Inertia-based method and system for measuring bending strain of pipeline, and device and medium

Abstract

An inertia-based method and system for measuring the bending strain of a pipeline, and a device and a medium. The method comprises: acquiring posture information of a pipeline having a preset length,

Description

Inertia-based pipeline bending strain detection method, system, equipment and medium Technical Field

[0001] The present invention relates to the technical field of oil and gas pipeline detection, and in particular to an inertia-based pipeline bending strain detection method, system, equipment and medium. Background Art

[0002] Due to their high pressure and long distances, pipelines are subject to displacement and deformation due to external forces, such as geological disasters (such as earthquakes, landslides, permafrost thaw or uplift), and other third-party damage. These deformations and displacements subject the pipelines to normal internal pressure loads and bending strain. The presence of bending strain seriously affects the structural integrity and operational safety of the pipelines. In particular, severe defects that induce bending strain are more likely to lead to failures. Therefore, the inspection of long-distance pipelines for bending strain has become a key focus for pipeline operators in recent years, and is of great significance for preventing accidents and ensuring pipeline safety.

[0003] The most important method for pipeline defect detection is online pipeline inspection. Compared with external pipeline inspection, it has the advantages of low cost, high efficiency, and identifiability. The IIT robot uses the conveying medium as the driving force to perform non-destructive inspections on pipelines for deformation, corrosion, cracks, etc., providing a scientific basis for pipeline operation, maintenance and safety assessment. The IIT robot uses an IMU (Inertial Measurement Unit) based on strapdown inertial navigation technology to map the centerline coordinates and calculate bending strain. However, the leather cup or support wheel of the ILI robot will encounter the pipe wall, girth weld or spiral weld and other features, resulting in slight oscillations and vibrations during the inspection process, and the posture information calculated by the IMU will be affected by these noises.

[0004] Summary of the Invention

[0005] In order to overcome the problem that the posture information calculated by the IMU is affected by noise, the present invention provides an inertia-based pipeline bending strain detection method, system, device and medium.

[0006] In a first aspect, in order to solve the above technical problems, the present invention provides a pipeline bending strain detection method based on inertia, comprising the following steps:

[0007] Get the attitude information of the pipeline of preset length collected by IMU;

[0008] Perform wavelet transform on the attitude information to remove the influence of high-frequency noise from system noise and vibration in the IMU, and obtain the first target attitude information;

[0009] Sliding and filtering the first target posture information according to a preset order and a preset window width through a sliding window, wherein each sliding of the sliding window on the first target posture information determines a data window;

[0010] According to each data window, the fitting coefficient is determined by the minimum mean square error;

[0011] The value at the center point of each data window is used as the filtering value, and the filtering sequence is determined according to each filtering value and the fitting coefficient. The filtering sequence is used as the second target attitude information. The second target attitude information is the information that removes the influence of other noises except the high-frequency noise from the system noise and vibration in the IMU;

[0012] determining the pipe bending strain force according to the second target posture information;

[0013] The pipeline bending strain is detected according to the pipeline bending strain force.

[0014] In a second aspect, the present invention provides an inertia-based pipeline bending strain detection system, comprising:

[0015] The attitude information acquisition module is used to obtain the attitude information of the pipeline of a preset length collected by the IMU;

[0016] A first target attitude information determination module is used to perform wavelet transform on the attitude information to remove the influence of high-frequency noise from system noise and vibration in the IMU to obtain the first target attitude information;

[0017] a data window determination module, configured to perform sliding filtering on the first target posture information according to a preset order and a preset window width through a sliding window, wherein each time the sliding window slides on the first target posture information, a data window is determined;

[0018] A fitting coefficient determination module is used to determine the fitting coefficient according to each data window through the minimum mean square error;

[0019] A fitting curve determination module, used for determining a fitting curve according to each data window and a fitting coefficient;

[0020] A second target attitude information determination module is configured to use the value at the center point of each data window as a filter value, determine a filter sequence based on each filter value and a fitting coefficient, and use the filter sequence as the second target attitude information, wherein the second target attitude information is the information from which the influence of other noises except for high-frequency noise from system noise and vibration in the IMU is removed;

[0021] a pipeline bending strain force determination module, configured to determine the pipeline bending strain force according to the second target posture information;

[0022] The detection module is used to detect the pipeline bending strain according to the pipeline bending strain force.

