A method, system and storage medium for oil and gas pipeline safety assessment based on stress monitoring and intelligent prediction

Abstract

The present invention discloses a method, system, and storage medium for oil and gas pipeline safety assessment based on stress monitoring and intelligent prediction, which relates to the field of oil and gas pipeline safety monitoring and assessment. The method aims to address the problem in the prior art of lacking a method for oil and gas pipeline safety monitoring and assessment based on stress monitoring and intelligent prediction, making it difficult to effectively identify potential safety hazards in pipelines. The method comprises: S100, using finite element analysis to perform numerical simulations on the stress distribution of pipelines under different working conditions to determine stress-sensitive locations in the pipeline; S200, using stress sensors to collect stress data from stress-sensitive locations in the pipeline in real time; S300, constructing an XLSTM model, which uses a gating mechanism and residual connections, and uses a Hippo optimization algorithm to optimize the model's hyperparameters, and predicting stress trends based on the optimized model; S400, using an exponentially weighted moving average control chart for stress anomaly warning, introducing an adaptive threshold adjustment mechanism, performing anomaly detection on real-time stress data, and identifying potential safety risks.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas pipeline safety monitoring and assessment, and in particular to a method, system and storage medium for oil and gas pipeline safety assessment based on stress monitoring and intelligent prediction. Background Art

[0002] Oil and gas pipelines are critical infrastructure for energy transportation. Over their long service lives, they are susceptible to factors such as foundation settlement, corrosion, aging, mechanical damage, and environmental changes, potentially leading to structural deformation, material fatigue, and even failure, potentially causing major safety incidents. To ensure the safe and stable operation of oil and gas pipelines, there is an urgent need to develop high-precision, real-time pipeline stress monitoring and safety assessment methods.

[0003] Currently, traditional pipeline safety monitoring methods primarily rely on manual inspections, periodic testing, and finite element simulations. However, these methods suffer from long monitoring cycles, poor real-time performance, data lag, and an inability to accurately predict potential risks. In recent years, online monitoring systems based on sensing technologies, such as fiber grating sensors, vibrating wire sensors, and acoustic emission monitoring, have been gradually applied to pipeline safety management. However, these systems still face technical bottlenecks such as irrational sensor placement, insufficient data analysis capabilities, and a lack of efficient prediction models.

[0004] With the continuous development of deep learning technology and neural network methods, neural network-based prediction methods have gradually begun to be applied to oil and gas pipeline safety monitoring. Deep learning, as an important branch of machine learning, has also developed rapidly. Among them, convolutional neural networks and recurrent neural networks appeared earlier and are more widely used. LSTM, which is improved on the basis of recurrent neural networks, has higher prediction accuracy for time series. For example, the prior art with document number CN116447528A discloses a pipeline oil and gas leak detection method based on graph neural networks and LSTM networks. It adopts an abnormal temperature positioning method, uses a graph neural network to construct the spatial relationship between each signal node and capture the spatial characteristics of the signal, and uses an LSTM network to capture the time domain characteristics of the signal to achieve pipeline oil and gas leak detection. The prior art with document number CN111539393A discloses a third-party construction early warning method for oil and gas pipelines based on EMD decomposition and LSTM. By classifying and identifying fiber optic sensor disturbance signals, it performs perimeter security alarms. Currently, in the research on oil and gas pipeline safety monitoring and assessment, LSTM has a great advantage in prediction accuracy, but there are still problems such as difficulty in hyperparameter optimization and slow training convergence. At the same time, there is a lack of oil and gas pipeline safety monitoring and assessment methods based on stress monitoring and intelligent prediction to effectively identify potential safety hazards in pipelines, optimize early warning mechanisms, and improve the operational safety and stability of oil and gas pipelines. Summary of the Invention

[0005] The technical problems to be solved by the present invention are:

[0006] The existing technology lacks oil and gas pipeline safety monitoring and assessment methods based on stress monitoring and intelligent prediction, making it difficult to effectively identify potential safety hazards in pipelines.

[0007] The present invention is to solve the above technical problems using the following technical solutions:

[0008] The present invention provides an oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction, the method comprising the following steps:

[0009] S100: Finite element analysis is used to numerically simulate the stress distribution of pipelines under different working conditions. Adaptive meshing is used to perform local mesh refinement in high stress concentration areas to identify stress-sensitive areas of the pipeline.

[0010] S200: Install stress sensors at stress-sensitive locations to collect pipeline stress data in real time.

