Pipeline stress on-line detection method

Abstract

The invention discloses a pipeline stress on-line detection method, belongs to the technical field of pipeline detection, and relates to a full-link technical scheme of'detection target-stress type-characteristic data-special model 'which is reversely constructed from'multivariate detection target' (such as fatigue life evaluation, defect safety monitoring and sudden impact early warning) of pipeline safety operation. According to the method, multiple most suitable detection technologies are intelligently selected and fused, multi-physics sensing technologies such as optics (FBG), acoustics (ultrasound) and vibration (acceleration) are fused, and static stress, local peak stress and dynamic alternating stress can be synchronously captured for a composite target; according to the invention, targeted, differentiated and collaborative online monitoring of stress states of different stress types and different risk parts is realized, the inherent defects of single sensing dimension and existence of a monitoring blind area of a single technology (such as only magnetism or only vibration) are effectively overcome, and the refinement and intelligence level of pipeline safety management is improved.

Description

Technical Field

[0001] This invention relates to the field of pipeline inspection technology, and more specifically, to a method for online pipeline stress detection. Background Technology

[0002] As a critical infrastructure for transporting oil and gas, chemical media, heat energy, and water, pipelines are directly related to production safety, public safety, and environmental protection in terms of structural integrity. Under the action of various loads such as internal and external pressure, temperature, gravity, wind load, earthquake, and foundation settlement, pipelines will generate complex stress states. Excessive stress is one of the root causes of pipeline fatigue cracking, plastic deformation, and even sudden failure. Therefore, continuous and accurate monitoring of pipeline stress is a core technical link to achieve predictive maintenance, prevent major accidents, and ensure long-term safe operation.

[0003] Currently, the detection of pipeline stress in engineering practice mainly relies on the following traditional methods: resistance strain gauge method, fiber optic sensor, periodic offline ultrasonic testing, strong and weak magnetic field testing method, etc. For example, patent number CN113933381A discloses a pipeline stress detection method based on strong and weak magnetic field testing. This patent determines the degree of pipeline stress damage by processing the ratio of the collected strong magnetic signal and weak magnetic signal, and can then identify stress concentration areas and fatigue damage that has not yet formed volume defects.

[0004] The CN113933381A patent describes the classic approach of deeply optimizing a single detection technology. However, when this single detection technology is applied to online monitoring of complex working conditions, it often has some limitations. This is because pipeline stress manifests in multiple forms (static pressure stress, bending stress, vibration stress, and local peak stress). The relationship between each stress and the magnetization properties of the material (permeability and remanence) is not linear. A single magnetic signal is difficult to distinguish the type and source of these stresses and cannot provide a comprehensive, real-time, accurate, and direct picture of the complex stress state of the pipeline throughout its entire life cycle, which can serve different operation and maintenance decisions.

[0005] To address the practical technical shortcomings, this paper proposes an online pipeline stress detection method. This method establishes a targeted, differentiated, and collaborative online monitoring system based on multiple detection targets, which monitors the stress state of different stress types and risk locations in a way that breaks away from the traditional approach of "focusing on a single sensor". Summary of the Invention

[0006] The purpose of this invention is to address the inherent shortcomings of traditional monitoring methods, such as single monitoring targets, isolated technical means, and one-sided data. It aims to provide a novel online pipeline stress detection method. The core objective of this invention is to break away from the traditional approach of "centered on a single sensor" and establish a full-link technical solution that is guided by multiple detection targets and constructs "detection target → stress type → feature data → dedicated model" in reverse. This enables targeted, differentiated, and collaborative online monitoring of stress states under different stress types, different risk locations, and different time scales.

[0007] The objective of this invention can be achieved through the following technical solution: an online pipeline stress detection method, comprising the following steps: S1. Multi-target identification and decomposition: Receives and parses the pipeline engineering safety requirements of the pipeline to be tested input by the user, and decomposes them into one or more specific and operable detection targets; S2. Detection Strategy Oriented Matching: Obtain the detection target, and based on the preset expert knowledge base, determine one or more key detection parts that need to be focused on to achieve the detection target, as well as the stress type that plays a dominant role in the key detection parts, and mark the key monitoring parts as the parts to be tested. Based on the stress type, a matching detection method is selected from a variety of preset stress detection methods to generate a "test part - stress type - detection method" detection scheme for a specific detection target; S3. Deployment of testing procedures and data acquisition: Deploy physical sensors corresponding to the selected testing methods at the parts to be tested, and perform multi-channel data acquisition for multiple testing schemes to obtain pipeline stress data; S4. Data fusion preprocessing: Receive pipeline stress data for the same detection target, perform data fusion preprocessing, and obtain target feature data. S5. Target-oriented dedicated model analysis: Obtain the correlation data of each detection target, and input it together with the target feature data into the dedicated model corresponding to the detection target for stress assessment, determine whether the stress level of each detection target meets the requirements, and output the assessment results simultaneously.