[0023] In a third aspect, the present invention further provides a computing device comprising a memory, a processor, and a program stored in the memory and running on the processor, wherein when the processor executes the program, the steps of the above-mentioned inertia-based pipeline bending strain detection method are implemented.

[0024] In a fourth aspect, the present invention further provides a computer-readable storage medium, in which instructions are stored. When the instructions are executed on a terminal device, the terminal device executes the steps of the inertia-based pipeline bending strain detection method.

[0025] The inertia-based pipeline bending strain detection system provided by the present invention has the following beneficial effects: by performing wavelet transform on attitude information, the influence of high-frequency noise from system noise and vibration in the IMU is removed to obtain first target attitude information, and then a data window is obtained by sliding a sliding window on the first target attitude information, and the corresponding value at the center point of the data window is used as a filter value, and a filter sequence is constructed in combination with a fitting coefficient, and the filter sequence is used as second target attitude information, thereby removing the influence of other noises except the high-frequency noise from system noise and vibration in the IMU. Finally, the pipeline bending strain force is determined based on the second target attitude information, and the pipeline bending strain is detected based on the pipeline bending strain force. The noise is removed twice, the accuracy of the attitude information calculated by the IMU is improved, and the problem that the attitude information calculated by the IMU is affected by noise is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention is further described below with reference to the accompanying drawings and embodiments.

[0027] FIG1 is a schematic flow chart of a method for detecting pipe bending strain based on inertia according to an embodiment of the present invention;

[0028] Figure 2 is a schematic diagram of the SG filter;

[0029] Figure 3 shows the original posture information obtained by the IIT robot before noise processing;

[0030] Figure 4 is a bending strain diagram calculated before noise processing;

[0031] Figure 5 is a diagram of the target posture information obtained by the IIT robot after noise removal;

[0032] Figure 6 is a bending strain diagram calculated after noise processing;

[0033] Figure 7 is a comparison of target posture information under different filtering methods;

[0034] Figure 8 is a comparison of bending strain under different filtering methods;

[0035] Figure 9 is a comparison of bending strain forces of straight sinking and different sinking before denoising using different filtering methods;

[0036] FIG10 is a diagram showing different displacements of bending strain under the method of this embodiment;

[0037] FIG11 is a comparison of the absolute deviation of bending strain force using different filtering methods;

[0038] FIG12 is a schematic structural diagram of a pipeline bending strain detection system based on inertia according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following examples are provided to further explain and supplement the present invention and do not constitute any limitation to the present invention.

[0040] The following describes the inertia-based pipeline bending strain detection method, system, device, and medium according to embodiments of the present invention with reference to the accompanying drawings.

[0041] As shown in FIG1 , an embodiment of the present invention provides a pipeline bending strain detection method based on inertia, which includes the following steps:

[0042] S1. Obtain the posture information of a pipeline of a preset length collected by the IMU.

[0043] The preset length can be adjusted according to actual conditions.

[0044] S2. Perform wavelet transform on the attitude information to remove the influence of high-frequency noise from system noise and vibration in the IMU, and obtain the first target attitude information.

[0045] The first target posture information is obtained through the Symlets wavelet transform filter.

[0046] S3. Slidingly filtering the first target posture information according to a preset order and a preset window width through a sliding window, wherein each time the sliding window slides on the first target posture information, a data window is determined.

[0047] S4. Determine the fitting coefficients based on the minimum mean square error for each data window.

[0048] S5. The value at the center point of each data window is used as a filter value. A filter sequence is determined according to each filter value and the fitting coefficient. The filter sequence is used as the second target attitude information. The second target attitude information is the information that removes the influence of other noises except the high-frequency noise from the system noise and vibration in the IMU.

[0049] The second target posture information is obtained through the SG (Savitzky-Golay) filter.

[0050] In addition to high-frequency noise from system noise and vibration in the IMU, other noises include girth weld noise, spiral weld noise, etc., which are added to the posture information of the ILI robot during long-distance and long-time detection.

[0051] S6. Determine the pipeline bending strain force according to the second target posture information.

[0052] S7. Detect the pipeline bending strain according to the pipeline bending strain force.

