.

[0065] The projection of the spatial location of the flow field source onto the pipe wall is determined as the spatial location of the corrosion product film rupture point that induces microturbulence events; The method for determining the spatial location of the corrosion product film rupture point that induces microturbulence events by projecting the spatial location of the flow field source onto the pipe wall includes: Obtain a three-dimensional geometric model of the inner wall of the pipeline at the site of structural abrupt change, as well as the layout coordinates of the electrochemical sensing nodes that generate synchronous transient responses.

[0066] Based on the three-dimensional geometric model of the inner wall, the spatial location of the transient disturbance source is calculated and projected perpendicularly onto the inner wall of the pipe to obtain the set of projection points of the spatial location of the flow field source on the inner wall of the pipe.

[0067] For example, a three-dimensional geometric model of the inner wall of the pipeline at the abrupt change in structure is obtained. This model can be imported from pipeline design drawings or obtained by performing a three-dimensional scanning model of the pipeline. The spatial geometry of the pipeline inner wall is expressed in the form of a discrete triangular mesh or parametric surface. Using the calculated spatial location of the flow field source as the query point, all facets of the three-dimensional geometric model of the pipeline inner wall are traversed. For each facet, the projection point of the flow field source spatial location along the outer normal of that facet is calculated. If a ray from the flow field source spatial location along the outer normal direction of a facet (i.e., pointing towards the pipe wall) intersects with the inner wall geometric model, then that intersection point is included in the projection point set.

[0068] Based on the correlation between each projection point in the projection point set and the electrochemical sensing node that generates a synchronous transient response, the spatial location of the corrosion product film damage point that induces the microturbulent event is determined from the projection point set.

[0069] The method for determining the spatial location of the corrosion product film damage point that induces microturbulent events from the set of projection points based on the correlation between each projection point in the set of projection points and the electrochemical sensing node that generates a synchronous transient response includes: The deployment coordinates of the electrochemical sensing nodes that generate synchronous transient responses and the electrochemical transient characteristics of the synchronous transient responses received by the electrochemical sensing nodes are obtained.

[0070] The spatial distance between each projection point in the projection point set and the layout coordinates of the electrochemical sensing node is calculated, and the electrochemical-spatial correlation confidence of each projection point is constructed based on the spatial distance and electrochemical transient characteristics.

[0071] The projection point corresponding to the maximum value of the electrochemical-spatial correlation confidence in the projection point set is determined as the spatial location of the corrosion product film damage point that induces microturbulence events.

[0072] And traced back to the source of corrosion.

[0073] For example, the deployment coordinates and corresponding electrochemical transient characteristics of each electrochemical sensing node that generates a synchronous transient response in this microturbulent event are obtained. The electrochemical transient characteristics may include the transient response amplitude and the frequency characteristics of the transient response. The transient response amplitude received by the electrochemical sensing node reflects the electrochemical signal transmission intensity between the corrosion product film failure point and the electrochemical sensing node. Electrochemical sensing nodes closer to the corrosion failure point typically receive a larger electrochemical transient response amplitude.

[0074] For each projection point in the projection point set, calculate the spatial distance between the projection point and the coordinates of each electrochemical sensing node that generates a synchronous transient response, and denote the distance between the projection point and the coordinates of each projection point. The projection point and the first The spatial distance between the electrochemical sensing nodes is Based on the amplitude of the electrochemical transient response received by each electrochemical sensing node. and spatial distance , construct the first Electrochemical-spatial correlation confidence level of each projection point : ;in, This represents the average transient response amplitude of each electrochemical sensing node. Coordinates are assigned to each sensing node up to the [number]th [node]. The average distance between each projection point This is a dimensionless adjustment parameter (can be set to 1). The total number of electrochemical sensing nodes that generate synchronous transient responses. (The two items in parentheses are...) and All are dimensionless quantities, confidence function The confidence function indicates that if a projection point is closer to an electrochemical sensing node with a large electrochemical transient response amplitude (i.e., the difference between the normalized distance and the reciprocal of the normalized amplitude is smaller), then the spatial correlation of that projection point as a corrosion source is higher, and the confidence function value is larger. The projection point corresponding to the maximum electrochemical-spatial correlation confidence value in the projection point set is determined as the spatial location of the corrosion product film failure point in this microturbulent event, and this location is traced back to the corrosion source location, serving as an important basis for pipeline health status assessment and predictive maintenance decisions.