[0011] S300. Construct an XLSTM model. The XLSTM model uses a gating mechanism and residual connections to enable the model to have stronger long-term temporal memory capabilities. The Hippo optimization algorithm is used to optimize the hyperparameters of the XSTM model. The stress trend is predicted based on the optimized XLSTM model.

[0012] S400 uses an exponentially weighted moving average control chart for stress anomaly warning, and introduces an adaptive threshold adjustment mechanism to detect anomalies in real-time stress data and identify potential safety risks.

[0013] Furthermore, S100 includes the following steps:

[0014] S110. Based on the pipeline's geometric structure, material properties, and operating conditions, a three-dimensional finite element analysis model is constructed, and boundary conditions and loads are applied to simulate actual operating conditions.

[0015] S120 simulates the stress distribution of pipelines under normal operation, foundation settlement, and uneven loading conditions, including the key indicators of maximum principal stress and von Mises stress, and identifies stress concentration areas. Adaptive meshing is used to perform local mesh refinement in high stress concentration areas.

[0016] S130. Based on the simulation results, analyze the weak areas of the structure, assess the failure risk of the pipeline, and identify stress-sensitive areas.

[0017] Furthermore, the adaptive meshing described in S120 is used to perform local mesh encryption in the high stress concentration area, specifically: based on the adaptive meshing strategy of error estimation and gradient change, the high stress gradient area is located in combination with the spatial change rate of the stress field, and local mesh encryption optimization is performed.

[0018] Furthermore, S200 includes the following steps:

[0019] S210, installing a vibrating wire stress sensor in the stress-sensitive area;

[0020] S220, build a data acquisition system, where sensors collect stress data in real time and store it on a cloud platform using a remote data transmission protocol;

[0021] S230. The cloud platform constructs a data preprocessing module for cleaning the collected data, eliminating invalid data, amplifying and normalizing the data in real time, and extracting the mean, variance, and amplitude characteristic parameters.

[0022] Furthermore, S200 also includes: installing a temperature sensor in the stress-sensitive area to collect real-time temperature information, and dynamically correcting the stress data in combination with a temperature compensation algorithm.

[0023] Furthermore, the functional implementation process of the XLSTM model in S300 is as follows: in 、 and They are Always forget the output values ​​of the gate, input gate, and output gate, is the activation function, for The candidate unit state value at time , tanh ( ) is the hyperbolic tangent function, for The unit status value at the moment, for The output value of the hidden layer unit at time , and For model weights and biases; introduce residual connections: ,in An additional gating mechanism.

[0024] Furthermore, the Hippo optimization algorithm described in S300 is used to optimize the hyperparameters of the XSTM model, specifically:

[0025] S310, set the hyperparameter vector that XLSTM needs to optimize ; Parameter search is performed through the Hippo optimization algorithm to minimize the loss function: ,in is the true stress value, is the predicted value;

[0026] S320, input stress data , the XLSTM model outputs the predicted value: ,in It is the optimal parameter after training by the Hippo optimization algorithm; calculate the prediction error. If the error is greater than the threshold, perform iterative optimization of the Hippo optimization algorithm and adjust the hyperparameters.

[0027] Furthermore, S400 includes the following steps:

[0028] S410, use the EWMA method to smooth the pipeline stress data and calculate the weighted mean of the stress data: ,in For the current moment The EWMA calculated value of is the original stress data at the current moment, is the EWMA value of the previous moment, is the smoothing factor;

[0029] S420, introduce an adaptive threshold adjustment mechanism to dynamically adjust the EWMA control limit: , ,in UCL is the upper control limit, LCL is the lower control limit, is the mean of the historical stress data, is the standard deviation of historical stress data, is the control limit coefficient;

[0030] S430: Perform adaptive anomaly detection. ,like If the level is lower than the first-level warning threshold, the abnormality will be recorded and observed; if In the second level warning range, it is recommended to arrange regular inspections; if If the third-level warning threshold is exceeded, the emergency safety plan will be activated.

[0031] The present invention provides an oil and gas pipeline safety assessment system based on stress monitoring and intelligent prediction. The system has a program module corresponding to the steps of the method described in any of the above technical solutions, and executes the steps in the above oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction during operation.