[0053] In this embodiment, wavelet transform is performed on the attitude information to remove the influence of high-frequency noise from system noise and vibration in the IMU, thereby obtaining first target attitude information. A sliding window is then slid over the first target attitude information to obtain a data window. The corresponding value at the center point of the data window is used as a filter value, and a filter sequence is constructed in combination with a fitting coefficient. The filtered sequence is used as the second target attitude information, thereby removing the influence of other noises except for the high-frequency noise from system noise and vibration in the IMU. Finally, the pipeline bending strain force is determined based on the second target attitude information, and the pipeline bending strain is detected based on the pipeline bending strain force. The noise is removed twice, thereby improving the accuracy of the attitude information calculated by the IMU and solving the problem that the attitude information calculated by the IMU is affected by noise.

[0054] Optionally, performing a wavelet transform on the posture information to remove the influence of high-frequency noise from system noise and vibration in the IMU to obtain first target posture information includes: performing a continuous wavelet transform or a discrete wavelet transform on the posture information to obtain the first target posture information.

[0055] Both continuous and discrete wavelet transforms are implemented using Symlets wavelet transform filters. The wavelet transform is a time-frequency signal analysis method with excellent positioning properties in both the time and frequency domains. Wavelets consist of a series of wavelet basis functions that can describe the local characteristics of a signal in both the time and frequency domains. It can perform multi-scale analysis of signals and analyze them in either the time or spatial domain. High frequency resolution is achieved in high-frequency portions of the signal, while high temporal resolution is achieved in low-frequency portions. Consequently, information loss is minimized when processing signals with minimal discontinuities. This reduces phase distortion during signal decomposition and reconstruction, thereby removing the effects of high-frequency noise from system noise and vibration within the IMU.

[0056] Optionally, continuous wavelet transform is performed on the posture information to obtain the first target posture information, and the formula is as follows:

[0057] Among them, CWT x (a, b) represents the first target posture information, x(t) represents the posture information, ψ * (x) represents the mother wavelet function, a represents the scale factor, b represents the translation factor, N represents the order of the wavelet transform filter, and w k represents the coefficient of the wavelet transform filter, φ(t) represents the scaling function, Q(z) represents a preset polynomial, and the result of W(z) is used as w k value.

[0058] Optionally, discrete wavelet transform is performed on the posture information to obtain the first target posture information:

[0059] Among them, DWT x (j,k) represents the first target posture information, j and k represent integer parameters, x(t) represents the posture information, ψ * (x) represents the mother wavelet function, a represents the scale factor, b represents the translation factor, N represents the order of the wavelet transform filter, and w k represents the coefficient of the wavelet transform filter, φ(t) represents the scaling function, Q(z) represents a preset polynomial, and the result of W(z) is used as w k value.

[0060] Whether using continuous wavelet transform or discrete wavelet transform, the mother wavelet function must meet the following conditions:

[0061] (1) Finite support: ψ(t) = 0, when t < 0 or t > 1.

[0062] (2) Orthogonality: where δ(k) is the Dirac function.

[0063] (3) Zero mean:

[0064] (3) High-pass filtering: in is the Fourier transform of ψ(t).

[0065] In addition, the Symlet wavelet is derived from the Daubechies wavelet. They are both generated by a low-pass filter W, which can be decomposed into:

[0066] Where N is the order of the filter, and Q(z) is a polynomial. The Daubechies wavelet selects the minimum phase root of Q(z) as the filter coefficient, while the Symlet wavelet selects the closest linear phase root of Q(z) as the filter coefficient. In this way, the Symlet wavelet has better symmetry and approximate linear phase, while the Daubechies wavelet has better frequency selectivity and faster decay.

[0067] Optionally, the first target posture information is slidingly filtered according to a preset order and a preset window width through a sliding window, and a data window is determined each time the sliding window slides on the first target posture information. The formula is as follows:

[0068] Among them, y(n) represents the nth data window, D represents the total number of sliding window sliding, n k represents an n-order polynomial, a k Indicates the kth slide.

[0069] The preset order and preset window width can be adjusted according to actual conditions. In this embodiment, the preset window width is 2M+1, where M represents the preset value. A large low-order sliding window will cause signal distortion and widen the absorption line type, making it difficult to retain the required information. A high-order small sliding window can better retain signal information, but its filtering effect on noise is also weaker. Therefore, it is necessary to select a suitable window width.

[0070] Optionally, the fitting coefficients are determined based on the minimum mean square error for each data window, including:

[0071] Among them, ε D Represents the fitting coefficient, x[n] represents the actual value of the nth data window, and M represents the number of data windows.

