High-temperature steam pipeline stress and strain state monitoring system based on characteristic safety parameters

Abstract

The present invention discloses a high-temperature steam pipeline stress and strain state monitoring system based on characteristic safety parameters, which includes a sensor acquisition module, a data processing module, a wireless communication module, an energy acquisition module, an edge computing module, a fault self-diagnosis module and a server. The sensor acquisition module acquires first sensing information through a heterogeneous sensor combination distributed on the pipeline surface, and the data processing module extracts characteristic parameters such as stress concentration coefficient, strain rate, temperature gradient and vibration frequency. The wireless communication module uses a multi-hop network mode to transmit data to the edge computing module for secondary processing to obtain pipeline operation information. The energy acquisition module obtains energy from the pipeline surface to power the sensors. The fault self-diagnosis module realizes abnormal device location and isolation through cross-validation and link error rate analysis. The server comprehensively receives various types of information to complete continuous monitoring of the pipeline state, forming a complete self-powered intelligent monitoring system.

Description

Technical field

[0001] The present invention relates to the field of stress monitoring, and particularly to a monitoring system for the stress and strain state of high-temperature steam pipelines based on characteristic safety parameters. Background art

[0002] As a core component of the energy transportation system, the real-time monitoring of the stress and strain state of high-temperature steam pipelines is of great significance for preventing structural failures. Traditional monitoring methods mostly use single-type sensors to collect local parameters, which are difficult to comprehensively characterize the multi-dimensional stress distribution characteristics of pipelines, and there are monitoring blind spots caused by insufficient complementarity of sensors. Existing wired transmission schemes rely on fixed wiring, which is prone to line aging and fracture under harsh working conditions such as high temperature and vibration, resulting in data interruption; at the same time, the external power supply mode is limited by the pipeline environment and it is difficult to meet the energy requirements for long-term monitoring. Most systems lack the ability of autonomous fault diagnosis, and are prone to overall monitoring failure when sensors are abnormal or communication links fail. In addition, traditional data processing methods have problems such as large transmission load and high response delay, making it difficult to realize the real-time evaluation of the pipeline health state. Moreover, the existing technologies mostly adopt the strategy of global shutdown and maintenance for abnormal working conditions, lacking a local fault tolerance mechanism, which seriously affects the monitoring continuity. Summary of the invention

[0003] In view of the above problems, the present invention provides a monitoring system that meets the intelligent monitoring requirements of high-temperature steam pipelines in terms of monitoring accuracy, environmental adaptability and operation and maintenance efficiency.

[0004] To achieve the above object, the present application provides a high-temperature steam pipeline stress and strain state monitoring system based on characteristic safety parameters, including a sensor acquisition module, a data processing module, a wireless communication module, an energy acquisition module, an edge computing module, a fault self-diagnosis module, and a server. The sensor acquisition module includes a plurality of heterogeneous sensors, which are distributed on the pipeline in a first preset manner. The heterogeneous sensors are configured as at least two of a strain gauge, an optical fiber sensor, and a vibration sensor. The sensor acquisition module is used to acquire first sensing information. The data processing module is used to receive and process the first sensing information and extract characteristic safety parameters, which include a stress concentration coefficient, a strain rate, a temperature gradient, and a vibration frequency. The wireless communication module is configured to use a multi-hop network communication mode to transmit the first sensing information and / or the characteristic safety parameters to the edge computing module. The energy acquisition module is arranged on the pipeline surface. The energy acquisition module is electrically connected to the sensor acquisition module. The energy acquisition module is used to collect energy from the pipeline surface and convert it into electrical energy to be delivered to the heterogeneous sensors. The edge computing module is used to process the characteristic safety parameters to obtain first processing information and transmit the first processing information to the server. The first processing information is the operation information of the pipeline. The fault self-diagnosis module is used to detect an abnormal state, and the abnormal state is configured as any one of acquisition abnormality, communication abnormality, and power supply abnormality. The fault self-diagnosis module is used to achieve fault location and isolation of abnormal devices through cross-verification and link error rate analysis. The server is used to receive the first sensing information and / or the first processing information and / or the characteristic safety parameters and continuously monitor the working state of the pipeline.