[0032] The present invention also provides a computer-readable storage medium, which stores a computer program. The computer program is configured to implement the steps of the oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction described in any one of the above technical solutions when called by a processor.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] The present invention determines the stress-sensitive areas of the pipeline through finite element analysis and precisely arranges vibrating-wire stress sensors at the corresponding locations, achieving real-time, high-precision data acquisition and remote transmission, significantly improving the accuracy and real-time performance of pipeline safety monitoring. The XLSTM model, which introduces residual connections and a gating mechanism, is used to efficiently predict pipeline stress trends and issue anomaly warnings, giving the model stronger long-term memory capabilities and enabling more accurate predictions of future changes in pipeline stress. The Hippo optimization algorithm is used to optimize the hyperparameters of the XLSTM model, and the EWMA method is used to detect anomalies in real-time stress data. An adaptive threshold adjustment mechanism is introduced to dynamically adjust control limits based on the mean and standard deviation of historical data, effectively improving the accuracy of predictions of potential pipeline risks and reducing the risk of safety accidents caused by prediction errors.

[0035] The present invention can achieve accurate monitoring, trend prediction and safety early warning of pipeline stress, thereby improving the safety and stability of pipeline operation and reducing the risk of accidents caused by foundation settlement, corrosion, aging, mechanical damage and environmental changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Flowchart of a method for safety assessment of oil and gas pipelines based on stress monitoring and intelligent prediction in an embodiment of the present invention;

[0037] Figure 2 This is the technical route of the oil and gas pipeline safety assessment method in the embodiment of the present invention;

[0038] Figure 3 This is a flowchart of the HO-XLSTM modeling training and optimization in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the present invention, exemplary embodiments or examples of the present invention will be described below with reference to the accompanying drawings. Obviously, the described embodiments or examples are only some of the embodiments or examples of the present invention, and not all of them. Based on the embodiments or examples of the present invention, all other embodiments or examples obtained by those skilled in the art without creative work should fall within the scope of protection of the present invention.

[0040] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0041] Specific implementation plan 1: Combined Figures 1 to 3As shown, the present invention provides an oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction, the method comprising the following steps:

[0042] S100: Finite element analysis is used to numerically simulate the stress distribution of pipelines under different working conditions. Adaptive meshing is used to perform local mesh refinement in high stress concentration areas to identify stress-sensitive areas of the pipeline.

[0043] S200: Install stress sensors at stress-sensitive locations to collect pipeline stress data in real time.

[0044] S300. Construct an XLSTM model. The XLSTM model uses a gating mechanism and residual connections to enable the model to have stronger long-term temporal memory capabilities. The Hippo optimization algorithm is used to optimize the hyperparameters of the XSTM model. Stress trend prediction is performed based on the XLSTM model.

[0045] S400 uses an exponentially weighted moving average control chart for stress anomaly warning, and introduces an adaptive threshold adjustment mechanism to detect anomalies in real-time stress data and identify potential safety risks.

[0046] Specific implementation scheme 2: S100 includes the following steps:

[0047] S110. Based on the pipeline's geometric structure, material properties, and operating conditions, a three-dimensional finite element analysis model was constructed using the finite element analysis software ABAQUS. The elastic-plastic parameters, temperature-related properties, and fatigue properties of the material were defined. Based on the pipeline's on-site operating conditions and actual operating conditions, boundary conditions and constraints were applied, along with actual load conditions such as internal and external pressure, thermal load, gravity load, and bending moment, to simulate actual operating conditions.

[0048] S120. Select the appropriate analysis step within the ABAQUS solution module, such as static analysis, dynamic analysis, or nonlinear analysis mode, determine the corresponding calculation parameters, including contact and friction conditions, and simulate the stress, strain, and displacement distribution of the pipeline under normal operation (internal fluid pressure fluctuations), foundation settlement, and uneven loading conditions. This includes key indicators such as maximum principal stress and von Mises stress, and generates detailed contour maps, vector diagrams, and graphical visualizations to intuitively analyze and identify areas of stress concentration. Furthermore, after identifying areas of stress concentration, adaptive meshing is used to refine the mesh locally in areas of high stress concentration to obtain more accurate stress distribution data.

[0049] S130. Identify stress-sensitive areas. Based on the simulation results, combined with the path analysis and cross-section analysis tools in the ABAQUS post-processing module, conduct focused analysis and assessment of weak areas in the pipeline. This will clarify the location and risk level of these structural weaknesses, and further provide a targeted and highly reliable optimization plan for the placement of field sensors, thereby ensuring the accuracy and effectiveness of monitoring and assessment. This implementation plan is otherwise identical to Specific Implementation Plan 1.