[0072] The purpose of determining the fitting coefficient by the minimum mean square error is to reduce the impact of the error and improve the accuracy of the fitting curve when performing weighted value fitting on the filter values ​​corresponding to each data window.

[0073] As shown in FIG2, the value y(i) corresponding to the center point of each data window (each black point in FIG1) is i As the filtered value, y(i) represents the filtered value, a i Represents the i-th window data, and each filter value is fitted by the fitting coefficient to obtain the filter sequence y[n] (i.e., the fitting curve in Figure 1), and y[n] is used as the second target posture information.

[0074] Optionally, the pipeline bending strain force is determined according to the second target posture information, and the formula is as follows:

[0075] Where ε represents the pipe bending strain, ε v , ε h They represent the vertical and horizontal strain components of the pipeline, D represents the nominal diameter of the pipeline, k v 、k h They represent the vertical curvature and horizontal curvature of the pipeline respectively, k represents the total curvature of the pipeline, P is the pitch angle, s is the mileage, A is the heading angle, and k v 、k h , k both come from the second target posture information.

[0076] Since the bending strain of the pipeline is proportional to the curvature change within the elastic deformation range, the curvature of the pipeline should be calculated first when calculating the bending strain of the pipeline. In the calculation of pipeline positioning, in addition to obtaining the pipeline coordinates under the UTM (Mercator projection system Universal Test Message) projection, the second target posture information of the IIT robot in the local coordinate system is also obtained. Among them, the pitch change of the IIT robot represents the change in the inclination angle of the pipeline relative to the horizontal plane in the fixed observation pipeline, and the heading angle represents the angle between the pipeline along the line and the north direction. Therefore, the curvature of the pipeline and the second target posture information have the following relationship:

[0077] The bending strain of a pipe is directly related to its curvature. It has two main components: circumferential strain and longitudinal strain. The longitudinal strain is further divided into axial and bending parameters. If the centerline of the pipe is assumed to be the neutral axis, the relationship between the bending strain and the centerline curvature is as follows:

[0078] If the pipeline is subjected to true bending strain, the range of its bending influence should be large, and features such as girth welds will cause local strain changes. Therefore, the strain changes caused by girth welds should be filtered out as the judgment condition for adaptive filtering. When obtaining the preliminary bending strain results, if the bending strain is greater than 0.125% within a range of 3m, the order of the SG filter is modified, and the adaptive SG filter processes the noise posture information until the bending strain meets the requirements.

[0079] Optionally, in order to better illustrate this embodiment, a specific example is used for illustration:

[0080] Nine pipelines of approximately 90 m in length were installed on five buttresses. Strain gauges were installed in the center of the pipelines. When the strain variation is the largest, it is obvious when comparing the bending strain detected by the IIT robot. The bending strain of straight lines and different pipeline displacements was analyzed. All pulling speeds tested were 1 m / s.

[0081] As shown in Figures 3 and 4, the IIT robot is subject to various vibrations during the inspection process, such as those caused by girth welds and spiral welds. This noise can interfere with posture information, such as tilt, and inaccurately calculates bending strain. During the inspection process, the IIT robot experiences three types of noise. High-frequency noise signals primarily originate from the IIT robot's internal noise and vibrations during the inspection process. Small periodic spikes are primarily due to the minor impact of the IIT robot passing through the pipeline's spiral welds. There are also some large periodic spikes, primarily due to the significant impact of the IIT robot passing through girth welds. Therefore, effective methods should be used to reduce the noise of posture information.

[0082] Using the method of this embodiment to process raw posture information, as shown in Figures 5 and 6, Figure 5 shows that not only does it eliminate the constant-frequency noise caused by the vibration and impact of spiral welds on the IIT robot during inspection, but it also reduces the posture changes caused by impact when the IIT robot passes through girth welds. The calculated bending strain before and after filtering is shown in Figure 6. The bending strain eliminates noise caused by vibration and pipe weld impact, and is more accurate for straight pipes.

[0083] To validate the proposed method, different filtering methods were compared. Figure 7 shows a comparison of target posture information obtained by denoising the original posture information using a wavelet transform filter alone and the method of this embodiment. A comparison of bending strain is shown in Figure 8. It can be seen that the method of this embodiment achieves better denoising results for the original posture information and more accurate bending strain.