[0075] Example 2: Based on the same inventive concept, such as Figure 2 As shown, this embodiment also provides a pipeline corrosion source tracing system based on multi-source data fusion, the system comprising: The sensing node data acquisition module is used to deploy flow field sensing nodes and electrochemical sensing nodes at the structural abrupt changes in the pipeline and at its upstream and downstream locations, respectively, to collect the flow field time series of each flow field sensing node and the electrochemical time series of each electrochemical sensing node.

[0076] The macroscopic turbulence disturbance mode time series extraction module is used to extract the macroscopic turbulence disturbance mode time series at locations of structural abrupt changes based on the spatial distribution characteristics of the flow field time series at different sensing nodes.

[0077] The residual flow field time series acquisition module is used to perform adaptive filtering on the flow field time series at the structural abrupt change location, with the macroscopic turbulence disturbance mode time series as a reference, to obtain the residual flow field time series.

[0078] The microturbulence event identification module is used to match transient events in time synchronization between the residual flow field time series and the electrochemical time series. When a transient disturbance is detected in the residual flow field time series and a synchronous transient response is generated in the electrochemical time series, the matching event is identified as a microturbulence event induced by the damage of the corrosion product film.

[0079] The corrosion source location determination module is used to calculate the spatial location of the transient disturbance source based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes; the projection of the spatial location of the flow field source on the pipe wall is determined as the spatial location of the corrosion product film rupture point that induces the microturbulent event, and traced back to the corrosion source location.

[0080] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0081] Finally, it should be noted that although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for tracing the source of pipeline corrosion based on multi-source data fusion, characterized in that, The method includes: Flow field sensing nodes and electrochemical sensing nodes were deployed at the structural abrupt change points of the pipeline, as well as upstream and downstream, to collect the flow field time series of each flow field sensing node and the electrochemical time series of each electrochemical sensing node. Based on the spatial distribution characteristics of the flow field time series at different sensing nodes, the time series of macroscopic turbulence disturbance modes at the locations of structural abrupt changes are extracted. Using the macroscopic turbulence disturbance mode time series as a reference, adaptive filtering is performed on the flow field time series at the structural abrupt change location to obtain the residual flow field time series; The residual flow field time series and the electrochemical time series are time-synchronized transient event matching. When a transient disturbance is detected in the residual flow field time series and a synchronous transient response is generated in the electrochemical time series, the matching event is identified as a microturbulent event induced by corrosion product film damage. Based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes, the spatial location of the flow field source of the transient disturbance is calculated; the projection of the spatial location of the flow field source on the pipe wall is determined as the spatial location of the corrosion product film rupture point that induces the microturbulent event, and traced back to the location of the corrosion source.

2. The pipeline corrosion source tracing method based on multi-source data fusion according to claim 1, characterized in that, The method for extracting the macroscopic turbulence disturbance mode time series at locations of structural abrupt changes based on the spatial distribution characteristics of the flow field time series at different sensing nodes includes: The time-frequency transformation of the flow field time series is performed to obtain the time spectrum of each sensing node; the spatial distribution of the time spectrum of different sensing nodes is decomposed into a specific spatial pattern. Based on the time spectrum corresponding to a specific spatial pattern, the flow field time series characterized only by the specific spatial pattern is reconstructed to obtain the macroscopic turbulence disturbance pattern time series at the structural abrupt change location.