[0005] In some embodiments, the fault self-diagnosis module is used to achieve fault location and isolation of abnormal devices through cross-verification and link error rate analysis, including:

[0006] Cross-verify a plurality of first sensing information to obtain several first stress calculation values, synchronously obtain the first stress actual value associated with the first sensing information corresponding to each first stress calculation value, and determine whether the deviation value between the first stress calculation value and the first stress actual value is within the range of a first preset threshold. If not, it means that the heterogeneous sensor corresponding to the first stress actual value is in acquisition abnormality;

[0007] And, monitor the communication link state of the wireless communication module and calculate the link error rate, and determine whether the link error rate is within the range of a second preset threshold. If not, it means that the wireless communication module is in communication abnormality;

[0008] And, monitor the power access value of each heterogeneous sensor and obtain the calibrated power value of the heterogeneous sensor;

[0009] When the power access value of the heterogeneous sensor is lower than the calibrated power value of the heterogeneous sensor, calculate the deviation between the power access value and the calibrated power value, denoted as the first deviation value;

[0010] Determine whether the first deviation value is within the range of the third preset threshold. If not, it indicates that the heterogeneous sensor corresponding to the power access value is in abnormal power supply;

[0011] If the abnormal device is a heterogeneous sensor, isolate the heterogeneous sensor from the fault and enable the backup acquisition mode associated with the heterogeneous sensor;

[0012] If the abnormal device is a wireless communication module, switch to the backup communication link, or adjust the backup communication mode of the wireless communication module.

[0013] In some embodiments, the backup acquisition mode is configured as:

[0014] Obtain the location information of the heterogeneous sensor in the abnormal state, denoted as the abnormal location information, and denote the heterogeneous sensor in the abnormal state as the abnormal sensor;

[0015] With the abnormal location information as the center and a preset length as the radius, divide a backup acquisition area. Denote the heterogeneous sensors other than the abnormal sensor in the backup acquisition area as backup sensors;

[0016] And obtain the established acquisition frequency of the abnormal sensor, denoted as the first acquisition frequency;

[0017] Generate a second acquisition frequency according to the first acquisition frequency;

[0018] Control the backup sensors to collect backup sensing information according to the second acquisition frequency;

[0019] Perform data fusion on multiple backup sensing information to obtain a second calculated stress value;

[0020] Take the second calculated stress value as the stress value corresponding to the abnormal location information.

[0021] In some embodiments, if the abnormal device is a wireless communication module, switching to the backup communication link, or adjusting the backup communication mode of the wireless communication module includes:

[0022] The backup communication link is a pre-configured redundant communication path;

[0023] If the backup communication link is unavailable, enable the backup communication mode of the wireless communication module. The backup communication mode includes reducing the communication rate and switching the communication protocol;

[0024] And record the communication anomaly event and transmit the communication anomaly event to the edge computing module and the server.

[0025] In some embodiments, the wireless communication module includes a plurality of sensor nodes, which are distributed on the pipeline in a second preset manner to form a multi-hop network. Each sensor node is configured to have data receiving and relay forwarding functions. Each sensor node establishes a communication connection with at least one neighbor node, and the neighbor node is configured as other sensor nodes adjacent to the sensor node.

[0026] In some embodiments, the multi-hop network communication mode is configured to include the following steps:

[0027] When a certain sensor node needs to send data, the nearest neighbor node is selected as the next-hop node according to the preset routing protocol;

[0028] After the neighbor node receives the data, it forwards the data to the next relay node according to the routing protocol until the data reaches the edge computing module. The relay node is the neighbor node adjacent to the neighbor node;

[0029] The edge computing module receives data from multiple sensor nodes and performs summarization and processing;

[0030] The communication status of each sensor node in the multi-hop network is monitored in real time. If a fault is detected, the routing protocol automatically adjusts the network topology and selects an alternative path for data transmission.