[0008] Furthermore, in S1, the inspection targets include long-term service safety inspection of pipelines, local defect / damage inspection of pipelines, and sudden load inspection of pipelines, which are respectively labeled as inspection target one, inspection target two, and inspection target three.

[0009] Furthermore, in S2, for the first detection target, the part to be tested is determined to be a stress concentration area, the stress type is determined to be axial stress / bending stress, the detection method is selected as fiber optic grating sensor array, and a detection scheme of "stress concentration area - axial stress / bending stress - fiber optic grating sensor array" is generated. For the second detection target, the area to be inspected is determined to be a defect / damage area, the stress type is determined to be peak stress, the detection method is selected as ultrasonic guided wave / acoustic elastic ultrasound, and a detection scheme of "defect / damage area - peak stress - ultrasonic guided wave / acoustic elastic ultrasound" is generated. For the third detection target, the parts to be inspected are identified as sensitive parts of the piping system (pipe connection points, midpoints of long-span pipes), the stress type is identified as alternating stress / vibration stress, the detection method is selected as piezoelectric accelerometer array, and a detection scheme of "sensitive parts of piping system - alternating stress / vibration stress - piezoelectric accelerometer array" is generated.

[0010] Furthermore, in S3, the pipeline stress data includes: wavelength drift, ultrasonic waveform data, and vibration acceleration time-domain signal, which correspond to the pipeline stress data of each detection target.

[0011] Furthermore, the specific process of data fusion preprocessing in S4 includes: For the first detection target, the wavelength drift is converted into micro-strain, and real-time temperature compensation is performed using the built-in strain and temperature dual gratings. Then, based on Hooke's law and pipeline geometric parameters, the axial and bending stress values ​​are calculated, and the "time-stress" sequence of each detection point is output. For the second detection target, digital filtering and noise reduction are performed on the ultrasonic wave data to extract ultrasonic wave features: wave velocity change, signal attenuation coefficient, and amplitude ratio of specific modes; For the third detection target, the collected acceleration time-domain signal is transformed to obtain the vibration spectrum, and the dynamic alternating stress amplitude is calculated as a sudden dynamic feature. The target feature data for each detection target are corresponding to the "time-stress" sequence, ultrasonic shape characteristics, and sudden dynamic characteristics.

[0012] Furthermore, the dedicated models corresponding to each detection target are the fatigue damage accumulation model, the fracture mechanics and defect assessment model, and the dynamic response and event diagnosis model.

[0013] Furthermore, the specific process for conducting stress assessment includes: For the first detection target, the "time-stress" sequence and its associated data are input into the fatigue damage accumulation model to output the current cumulative damage degree D, predict the remaining life, and assess the overall stress level and fatigue life of the pipeline. For the second detection target, the ultrasonic shape characteristics and their associated data are input into the fracture mechanics and defect assessment model, and the safety margin FAD coordinates and crack propagation rate are output to assess the stress concentration and propagation risk at the defect or damage site. For the third detection target, the sudden dynamic characteristics and their associated data are input into the dynamic response and event diagnosis model to output the resonance risk level (high / medium / low), transient event type, and dynamic fatigue damage contribution rate. The transient event types include Class I (severe event), Class II (major event), and Class III (general event) to assess the dynamic stress hazards caused by sudden impact loads.

[0014] Furthermore, for the first detection target, when the cumulative damage at all detection points is less than the safety threshold and the predicted remaining life is greater than the preset planned life cycle, the overall stress level of the pipeline is determined to meet the requirements. For the second detection target, when the safety margin FAD coordinate is within the FAD safety zone of the judgment criterion and the safety margin is greater than the preset safety margin value, and the crack propagation rate is less than the preset propagation rate threshold, the defect is judged to be stable in the current environment and the stress level meets the requirements.

[0015] For the third detection target, if there is no high / medium resonance risk level, no Class I / II events, and the dynamic fatigue damage contribution rate is less than the preset dynamic load contribution threshold, it is determined that the stress level of the pipeline meets the requirements when subjected to sudden impact load.