[0084] To further validate the accuracy of the proposed method for measuring bending strain changes caused by pipe bending, strain gauges were used to compare measurement results from an IIT robot installed in the center of a pull-through pipe. The pull-through pipe was subjected to different loads and displacements of 5, 10, 15, and 26.5 cm to simulate varying degrees of bending deformation. These varying bending deformations resulted in varying bending strains in the pipe.

[0085] Figure 9 shows a comparison of bending strain for straight sinking and different sinking conditions before denoising. Due to the noise, it is difficult to obtain the true change in bending strain. The bending strain for different displacements is shown in Figure 10. It can be clearly seen that the bending strain changes accordingly when the pipe has different bending deformations.

[0086] To verify the effectiveness of the method of this embodiment, the pull-through test data was calculated and compared using the cubic spline interpolation filtering algorithm (CSP), the Symlets wavelet (SW) filtering method, and the method of this embodiment. Table 1 shows a comparison of bending strain using different filtering methods. According to Table 1, when the pipeline sinks 5 cm, the strain gauge changes by 0.011%. Before filtering the data, the calculated strain change in bending strain was 0.0007%, with an absolute deviation of 0.0040%. When using the three filtering methods, the calculated strain change in bending strain was 0.015%, 0.014%, and 0.014%, respectively. The absolute deviations were 0.004%, 0.003%, and 0.003%, respectively.

[0087] For a pipe sinking 10 cm, the strain gauge change is 0.021%. Before data filtering, the calculated strain change in the PBS is 0.0362%, with an absolute deviation of 0.0152%. Using the three filtering methods, the calculated strain changes in the PBS are 0.0278%, 0.0263%, and 0.025%, respectively. The absolute deviations are 0.0068%, 0.0053%, and 0.004%, respectively.

[0088] For a pipe sinking 15 cm, the strain gauge change is 0.032%. Before data filtering, the PBS strain change is calculated to be 0.033%, with an absolute deviation of 0.001%. Using the three filtering methods, the PBS strain changes are calculated to be 0.029%, 0.0342%, and 0.033%, respectively. The absolute deviations are 0.003%, 0.0022%, and 0.001%, respectively.

[0089] For a pipe sinking 26.5 cm, the strain gauge change is 0.055%. Before data filtering, the PBS strain change is calculated to be 0.0778%, with an absolute deviation of 0.0228%. Using the three filtering methods, the PBS strain changes are calculated to be 0.0712%, 0.0701%, and 0.068%, respectively. The absolute deviations are 0.0162%, 0.0151%, and 0.013%, respectively.

[0090] Table 1 Comparison of bending strain and different filtering methods

[0091] Wherein, Pipeline Sank indicates pipeline sinking, Strain changes for gauge measurement indicates changes in strain gauges, Strain changes indicates strain changes in bending strain force, Absolute Deviation indicates absolute deviation, CSP indicates cubic spline interpolation filtering algorithm, SW indicates Symlets wavelet (SW) filtering method, and W-SG indicates the method of this embodiment.

[0092] A comparison of the absolute deviations of bending strain under different filtering methods is shown in Figure 11. The average absolute deviation before data filtering is 0.0108%. When using the three filtering methods, the average absolute deviations of bending strain are 0.0075%, 0.0064%, and 0.0053%, respectively, and the average relative deviations are 26.9%, 21.7%, and 18.2%, respectively. Clearly, the method of this embodiment offers superior performance and is more accurate in calculating bending strain. Based on the pull-through tests and results, the method proposed in this invention can achieve superior performance in calculating bending strain.

[0093] As shown in FIG12 , an embodiment of the present invention further provides an inertia-based pipeline bending strain detection system, comprising:

[0094] The posture information acquisition module 101 is used to obtain the posture information of the pipeline of a preset length collected by the IMU;

[0095] A first target attitude information determination module 102 is configured to perform a wavelet transform on the attitude information to remove the influence of high-frequency noise from system noise and vibration in the IMU to obtain first target attitude information;

[0096] A data window determining module 103 is configured to perform sliding filtering on the first target posture information using a sliding window according to a preset order and a preset window width, wherein each sliding of the sliding window on the first target posture information determines a data window;

[0097] A fitting coefficient determination module 104 is used to determine the fitting coefficient according to each data window by using the minimum mean square error;

[0098] The second target attitude information determination module 105 is configured to use the value at the center point of each data window as a filter value, determine a filter sequence based on each filter value and a fitting coefficient, and use the filter sequence as the second target attitude information, wherein the second target attitude information is the information from which the influence of other noises except for high-frequency noise from system noise and vibration in the IMU is removed;

[0099] a pipeline bending strain force determination module 106, configured to determine the pipeline bending strain force according to the second target posture information;

[0100] The detection module 107 is used to detect the pipeline bending strain according to the pipeline bending strain force.