3. The pipeline corrosion source tracing method based on multi-source data fusion according to claim 2, characterized in that, The method for obtaining a specific spatial pattern by pattern decomposing the spatial distribution of the temporal spectrum of different sensing nodes includes: A space-frequency matrix reflecting the distribution of time-frequency energy along the pipeline axis is constructed based on the time spectrum of each sensing node and the spatial location of the sensing nodes. Perform eigenorthogonal decomposition on the space-frequency matrix to obtain a set of spatial modes and the energy contribution rate corresponding to each spatial mode; From the set of spatial modes, the spatial mode that has the largest energy contribution rate at the structural abrupt change site and shows a monotonically decreasing trend along the downstream direction is selected as a specific spatial mode.

4. The pipeline corrosion source tracing method based on multi-source data fusion according to claim 3, characterized in that, The method for obtaining the residual flow field time series by performing adaptive filtering on the flow field time series at locations of structural abrupt changes, using the macroscopic turbulence disturbance mode time series as a reference, includes: The time series of the macroscopic turbulence disturbance mode is used as the reference input signal of the adaptive filter, and the time series of the flow field at the structural abrupt change is used as the main input signal of the adaptive filter. The adaptive filter iteratively updates its filter weight coefficients based on the correlation between the reference input signal and the main input signal using the minimum mean square error criterion, so as to cancel the signal component characterized by the macroscopic turbulence interference mode from the main input signal, and uses the error signal output by the adaptive filter as the residual flow field time series.

5. The pipeline corrosion source tracing method based on multi-source data fusion according to claim 1, characterized in that, The method for calculating the spatial location of the transient disturbance source based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes includes: The arrival time and amplitude of transient disturbances received by each downstream flow field sensing node in the same identified microturbulence event are obtained. The arrival time delay difference between two flow field sensing nodes is obtained based on the spatial location of any two flow field sensing nodes and the arrival time. A time delay difference equation set consisting of multiple sets of arrival time delay differences is established. Based on the attenuation mapping relationship satisfied by the ratio of the disturbance amplitude of any two flow field sensing nodes and the ratio of the distance from the flow field source of the transient disturbance to the two flow field sensing nodes, an amplitude attenuation gradient constraint equation is established. By jointly solving the time delay difference equation set and the amplitude attenuation gradient constraint equation, the spatial location of the transient disturbance source in the fluid medium inside the pipe is obtained.

6. The pipeline corrosion source tracing method based on multi-source data fusion according to claim 5, characterized in that, The method for establishing the amplitude attenuation gradient constraint equation based on the attenuation mapping relationship satisfied by the ratio of the disturbance amplitude of any two flow field sensing nodes and the ratio of the distances from the flow field source of the transient disturbance to the two flow field sensing nodes includes: Obtain the propagation attenuation law of the transient disturbance in the fluid medium inside the pipe, and establish an attenuation model of the disturbance amplitude as a function of propagation distance based on the propagation attenuation law; Substituting the disturbance amplitudes of any two flow field sensing nodes into the attenuation model yields an amplitude ratio expression with the distance from the flow field source to the two flow field sensing nodes as the unknown. The amplitude ratio expression is used to establish an equation with the measured ratio of the disturbance amplitudes of the two flow field sensing nodes, which is then transformed into an attenuation mapping relationship corresponding to the ratio of the disturbance amplitudes of the two flow field sensing nodes and the ratio of the distances from the flow field source to the two flow field sensing nodes. Multiple sets of attenuation mapping relationships corresponding to the flow field sensing node pairs are combined to form an amplitude attenuation gradient constraint equation.