[0016] Compared with the prior art, the advantages of this invention are: 1. Starting from the "multi-dimensional detection targets" (such as fatigue life assessment, defect safety monitoring, and sudden impact early warning) of pipeline safety operation, a full-link technical solution of "detection target → stress type → feature data → dedicated model" is constructed in reverse. Based on this, multiple most suitable detection technologies are intelligently selected and integrated. By integrating multi-physics field sensing technologies such as optical (FBG), acoustic (ultrasound), and vibration (accelerometer), static stress, local peak stress, and dynamic alternating stress can be captured simultaneously for composite targets. This enables targeted, differentiated, and collaborative online monitoring of stress states of different stress types and different risk locations. It effectively overcomes the inherent defects of single technology (such as only magnetic or only vibration) in terms of single perception dimension and monitoring blind spots, and improves the refinement and intelligence level of pipeline safety management.

[0017] 2. This invention also synchronously binds target feature data (real-time, dynamic monitoring results) with related data (design, historical, static known information) for dedicated model analysis. Essentially, it gives these feature signals specific engineering background and physical constraints, which greatly improves accuracy and reduces false alarms and missed alarms. Attached Figure Description

[0018] Figure 1 This is a flowchart of the modular method of the present invention; Figure 2 This is a flowchart of the overall method of the present invention. Detailed Implementation

[0019] 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. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] Example 1: This invention discloses an online method for detecting pipeline stress. Please refer to [link / reference]. Figures 1-2 This includes the following steps: S1. Multi-target identification and decomposition: Receives and parses the pipeline engineering safety requirements of the pipeline to be tested input by the user, and decomposes them into one or more specific and operable detection targets; S2. Detection Strategy Oriented Matching: Obtain the detection target, and based on the preset expert knowledge base, determine one or more key detection parts that need to be focused on to achieve the detection target, as well as the stress type that plays a dominant role in the key detection parts, and mark the key monitoring parts as the parts to be tested. Based on the stress type, a matching detection method is selected from a variety of preset stress detection methods to generate a "test part - stress type - detection method" detection scheme for a specific detection target; S3. Deployment of testing procedures and data acquisition: Deploy physical sensors corresponding to the selected testing methods at the parts to be tested, and perform multi-channel data acquisition for multiple testing schemes to obtain pipeline stress data; S4. Data fusion preprocessing: Receive pipeline stress data for the same detection target, perform data fusion preprocessing, and obtain target feature data; S5. Target-oriented dedicated model analysis: Obtain the correlation data of each detection target, and input it together with the target feature data into the dedicated model corresponding to the detection target for stress assessment, determine whether the stress level of each detection target meets the requirements, and output the assessment results simultaneously.

[0021] This paper presents a novel online pipeline stress detection method. Its core objective is to break away from the traditional approach of "focusing on a single sensor" and establish a comprehensive technical solution that is guided by multiple detection targets and constructs a complete chain of "detection target → stress type → feature data → dedicated model" in reverse. This solution enables targeted, differentiated, and collaborative online monitoring of stress states under different stress types, risk locations, and time scales.

[0022] Example 2: The following are the specific methods for each process step; In S1, the detection targets specifically include long-term service safety detection of pipelines, local defect / damage detection of pipelines, and sudden load detection of pipelines, which are respectively labeled as detection target one, detection target two, and detection target three.

[0023] In S2, for the first detection target, the area to be tested is determined to be a stress concentration area (near elbows, tees, and supports), the stress type is determined to be axial stress / bending stress, the detection method is selected as fiber optic grating sensor array, and a detection scheme of "stress concentration area - axial stress / bending stress - fiber optic grating sensor array" is generated. The FBG sensor is essentially glass fiber, which is resistant to electromagnetic interference, corrosion resistant, and has a long life (up to 20 years or more), making it very suitable for monitoring the entire life cycle of pipelines. For the second inspection target, the area to be inspected is determined to be a defect / damage area (weld, corrosion area, scratch), the stress type is determined to be peak stress, and the inspection method is ultrasonic guided wave / acoustic elastic ultrasound. The inspection scheme of "defect / damage area - peak stress - ultrasonic guided wave / acoustic elastic ultrasound" is generated. Under this target, the focus is on the "peak stress" at the defect tip or damage edge. For the third detection target, the parts to be inspected are identified as sensitive parts of the piping system (pipe connection points, midpoints of long-span pipes), the stress type is identified as alternating stress / vibration stress, and the detection method is selected as piezoelectric accelerometer array. The detection scheme of "sensitive parts of piping system - alternating stress / vibration stress - piezoelectric accelerometer array" is generated. Under this target, the focus is on the response of the entire piping system structure to dynamic loads.