[0031] In some embodiments, the energy harvesting module includes a vibration energy harvester, a wind speed energy harvester, a power reserve unit, and a power management circuit. The energy harvesting module is used to harvest energy from the pipeline surface and convert it into electrical energy for delivery to the heterogeneous sensors, including:

[0032] Vibration energy harvesters and wind speed energy harvesters are arranged at key positions of the pipeline. The key positions include pipeline elbow areas, weld areas, gas flow ports, and high stress concentration areas. The vibration energy harvester is used to convert the mechanical vibration energy on the pipeline surface into electrical energy, and the wind speed energy harvester is used to convert the air flow velocity around the pipeline into electrical energy;

[0033] The electrical energy generated by the vibration energy harvester and the wind speed energy harvester is stored in the power reserve unit;

[0034] The power management circuit monitors the reserved electrical energy value of the power reserve unit and the electrical energy access value of the heterogeneous sensors in real time;

[0035] When the electrical energy access value of the heterogeneous sensor is lower than the calibrated electrical energy value of the heterogeneous sensor, the deviation between the electrical energy access value and the calibrated electrical energy value is calculated and denoted as the second deviation value;

[0036] It is judged whether the reserved electrical energy value is greater than the second deviation value. If so, the power reserve unit and the heterogeneous sensor are connected;

[0037] Otherwise, it is not connected.

[0038] In some embodiments, the data processing module is configured to receive and process the first sensing information and extract feature security parameters, including:

[0039] Generate a data processing task according to the first sensing information, decompose the data processing task into multiple processing subtasks, and allocate the multiple processing subtasks to a multi-core CPU for parallel computing;

[0040] The multi-core CPU parallel computing includes the following steps:

[0041] Adopt wavelet transform or Fourier transform to compress and denoise the first sensing information to reduce the data volume of the first sensing information;

[0042] Use the Kalman filtering algorithm to eliminate the sensor noise and environmental interference noise in the first sensing information to obtain the second sensing information;

[0043] Adopt a data fusion algorithm to fuse the second sensing information to obtain the feature security parameters;

[0044] Send the feature security parameters to the edge computing module.

[0045] In some embodiments, the edge computing module is configured to process the feature security parameters to obtain the first processing information, including:

[0046] Adopt an incremental calculation method to process the feature security parameters. The data processing includes the following steps:

[0047] Perform Huffman coding on the feature security parameters, and record the processed feature security parameters as the first packed data;

[0048] And construct an operating state feature vector of the pipeline according to the feature security parameters;

[0049] Perform state evaluation on the operating state feature vector to obtain an evaluation result. The evaluation result includes a warning state, a maintenance state, an overhaul state, and a normal state;

[0050] Perform Huffman coding on the evaluation result, and record the processed evaluation result as the second packed data;

[0051] Integrate the second packed data with the first packed data to form the first processing information, and transmit the first processing information to the server;

[0052] And store the feature security parameters and the evaluation result in the local database.

[0053] In some embodiments, performing state evaluation on the operating state feature vector to obtain an evaluation result includes:

[0054] Input the operating status feature vector into the multi-level status classification model. The operating status feature vector includes the axial stress gradient of the pipeline, the circumferential strain volatility, the temperature-stress coupling coefficient, the entropy value of steam pressure change, and the vibration spectrum distortion degree.

[0055] The multi-level status classification model performs the following hierarchical judgments:

[0056] The first-level judgment: When the temperature-stress coupling coefficient exceeds the preset dynamic threshold and lasts for a duration exceeding the preset duration threshold range, a warning state is triggered.

[0057] The second-level judgment: If the product of the axial stress gradient and the circumferential strain volatility exceeds the material yield critical value, a maintenance state is triggered.

[0058] The third-level judgment: When the vibration spectrum distortion degree and the vibration spectrum distortion degree under the preset fault mode are within the range of the same cosine similarity threshold, an overhaul state is triggered.