[0050] Specific Implementation Plan 3: Adaptive meshing, as described in S120, is used to refine the mesh locally in areas of high stress concentration. Specifically, an adaptive meshing strategy based on error estimation and gradient variation is employed, combining the spatial rate of change of the stress field to locate areas of high stress gradients and optimize local mesh refinement. Smaller element sizes are used in stress concentration areas to improve the spatial resolution of the stress solution, enhance the accuracy of local stress calculations, and ensure reliable results in stress-sensitive areas. This implementation plan is otherwise identical to Specific Implementation Plan 2.

[0051] Specific implementation scheme 4: S200 includes the following steps:

[0052] S210. Install a vibrating-wire stress sensor in the stress-sensitive area, using mechanical fixation or welding to ensure good contact between the sensor and the pipe surface and reduce measurement errors. To reduce installation errors, the present invention uses high-precision laser alignment technology, and uses a laser positioning system to guide and adjust the installation angle and position in real time to ensure that the sensor axis is strictly aligned with the normal of the pipe outer wall.

[0053] S220. Build a data acquisition system. Sensors are used to collect stress data in real time and store it in a cloud platform using a remote data transmission protocol.

[0054] S230: The cloud platform constructs a data preprocessing module to clean the collected data in real time, remove invalid data, amplify and normalize it, extract the mean, variance, and amplitude characteristic parameters, and store the valid characteristic parameters, thereby providing timely, reliable, and high-quality data foundation support for subsequent pipeline status monitoring, trend analysis, and safety assessment. This implementation plan is otherwise identical to Specific Implementation Plan 3.

[0055] Specific implementation scheme 5: S200 further includes: installing a temperature sensor in the stress-sensitive area to collect real-time temperature information, and dynamically correcting the stress data in combination with a temperature compensation algorithm. The rest of this implementation scheme is the same as specific implementation scheme 4.

[0056] During the sensor installation phase, this implementation plan should carry out pre-treatment based on the surface condition and installation location of the pipeline. First, the pipeline surface should be thoroughly cleaned and polished to ensure a tight fit between the vibrating string sensor and the pipeline. The sensor should be firmly fixed in the selected position by mechanical clamps or precision welding to ensure good contact between the sensor and the pipeline surface and reduce measurement errors. The measuring axis of the sensor must be consistent with the main stress direction to ensure data measurement accuracy. In terms of setting up the data acquisition system, a high-precision, high-resolution data acquisition module should be selected to accurately record the signal data output by the sensor in real time. At the same time, a remote data transmission unit should be configured, using standard and mature 5G, LoRa, and MQTT data communication protocols to achieve secure, stable and efficient data transmission, and synchronize it to the cloud storage system in real time. In addition, a complete data reception and unified management interface is built on the cloud platform to achieve efficient access and centralized management of real-time data, ensuring the reliability and stability of data transmission.

[0057] Specific implementation plan six: S230 also includes data optimization processing, specifically: based on the stress data of the collected pipeline Establish a time series data set; clean the data, remove invalid data, amplify and normalize it. The normalization formula is: , so that the data range is normalized to [0,1]. The rest of this embodiment is the same as the specific embodiment five.

[0058] Specific implementation plan seven: The functional implementation process of the XLSTM model described in S300 is as follows: in 、 and They are Always forget the output values ​​of the gate, input gate, and output gate, is the activation function, for The candidate unit state value at time , tanh ( ) is the hyperbolic tangent function, for The unit status value at the moment, for The output value of the hidden layer unit at time , and For model weights and biases; introduce residual connections: ,in This embodiment is otherwise the same as the sixth embodiment.

[0059] Specific implementation plan eight: The Hippo optimization algorithm is used to optimize the hyperparameters of the XSTM model as described in S300, specifically:

[0060] S310, set the hyperparameter vector that XLSTM needs to optimize ; Parameter search is performed through the Hippo optimization algorithm to minimize the loss function: ,in is the true stress value, is the predicted value;

[0061] S320, input stress data , the XLSTM model outputs the predicted value: ,in It is the optimal parameter after training by the Hippo optimization algorithm; calculate the prediction error. If the error is greater than the threshold, perform iterative optimization of the Hippo optimization algorithm and adjust the hyperparameters.