[0024] In S3, the specific process for acquiring pipe stress data includes: The detection scheme for "stress concentration zone - axial stress / bending stress - fiber optic grating sensor array" for long-term service safety inspection of pipelines involves dividing the cross section of the stress concentration zone into 4-8 detection points symmetrically in the circumferential direction. Each detection point is equipped with one FBG sensor, which is connected in series to form a fiber optic network. Each FBG sensor point integrates strain and temperature dual gratings for temperature compensation and collects the wavelength drift of each FBG. The "defect / damage zone-peak stress-ultrasonic guided wave / acoustic elastic ultrasound" detection scheme for local defects / damage detection in pipelines involves symmetrically dividing detection points on both sides of the defect / damage zone. Ultrasonic transmitting and receiving transducers are arranged at each pair of detection points. Guided waves of a specific mode are emitted, and reflected signals passing through or reflecting the defect / damage zone are received. This signal is sampled by a high-speed data acquisition card at a fixed sampling rate and quantized into a digital signal to obtain discrete time series ultrasonic wave data. The detection scheme of "pipeline sensitive part - alternating stress / vibration stress - piezoelectric acceleration sensor array" for pipeline sudden load detection involves high-density deployment of multiple detection points in three or more directions at the part to be inspected, with each detection point set with a vibration acceleration, and high-speed acquisition of vibration acceleration time domain signals. Wavelength drift, ultrasonic waveform data, and vibration acceleration time-domain signals correspond to the pipe stress data for each detection target.

[0025] The specific process of data fusion preprocessing in S4 includes: For the first detection target, the wavelength drift of each FBG is received, the wavelength drift is converted into micro-strain, and real-time temperature compensation is performed using the built-in strain and temperature dual gratings. Then, based on Hooke's law and pipeline geometric parameters, the axial and bending stress values ​​are calculated, and the "time-stress" sequence of each detection point is output. For the second detection target, ultrasonic waveform data is received and digitally filtered and denoised (to eliminate environmental and system noise). Ultrasonic waveform features are extracted: wave velocity variation, signal attenuation coefficient, and specific mode amplitude ratio. Wave velocity variation is calculated using a cross-correlation algorithm to determine the time offset between the current waveform and the reference waveform, combined with the known propagation path length. The signal attenuation coefficient represents the degree of amplitude attenuation during wave propagation and is related to the material's microstructure, defects, and stress state. It is typically estimated by measuring the signal amplitude at two receiving points at different distances or by comparing the signal amplitude at different times along the same path (e.g., before and after stress changes). The specific mode amplitude ratio represents the ratio of the calculated amplitude of a specific mode to the amplitude under the reference state. These ultrasonic waveform features (wave velocity variation, attenuation coefficient, and mode amplitude ratio) are not the peak stress itself, but rather a direct reflection of the changes in the acoustic properties of the material under stress. They have a mapping relationship with the peak stress defined by physical laws. For the third detection target, the collected acceleration time-domain signal is transformed to obtain the vibration spectrum, and the dynamic alternating stress amplitude is calculated as a sudden dynamic feature for spectrum analysis and trend change monitoring. The target feature data for each detection target are corresponding to the "time-stress" sequence, ultrasonic shape characteristics, and sudden dynamic characteristics.

[0026] In S5, the dedicated models for each detection target are the fatigue damage accumulation model, the fracture mechanics and defect assessment model, and the dynamic response and event diagnosis model. These models can be machine learning models, and the historical data accumulated by the system (feature data and subsequent validation results) can be continuously used to optimize and train the dedicated models. For the first detection target, the associated data includes pipe material properties, pipe operating temperature, and pressure; for the second detection target, the associated data includes pipe defect characteristics, dimensions, and material fracture toughness; for the third detection target, the associated data includes pipe structural dynamic property parameters (natural frequency and mode shape data, mass distribution and stiffness parameters) and fluid dynamic characteristic parameters (density, sound velocity, viscosity), where the sound velocity is used to calculate the water hammer wave velocity. The specific process of inputting specific target feature data into a dedicated model corresponding to the detection target for stress assessment includes: For the first detection target, multiple "time-stress" sequences and their associated data are input into the fatigue damage accumulation model, and the current cumulative damage degree D (between 0 and 1) and the predicted remaining life are output for each detection point to assess the overall stress level and fatigue life of the pipeline. For the second detection target, the ultrasonic shape characteristics and their associated data are input into the fracture mechanics and defect assessment model, and the safety margin FAD coordinates and crack propagation rate are output to assess the stress concentration and propagation risk at specific defects or damage sites. For the third detection target, the sudden dynamic characteristics and their associated data are input into the dynamic response and event diagnosis model to output the resonance risk level (high / medium / low), transient event type, and dynamic fatigue damage contribution rate. The transient event types include Class I (severe event), Class II (major event), and Class III (general event) to assess the dynamic stress hazards caused by sudden impact loads.