[0059] The fourth-level judgment: If the operating status feature vector is within the range of the preset operating threshold, a normal state is triggered.

[0060] Establish a decision fusion mechanism based on fuzzy logic to perform weighted fusion on the judgment results output by the multi-level status classification model.

[0061] Adopt a sliding time window algorithm to perform trend analysis on the evaluation results of a preset number of consecutive sampling periods. When the same operating status appears continuously for a preset number of times, the evaluation result output is triggered.

[0062] And, match the evaluation result with the historical fault cases in the local database to correct the credibility level of the evaluation result.

[0063] Different from the prior art, the above technical solution has the following beneficial effects:

[0064] The present invention provides a high-temperature steam pipeline stress and strain state monitoring system based on characteristic safety parameters, including a sensor acquisition module, a data processing module, a wireless communication module, an energy acquisition module, an edge computing module, a fault self-diagnosis module, and a server. The sensor acquisition module acquires the first sensing information through a heterogeneous sensor combination distributed on the pipeline surface, and extracts stress concentration coefficient, strain rate, temperature gradient, and vibration frequency characteristic parameters through the data processing module. The wireless communication module uses a multi-hop network mode to transmit data to the edge computing module for secondary processing to obtain pipeline operation information. The energy acquisition module obtains energy from the pipeline surface to power the sensors. The fault self-diagnosis module realizes abnormal device location and isolation through cross-validation and link error rate analysis. The server comprehensively receives various types of information to complete continuous monitoring of the pipeline state, forming a complete self-powered intelligent monitoring system.

[0065] The above description of the invention content is only an overview of the technical solution of this application. In order to enable those of ordinary skill in the art to more clearly understand the technical solution of this application, and then implement it according to the content recorded in the description and the drawings, and in order to make the above objects, other objects, features and advantages of this application more easily understood, the following will be described in conjunction with the specific embodiments of this application and the drawings. Brief Description of the Drawings

[0066] The drawings are only used to illustrate the principles, implementation methods, applications, features, effects, etc. of the specific embodiments of the present invention and other related contents, and should not be considered as a limitation to this application.

[0067] In the drawings of the specification:

[0068] Figure 1 It is a schematic diagram of the steps from step S101 to step S106 of the standby acquisition mode described in the specific embodiment;

[0069] Figure 2 It is a schematic diagram of the steps from step S201 to step S204 of the multi-hop network communication mode described in the specific embodiment;

[0070] Figure 3 It is a schematic diagram of the steps from step S301 to step S307 of the energy acquisition module described in the specific embodiment. Detailed Description of the Invention

[0071] In order to elaborate in detail on the possible application scenarios, technical principles, specific implementable solutions, achievable purposes and effects, etc. of this application, the following will be described in detail in conjunction with the listed specific embodiments and the drawings. The embodiments described herein are only used to more clearly illustrate the technical solutions of this application, so they are only examples and cannot be used to limit the protection scope of this application.

[0072] Referring to "embodiments" herein means that the specific features, structures or characteristics described in connection with the embodiments can be included in at least one embodiment of this application. The term "embodiment" appearing in various positions in the specification does not necessarily refer to the same embodiment, nor does it particularly limit its independence or relevance to other embodiments. In principle, in this application, as long as there is no technical contradiction or conflict, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0073] Unless otherwise defined, the meanings of the technical terms used herein are the same as those commonly understood by those skilled in the technical field to which this application belongs; the use of the relevant terms herein is only for describing specific embodiments and is not intended to limit this application.

[0074] In the description of the present application, the term "and / or" is an expression used to describe the logical relationship between objects, indicating that there can be three relationships. For example, A and / or B means: there is A, there is B, and there is both A and B at the same time. In addition, the character " / " in this text generally represents an "or" logical relationship between the associated objects before and after.

[0075] In the present application, terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual quantitative, primary-secondary, or sequential relationships between these entities or operations.