[0062] like Figure 3 As shown, the preprocessed data set is reasonably divided into a training set and a test set, and the parameters of the extended long short-term memory network model XLSTM are initialized. Subsequently, the initial values ​​of the relevant parameters of the Hippo swarm optimization algorithm are set, and a fitness function with prediction error as the core is defined. Based on this, the key hyperparameters of XLSTM, such as the number of hidden layers, the number of hidden units, and the learning rate, are efficiently searched and optimized. During the optimization process, the position and speed of the Hippo are continuously iterated and updated, and the optimal parameter combination is gradually approached until the maximum number of iterations is reached or the pre-set convergence conditions are met, and the optimal XLSTM model parameters are output. Finally, the optimized XLSTM model is applied to the actual stress prediction task, and the prediction error MES is calculated to evaluate the prediction accuracy; if the prediction error exceeds the set threshold, the HO algorithm is re-called, the model hyperparameters are adjusted, and the prediction performance of the model is continuously improved to meet the high-precision requirements of actual engineering applications. The rest of this implementation plan is the same as Specific Implementation Plan Seven.

[0063] Specific implementation scheme nine: S400 includes the following steps:

[0064] S410, use the EWMA method to smooth the pipeline stress data and calculate the weighted mean of the stress data: ,in For the current moment The EWMA calculated value of is the original stress data at the current moment, is the EWMA value of the previous moment, is the smoothing factor;

[0065] S420, introduce an adaptive threshold adjustment mechanism to dynamically adjust the EWMA control limit: , ,in UCL is the upper control limit, LCL is the lower control limit, is the mean of historical stress data, is the standard deviation of historical stress data, is the control limit coefficient;

[0066] S430: Perform adaptive anomaly detection. ,like If the level is lower than the first-level warning threshold, the abnormality will be recorded and observed; if In the second level warning range, it is recommended to arrange regular inspections; if If the level 3 warning threshold is exceeded, the emergency safety plan will be activated. The rest of this implementation plan is the same as the specific implementation plan eight.

[0067] The EWMA control chart in this implementation is more effective at detecting smaller stress changes, making it suitable for anomaly monitoring of long-term oil and gas pipeline operations. An adaptive threshold adjustment mechanism dynamically adjusts the control limits (UCL and LCL) based on the mean and standard deviation of historical data, thereby improving anomaly detection sensitivity and reducing false alarm rates.

[0068] This implementation plan determines the appropriate smoothing factor in the EWMA method by analyzing historical pipeline stress monitoring data , balancing the sensitivity to data fluctuations and the false alarm rate. During implementation, computational tools or custom programs are used to calculate the EWMA value at each moment in real time, dynamically generating corresponding control limits (UCL and LCL) to monitor data stability in real time. Furthermore, a hierarchical anomaly detection system is established, setting first-, second-, and third-level warning thresholds based on actual project requirements. Automatic alarm mechanisms and response processes are established for different levels of anomaly thresholds. When anomaly indicators exceed the set thresholds, on-site inspections, maintenance checks, or emergency safety plans are automatically triggered, ensuring timely detection of potential problems and rapid response, thereby improving the safety and stability of pipeline operations.

[0069] The oil and gas pipeline safety assessment method (algorithm) based on stress monitoring and intelligent prediction proposed in the present invention is the underlying technical core of the present invention. Various products can be derived based on the algorithm.

[0070] Based on the method proposed in the present invention, a stress monitoring and intelligent prediction-based oil and gas pipeline safety assessment system is developed using a programming language. The system has program modules corresponding to the steps of the above-mentioned technical solution, and executes the steps of the above-mentioned stress monitoring and intelligent prediction-based oil and gas pipeline safety assessment method during operation.

[0071] The developed system (software) computer program is stored on a computer-readable storage medium. When invoked by a processor, the computer program is configured to implement the steps of the aforementioned oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction. This materializes the present invention on a carrier, becoming a computer program product.