[0027] For the first detection target, if the cumulative damage at all detection points is less than the safety threshold and the predicted remaining life is greater than the preset planned life cycle, the overall stress level of the pipeline is determined to meet the requirements, and it is determined whether the cumulative damage is close to the design life limit. Otherwise, the overall stress level of the pipeline is determined to not meet the requirements. For the second detection target, when the safety margin FAD coordinate is within the FAD safety zone of the judgment criterion and the safety margin is greater than the preset safety margin value, and the crack propagation rate is less than the preset propagation rate threshold, the defect is judged to be stable in the current environment and the stress level meets the requirements. Otherwise, the stress level does not meet the requirements, and the detection point with the stress level not meeting the requirements is output and marked as a risk defect area.

[0028] For the third detection target, if there is no high / medium resonance risk level, no Class I / Class II events, and the dynamic fatigue damage contribution rate is less than the preset dynamic load contribution threshold, it is determined that the stress level meets the requirements when the pipeline is subjected to a sudden impact load. Otherwise, the stress level does not meet the requirements, and the detection points where the stress level does not meet the requirements are marked as risky piping sensitive areas. Output the evaluation results obtained by the dedicated model corresponding to each detection target.

[0029] The article involves comparisons of various thresholds. Thresholds, preset values, preset ranges, etc., are set for result comparison and analysis to determine good or bad. The magnitude of these thresholds is determined by a combination of large-scale model analysis of sample data and human experience. They can also be appropriately adjusted based on seasonal or common-sense influencing factors.

[0030] In summary, this approach starts from the "multi-dimensional detection targets" (such as fatigue life assessment, defect safety monitoring, and sudden impact early warning) of pipeline safety operation, and reversely constructs a full-link technical solution of "detection target → stress type → feature data → dedicated model". Based on this, it intelligently selects and integrates multiple most suitable detection technologies. By integrating multi-physics sensing technologies such as optical (FBG), acoustic (ultrasound), and vibration (accelerometer), it can simultaneously capture static stress, local peak stress, and dynamic alternating stress for composite targets, realizing targeted, differentiated, and collaborative online monitoring of stress states for different stress types and different risk locations. The core of online stress detection is to indirectly or directly calculate the stress state by monitoring stress-related physical parameters in real time and combining them with mechanical or data models. This solution needs to meet the following requirements: continuity, non-destructiveness, anti-interference, and easy deployment. It can effectively overcome the inherent defects of single technologies (such as only magnetism or only vibration) in terms of single sensing dimension and monitoring blind spots, and improve the refinement and intelligence of pipeline safety management. This detection method, with its adaptive design and lifecycle management capabilities, fundamentally addresses the limitations of traditional methods in online monitoring under complex conditions, representing an important direction for the intelligent evolution of pipeline integrity management technology.

[0031] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto; any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. A method for online detection of pipeline stress, characterized in that: Includes the following steps: S1. Multi-target identification and decomposition: Receives and parses the pipeline engineering safety requirements of the pipeline to be tested input by the user, and decomposes them into one or more specific and operable detection targets; S2. Detection Strategy Oriented Matching: Obtain the detection target, and based on the preset expert knowledge base, determine one or more key detection parts that need to be focused on to achieve the detection target, as well as the stress type that plays a dominant role in the key detection parts, and mark the key monitoring parts as the parts to be tested. Based on the stress type, a matching detection method is selected from a variety of preset stress detection methods to generate a "test part - stress type - detection method" detection scheme for a specific detection target; S3. Deployment of testing procedures and data acquisition: Deploy physical sensors corresponding to the selected testing methods at the parts to be tested, and perform multi-channel data acquisition for multiple testing schemes to obtain pipeline stress data; S4. Data fusion preprocessing: Receive pipeline stress data for the same detection target, perform data fusion preprocessing, and obtain target feature data. S5. Target-oriented dedicated model analysis: Obtain the correlation data of each detection target, and input it together with the target feature data into the dedicated model corresponding to the detection target for stress assessment, determine whether the stress level of each detection target meets the requirements, and output the assessment results simultaneously.