[0101] Optionally, the first target posture information determining module 102 is specifically configured to:

[0102] Performing continuous wavelet transform or discrete wavelet transform on the posture information to obtain first target posture information.

[0103] Optionally, the first target posture information determining module 102 is specifically configured to:

[0104] Perform continuous wavelet transform on the posture information to obtain the first target posture information. The formula is as follows:

[0105] Among them, CWT x (a, b) represents the first target posture information, x(t) represents the posture information, ψ * (x) represents the mother wavelet function, a represents the scale factor, b represents the translation factor, N represents the order of the wavelet transform filter, and w k represents the coefficient of the wavelet transform filter, φ(t) represents the scaling function, Q(z) represents a preset polynomial, and the result of W(z) is used as w k value.

[0106] Optionally, the first target posture information determining module 102 is specifically configured to:

[0107] Perform discrete wavelet transform on the posture information to obtain the first target posture information:

[0108] Among them, DWT x (j,k) represents the first target posture information, j and k represent integer parameters, x(t) represents the posture information, ψ * (x) represents the mother wavelet function, a represents the scale factor, b represents the translation factor, N represents the order of the wavelet transform filter, and w k represents the coefficient of the wavelet transform filter, φ(t) represents the scaling function, Q(z) represents a preset polynomial, and the result of W(z) is used as w k value.

[0109] Optionally, the data window determination module 103 is specifically configured to:

[0110] The first target posture information is filtered by sliding the sliding window according to a preset order and a preset window width. Each time the sliding window slides on the first target posture information, a data window is determined. The formula is as follows:

[0111] Among them, y(n) represents the nth data window, D represents the total number of sliding window sliding, n k represents an n-order polynomial, a k Indicates the kth slide.

[0112] Optionally, the fitting coefficient determination module 104 is specifically configured to:

[0113] Based on each data window, the fitting coefficients are determined by the minimum mean square error, including:

[0114] Among them, ε D Represents the fitting coefficient, x[n] represents the actual value of the nth data window, and M represents the number of data windows.

[0115] Optionally, the pipeline bending strain determination module 106 is specifically configured to:

[0116] The pipe bending strain is determined based on the second target posture information. The formula is as follows:

[0117] Where ε represents the pipe bending strain, ε v , ε h They represent the vertical and horizontal strain components of the pipeline, D represents the nominal diameter of the pipeline, k v 、k h They represent the vertical curvature and horizontal curvature of the pipeline respectively, k represents the total curvature of the pipeline, P is the pitch angle, s is the mileage, A is the heading angle, and k v 、k h , k both come from the second target posture information.

[0118] A computing device according to an embodiment of the present invention includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, some or all steps of the above-mentioned inertia-based pipeline bending strain detection method are implemented.

[0119] Among them, the computing device can be a computer, and correspondingly, its program is computer software. The above-mentioned parameters and steps in a computing device of the present invention can refer to the parameters and steps in the embodiment of the inertia-based pipeline bending strain detection method above, and are not repeated here.

[0120] Those skilled in the art will appreciate that the present invention may be implemented as a system, method, or computer program product. Therefore, the present disclosure may be specifically implemented in the following forms, namely: in the form of complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, the present invention may also be implemented in the form of a computer program product in one or more computer-readable media, the computer-readable media containing computer-readable program code. Computer-readable storage media may be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination thereof.