[0072] Various implementations of the systems and techniques described herein can be realized in digital electronic circuitry, integrated circuitry, dedicated ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system comprising at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0073] The computer programs (also referred to as programs, software, software applications, or code) herein comprise machine instructions for a programmable processor and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or apparatus (e.g., a magnetic disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including machine-readable media that receive machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0074] Although the present invention is disclosed as above, the scope of protection disclosed by the present invention is not limited thereto. Those skilled in the art of the present invention may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for oil and gas pipeline safety assessment based on stress monitoring and intelligent prediction, characterized in that: The method comprises the following steps: S100: Finite element analysis is used to numerically simulate the stress distribution of pipelines under different working conditions. Adaptive meshing is used to perform local mesh refinement in high stress concentration areas to identify stress-sensitive areas of the pipeline. S200: Install stress sensors at stress-sensitive locations to collect pipeline stress data in real time. Specifically: S210, installing a vibrating wire stress sensor in the stress-sensitive area; S220, build a data acquisition system, where sensors collect stress data in real time and store it on a cloud platform using a remote data transmission protocol; S230: The cloud platform constructs a data preprocessing module for cleaning the collected data, eliminating invalid data, amplifying and normalizing the data in real time, and extracting the mean, variance, and amplitude characteristic parameters; S300. Construct an XLSTM model. The XLSTM model uses a gating mechanism and residual connections to enable the model to have stronger long-term temporal memory capabilities. The Hippo optimization algorithm is used to optimize the hyperparameters of the XSTM model. The stress trend is predicted based on the optimized XLSTM model. S400 uses an exponentially weighted moving average control chart for stress anomaly warning and introduces an adaptive threshold adjustment mechanism to detect anomalies in real-time stress data and identify potential safety risks; S100 specifically includes the following steps: S110. Based on the pipeline's geometric structure, material properties, and operating conditions, a three-dimensional finite element analysis model is constructed, and boundary conditions and loads are applied to simulate actual operating conditions. S120 simulates the stress distribution of pipelines under normal operation, foundation settlement, and uneven loading conditions, including the key indicators of maximum principal stress and von Mises stress, and identifies stress concentration areas. Adaptive meshing is used to perform local mesh refinement in high stress concentration areas. S130. Based on the simulation results, analyze the structural weak areas, assess the failure risk of the pipeline, and identify the stress-sensitive areas; Adaptive meshing is used in S120 to perform local mesh refinement in high stress concentration areas. Specifically, the adaptive meshing strategy based on error estimation and gradient change is used to locate high stress gradient areas in combination with the spatial change rate of the stress field, and perform local mesh refinement optimization. As described in S300, the Hippo optimization algorithm is used to optimize the hyperparameters of the XSTM model, specifically: S310, set the hyperparameter vector that XLSTM needs to optimize ; Parameter search is performed through the Hippo optimization algorithm to minimize the loss function: ,in is the true stress value, is the predicted value; S320, input stress data , the XLSTM model outputs the predicted value: ,in is the optimal parameter after training with the Hippo optimization algorithm; calculate the prediction error. If the error is greater than the threshold, perform iterative optimization with the Hippo optimization algorithm to adjust the hyperparameters; S400 includes the following steps: S410, use the EWMA method to smooth the pipeline stress data and calculate the weighted mean of the stress data: ,in For the current moment The EWMA calculated value of is the original stress data at the current moment, is the EWMA value of the previous moment, is the smoothing factor; S420, introduce an adaptive threshold adjustment mechanism to dynamically adjust the EWMA control limit: , , where UCL is the upper control limit and LCL is the lower control limit, is the mean of historical stress data, is the standard deviation of historical stress data, is the control limit coefficient; S430: Perform adaptive anomaly detection. ,like If the level is lower than the first-level warning threshold, the abnormality will be recorded and observed; if In the second level warning range, it is recommended to arrange regular inspections; if If the third-level warning threshold is exceeded, the emergency safety plan will be activated.

2. The oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction according to claim 1 is characterized in that: S200 also includes: installing a temperature sensor in the stress-sensitive area to collect real-time temperature information, and dynamically correcting the stress data in combination with a temperature compensation algorithm.

3. The oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction according to claim 2 is characterized in that: The functional implementation process of the XLSTM model described in S300 is as follows: , , , , , ,in 、 and are the output values ​​of the forget gate, input gate, and output gate at time t, respectively. ( ) is the activation function, is the state value of the candidate unit at time t, tanh( ) is the hyperbolic tangent function, is the unit state value at time t, is the output value of the hidden layer unit at time t, and For model weights and biases; introduce residual connections: ,in An additional gating mechanism.

4. An oil and gas pipeline safety assessment system based on stress monitoring and intelligent prediction, characterized in that: The system has a program module corresponding to the steps of the method described in any one of claims 1 to 3 above, and executes the steps of the above-mentioned oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction when running.

5. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and the computer program is configured to implement the steps of the oil and gas pipeline safety assessment method based on stress monitoring and intelligent prediction according to any one of claims 1 to 3 when called by a processor.

Images