2. The method for online detection of pipeline stress according to claim 1, characterized in that: In S1, the inspection targets include long-term service safety inspection of pipelines, local defect / damage inspection of pipelines, and sudden load inspection of pipelines, which are marked as inspection target one, inspection target two, and inspection target three, respectively.

3. The method for online detection of pipeline stress according to claim 2, characterized in that: In S2, for the first detection target, the part to be tested is determined to be a stress concentration area, the stress type is determined to be axial stress / bending stress, the detection method is selected as fiber optic grating sensor array, and a detection scheme of "stress concentration area - axial stress / bending stress - fiber optic grating sensor array" is generated. For the second detection target, the area to be inspected is determined to be a defect / damage area, the stress type is determined to be peak stress, the detection method is selected as ultrasonic guided wave / acoustic elastic ultrasound, and a detection scheme of "defect / damage area - peak stress - ultrasonic guided wave / acoustic elastic ultrasound" is generated; For the third detection target, the parts to be inspected are identified as sensitive parts of the piping system (pipe connection points, midpoints of long-span pipes), the stress type is identified as alternating stress / vibration stress, the detection method is selected as piezoelectric accelerometer array, and a detection scheme of "sensitive parts of piping system - alternating stress / vibration stress - piezoelectric accelerometer array" is generated.

4. The method for online detection of pipeline stress according to claim 3, characterized in that: In S3, the pipeline stress data includes: wavelength drift, ultrasonic waveform data, and vibration acceleration time-domain signal, which correspond to the pipeline stress data of each detection target.

5. The method for online detection of pipeline stress according to claim 4, characterized in that: The specific process of data fusion preprocessing in S4 includes: For the first detection target, the wavelength drift is converted into micro-strain, and real-time temperature compensation is performed using the built-in strain and temperature dual gratings. Then, based on Hooke's law and pipeline geometric parameters, the axial and bending stress values ​​are calculated, and the "time-stress" sequence of each detection point is output. For the second detection target, digital filtering and noise reduction are performed on the ultrasonic wave data to extract ultrasonic wave features: wave velocity change, signal attenuation coefficient, and amplitude ratio of specific modes; For the third detection target, the collected acceleration time-domain signal is transformed to obtain the vibration spectrum, and the dynamic alternating stress amplitude is calculated as a sudden dynamic feature. The target feature data for each detection target are the "time-stress" sequence, ultrasonic shape characteristics, and sudden dynamic characteristics.

6. The method for online detection of pipeline stress according to claim 5, characterized in that: The dedicated models for each detection target are the fatigue damage accumulation model, the fracture mechanics and defect assessment model, and the dynamic response and event diagnosis model.

7. The method for online detection of pipeline stress according to claim 6, characterized in that: The specific process of stress assessment includes: For the first detection target, the "time-stress" sequence and its associated data are input into the fatigue damage accumulation model, which outputs the current cumulative damage degree D, predicts the remaining life, and assesses the overall stress level and fatigue life of the pipeline. For the second detection target, the ultrasonic shape characteristics and their associated data are input into the fracture mechanics and defect assessment model, and the safety margin FAD coordinates and crack propagation rate are output to assess the stress concentration and propagation risk at the defect or damage site. For the third detection target, the sudden dynamic characteristics and their associated data are input into the dynamic response and event diagnosis model to output the resonance risk level (high / medium / low), transient event type, and dynamic fatigue damage contribution rate. The transient event types include Class I (severe event), Class II (major event), and Class III (general event) to assess the dynamic stress hazards caused by sudden impact loads.

8. The method for online detection of pipeline stress according to claim 7, characterized in that: For the first detection target, when the cumulative damage at all detection points is less than the safety threshold and the predicted remaining life is greater than the preset planned life cycle, the overall stress level of the pipeline is determined to meet the requirements. For the second detection target, if the safety margin FAD coordinate is within the FAD safety zone of the judgment criterion and the safety margin is greater than the preset safety margin value, and the crack propagation rate is less than the preset propagation rate threshold, the defect is judged to be stable in the current environment and the stress level meets the requirements. For the third detection target, if there is no high / medium resonance risk level, no Class I / II events, and the dynamic fatigue damage contribution rate is less than the preset dynamic load contribution threshold, it is determined that the stress level of the pipeline meets the requirements when subjected to sudden impact load.

Images