[0076] Without further limitations, in the present application, the open-ended expressions such as "comprising", "including", "having", or other similar expressions used in a statement are intended to cover non-exclusive inclusion. These expressions do not exclude the possibility that there may be additional elements in the process, method, or product including the said elements. Thus, in a process, method, or product including a series of elements, it may include not only those defined elements, but also other elements not explicitly listed, or elements inherent to such a process, method, or product.

[0077] Similar to the understanding in the "Examination Guidelines", in the present application, expressions such as "greater than", "less than", "exceeding", etc. are understood not to include the recited number; expressions such as "above", "below", "within", etc. are understood to include the recited number. In addition, in the description of the embodiments of the present application, the meaning of "a plurality of" is two or more (including two). Similar expressions related to "many", such as "multiple groups", "multiple times", etc., are understood in the same way, unless otherwise specifically defined.

[0078] In the description of the embodiments of the present application, the spatially related expressions used, such as "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "perpendicular", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the specific embodiment or the drawing. It is only for the convenience of describing the specific embodiments of the present application or facilitating the understanding of the reader, rather than indicating or implying that the device or component referred to must have a specific position, a specific orientation, or be constructed or operated in a specific orientation. Therefore, it should not be construed as a limitation to the embodiments of the present application.

[0079] The processor described in the embodiments of the present application may be implemented by hardware, firmware, software, or a combination thereof, and may use circuits, one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), central processing units (CPUs), controllers, microcontrollers, microprocessors, or at least one of other physical, biological, or chemical structures capable of implementing functions similar to or equivalent to those of the above-listed processors, such as biological neurons, quantum computing units, DNA computing units, etc., so that the processor can execute some steps, all steps, or any combination of the steps mentioned in the computer programs or methods involved in the various embodiments of the present application.

[0080] The computer programs involved in the embodiments can be stored in a computer device-readable storage medium, which includes but is not limited to magnetic disks, magnetic tapes, magnetic cards, floppy disks, flash memories, optical discs, optical cards, read-only memories (ROMs), random access memories (RAMs), erasable programmable ROMs (EPROMs), and electrically erasable programmable ROMs (EEPROMs), etc. It also includes other biological, physical, or chemical structures that can achieve functions similar to or equivalent to the above-listed storage media, such as units with information storage capabilities like DNA, RNA, proteins, etc. In a specific embodiment, the storage medium involved can be one of the above medium types or a combination of the above medium types. In different embodiments, the computer programs involved in the embodiments can be stored centrally in a single medium or distributed among multiple media. The memory containing the computer device-readable storage medium can be a non-volatile memory or a random access memory. These computer device-readable storage media can be built into the device or connected to the device involved in the embodiments as an external device or a part of an external device. In some embodiments, the memory with the computer device-readable storage medium is deployed locally; in other embodiments, a scheme of deploying the memory away from the processor can also be adopted, such as a network-attached memory accessed via an RF circuit or an external port and a communication network, where the communication network can be the Internet, one or more intranets, local area networks (LANs), wide area wireless networks (WLANs), storage area networks (SANs), etc., or a suitable combination thereof, as long as the computer device can access the memory. In addition, the computer programs involved in the embodiments can be stored in plaintext / ciphertext form or designed as training data and integrated and recombinantly hidden in the parameter states of a deep neural network or other machine learning models through model training.