7. The pipeline corrosion source tracing method based on multi-source data fusion according to claim 5, characterized in that, The method for obtaining the spatial location of the transient disturbance source in the fluid medium within the pipe by jointly solving the time delay difference equation set and the amplitude attenuation gradient constraint equation includes: By combining the time delay difference equations and the amplitude attenuation gradient constraint equations, a joint solution model is constructed with the spatial location of the flow field source as the unknown variable. Using the distance ratio from the flow field source to each flow field sensing node determined by the amplitude attenuation gradient constraint equation as the constraint condition, the time delay difference equation set is constrained to obtain the constrained time delay difference equation set. An iterative optimization algorithm is used to solve the constrained time delay difference equations. In each iteration, the theoretical arrival time delay to each flow field sensing node is calculated based on the current estimated spatial location of the flow field source. The residual between the theoretical arrival time delay and the measured arrival time is used to update the current estimated spatial location of the flow field source until the preset convergence condition is met, thereby obtaining the spatial location of the flow field source of the transient disturbance in the fluid medium inside the pipe.

8. The pipeline corrosion source tracing method based on multi-source data fusion according to claim 1, characterized in that, The method for determining the spatial location of the corrosion product film rupture point that induces microturbulence events by projecting the spatial location of the flow field source onto the pipe wall includes: Obtain the three-dimensional geometric model of the inner wall of the pipeline at the structural abrupt change location, as well as the layout coordinates of the electrochemical sensing nodes that generate synchronous transient responses; Based on the three-dimensional geometric model of the inner wall, the spatial location of the transient disturbance source is calculated and projected perpendicularly onto the inner wall of the pipe to obtain the set of projection points of the spatial location of the flow field source on the inner wall of the pipe. Based on the correlation between each projection point in the projection point set and the electrochemical sensing node that generates a synchronous transient response, the spatial location of the corrosion product film damage point that induces the microturbulent event is determined from the projection point set.

9. The pipeline corrosion source tracing method based on multi-source data fusion according to claim 8, characterized in that, The method for determining the spatial location of the corrosion product film damage point that induces microturbulent events from the set of projection points based on the correlation between each projection point in the set of projection points and the electrochemical sensing node that generates a synchronous transient response includes: Acquire the deployment coordinates of the electrochemical sensing nodes that generate synchronous transient responses and the electrochemical transient characteristics of the synchronous transient responses received by the electrochemical sensing nodes. Calculate the spatial distance between each projection point in the projection point set and the layout coordinates of the electrochemical sensing node, and construct the electrochemical-spatial correlation confidence of each projection point based on the spatial distance and electrochemical transient characteristics; The projection point corresponding to the maximum value of the electrochemical-spatial correlation confidence in the projection point set is determined as the spatial location of the corrosion product film damage point that induces microturbulence events.

10. A pipeline corrosion source tracing system based on multi-source data fusion, characterized in that, The system includes: The sensing node data acquisition module is used to deploy flow field sensing nodes and electrochemical sensing nodes at the structural change points of the pipeline and at its upstream and downstream locations, respectively, to collect the flow field time series of each flow field sensing node and the electrochemical time series of each electrochemical sensing node. The macroscopic turbulence disturbance mode time series extraction module is used to extract the macroscopic turbulence disturbance mode time series at the locations of structural abrupt changes based on the spatial distribution characteristics of the flow field time series at different sensing nodes. The residual flow field time series acquisition module is used to perform adaptive filtering on the flow field time series at the structural abrupt location, with the macroscopic turbulence disturbance mode time series as a reference, to obtain the residual flow field time series. The microturbulence event identification module is used to match transient events in time synchronization between the residual flow field time series and the electrochemical time series. When a transient disturbance is detected in the residual flow field time series and a synchronous transient response is generated in the electrochemical time series, the matching event is identified as a microturbulence event induced by the damage of the corrosion product film. The corrosion source location determination module is used to calculate the spatial location of the transient disturbance source based on the time delay difference and amplitude attenuation gradient of the transient disturbance corresponding to the identified microturbulent event reaching different downstream flow field sensing nodes; the projection of the spatial location of the flow field source on the pipe wall is determined as the spatial location of the corrosion product film rupture point that induces the microturbulent event, and traced back to the corrosion source location.

Images