[0121] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0122] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. The pipeline bending strain detection method based on inertia is characterized by: The steps include: Obtain the attitude information of the pipeline of preset length collected by IMU; Performing wavelet transform on the attitude information to remove the influence of high-frequency noise from system noise and vibration in the IMU, and obtaining first target attitude information; Sliding and filtering the first target posture information according to a preset order and a preset window width through a sliding window, wherein the sliding window determines a data window each time it slides on the first target posture information; According to each data window, the fitting coefficient is determined by the minimum mean square error; Taking the value at the center point of each of the data windows as a filtering value, determining a filtering sequence according to each of the filtering values ​​and the fitting coefficient, and taking the filtering sequence as the second target attitude information, wherein the second target attitude information is the information from which the influence of other noises except the high-frequency noise from the system noise and vibration in the IMU is removed; determining a pipe bending strain force according to the second target posture information; The pipeline bending strain is detected according to the pipeline bending strain force.

2. The method according to claim 1, characterized in that Performing wavelet transform on the attitude information to remove the influence of high-frequency noise from system noise and vibration in the IMU to obtain first target attitude information, including: Perform continuous wavelet transform or discrete wavelet transform on the posture information to obtain first target posture information.

3. The method according to claim 2, characterized in that The posture information is subjected to continuous wavelet transform to obtain the first target posture information, and the formula is as follows: Among them, CWT x (a, b) represents the first target posture information, x(t) represents the posture information, ψ * (x) represents the mother wavelet function, a represents the scale factor, b represents the translation factor, N represents the order of the wavelet transform filter, and w k represents the coefficient of the wavelet transform filter, φ(t) represents the scaling function, Q(z) represents a preset polynomial, and the result of W(z) is used as w k The value of .

4. The method according to claim 2, characterized in that: Perform discrete wavelet transform on the posture information to obtain the first target posture information: Among them, DWT x (j, k) represents the first target posture information, j and k represent integer parameters, x(t) represents the posture information, ψ * (x) represents the mother wavelet function, a represents the scale factor, b represents the translation factor, N represents the order of the wavelet transform filter, and w k represents the coefficient of the wavelet transform filter, φ(t) represents the scaling function, Q(z) represents a preset polynomial, and the result of W(z) is used as w k The value of .

5. The method according to claim 1, characterized in that: The first target posture information is subjected to sliding filtering by a sliding window according to a preset order and a preset window width, and the sliding window determines a data window each time it slides on the first target posture information, and the formula is as follows: Among them, y(n) represents the nth data window, D represents the total number of sliding window sliding, n k represents an n-order polynomial, a k Indicates the kth slide.

6. The method according to claim 5, characterized in that Based on each data window, the fitting coefficients are determined by the minimum mean square error, including: Among them, ε D represents the fitting coefficient, x[n] represents the actual value of the nth data window, and M represents the number of data windows.

7. The method according to any one of claims 1 to 5, characterized in that: The pipeline bending strain force is determined according to the second target posture information, and the formula is as follows: Where ε represents the pipe bending strain, ε v , ε h They represent the vertical and horizontal strain components of the pipeline, D represents the nominal diameter of the pipeline, and k v , k h They represent the vertical curvature and horizontal curvature of the pipeline respectively, k represents the total curvature of the pipeline, P is the pitch angle, s is the mileage, A is the heading angle, and k v , k h , k both come from the second target posture information.

8. The inertia-based pipeline bending strain detection system is characterized by: include: The posture information acquisition module is used to obtain the posture information of the pipeline of a preset length collected by the IMU; A first target attitude information determination module is used to perform wavelet transform on the attitude information to remove the influence of high-frequency noise from system noise and vibration in the IMU to obtain first target attitude information; A data window determination module, configured to perform sliding filtering on the first target posture information according to a preset order and a preset window width through a sliding window, wherein the sliding window determines a data window each time it slides on the first target posture information; A fitting coefficient determination module is used to determine the fitting coefficient according to each data window through the minimum mean square error; A second target attitude information determination module is used to use the value at the center point of each of the data windows as a filter value, determine a filter sequence according to each of the filter values ​​and the fitting coefficient, and use the filter sequence as the second target attitude information, wherein the second target attitude information is the information from which the influence of other noises except the high-frequency noise from the system noise and vibration in the IMU is removed; A pipeline bending strain force determination module, used to determine the pipeline bending strain force according to the second target posture information; The detection module is used to detect the pipeline bending strain according to the pipeline bending strain force.

9. A computing device comprising a memory, a processor, and a program stored in the memory and running on the processor, characterized in that: When the processor executes the program, the steps of the inertia-based pipeline bending strain detection method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device executes the steps of the inertia-based pipeline bending strain detection method according to any one of claims 1 to 7.

Images