[0081] Please refer to Figures 1 to 3, To achieve the above object, this embodiment provides a high-temperature steam pipeline stress and strain state monitoring system based on characteristic safety parameters, including a sensor acquisition module, a data processing module, a wireless communication module, an energy acquisition module, an edge computing module, a fault self-diagnosis module, and a server. The sensor acquisition module includes multiple heterogeneous sensors, which are distributed on the pipeline in a first preset manner. The heterogeneous sensors are configured as at least two of strain gauges, fiber optic sensors, and vibration sensors. The sensor acquisition module is used to acquire first sensing information. The data processing module is used to receive and process the first sensing information and extract characteristic safety parameters, which include stress concentration coefficient, strain rate, temperature gradient, and vibration frequency. The wireless communication module is configured to use a multi-hop network communication mode to transmit the first sensing information and / or characteristic safety parameters to the edge computing module. The energy acquisition module is arranged on the pipeline surface. The energy acquisition module is electrically connected to the sensor acquisition module. The energy acquisition module is used to collect energy from the pipeline surface and convert it into electrical energy to be delivered to the heterogeneous sensors. The edge computing module is used to process the characteristic safety parameters to obtain first processed information and transmit the first processed information to the server. The first processed information is the operation information of the pipeline. The fault self-diagnosis module is used to detect abnormal states, which are configured as any one of acquisition abnormality, communication abnormality, and power supply abnormality. The fault self-diagnosis module is used to achieve fault location and isolation of abnormal devices through cross-verification and link error rate analysis. The server is used to receive the first sensing information and / or the first processed information and / or the characteristic safety parameters and continuously monitor the working state of the pipeline.

[0082] In this embodiment, the heterogeneous sensors of the sensor acquisition module are understood as sensors with various different sensing principles, including optical sensors, infrared sensors, acoustic wave sensors, electromagnetic sensors, mechanical sensors, etc. The sensors are encapsulated with high-temperature resistant ceramics and have an anti-corrosion alloy structure to withstand the high temperature on the pipe surface and the steam corrosion environment. The first preset method refers to arranging sensor nodes in a combined layout of circular array and axial gradient at elbows, welds and support parts according to the distribution characteristics of the pipe stress concentration areas, so as to ensure the monitoring coverage of key areas. Heterogeneous sensors are set on the equipment surface, and cross-verification is carried out by using different types of sensors to avoid the limitations of single-type sensors. The data processing module is used to receive and process the first sensing information, eliminate environmental interference, and accurately extract characteristic parameters such as stress concentration coefficients. The wireless communication module uses wireless communication technology for data transmission, reducing the difficulty and cost of wiring. The multi-hop network communication mode means that each sensor is regarded as a node, and based on the dynamic routing protocol, it autonomously selects adjacent nodes as relays to form a self-organizing mesh topology structure. When a single node fails, it automatically switches to the standby path to maintain the communication link. The multi-hop communication technology is adopted to ensure that data can be transmitted to the central node through multiple paths, avoiding single-point communication failures. The specific content will be described in detail later. The energy acquisition module can convert other types of energy in the working environment of the pipe into electric energy for temporary storage and use it as a backup power source to ensure the normal power supply of the heterogeneous sensors. It is ensured that when the main power supply fails, the backup power supply can take over temporarily and send a power failure information to the server at the same time, avoiding the missed acquisition of information by the heterogeneous sensors due to abnormal power supply and reducing the accident rate. Preliminary data processing and analysis are carried out on the edge device through the edge computing module to reduce the data transmission volume. The fault self-diagnosis module verifies the acquisition consistency through cross-comparison of the heterogeneous sensor data, combines the judgment of the communication link bit error rate threshold and the detection of power supply deviation, and triggers the isolation of abnormal devices, which can significantly improve the maintainability and reliability of the system and reduce the need for manual intervention. The specific content will be described in detail later.

[0083] The high-temperature steam pipeline stress and strain state monitoring system based on characteristic safety parameters provided in this embodiment realizes the full-process closed-loop monitoring from data acquisition to state evaluation through a multi-module collaborative working mechanism. First, based on the distribution characteristics of the pipeline stress concentration areas, a heterogeneous sensor layout method combining a circular array and an axial gradient is adopted at the elbows, welds, and support parts. The strain gauges, fiber optic sensors, and vibration sensors with high-temperature resistant ceramic encapsulation and anti-corrosion alloy structures are used to synchronously collect multi-dimensional physical quantity data. Through the heterogeneous sensor layout and cross-validation mechanism, the three-dimensional strain field and vibration characteristics of the key areas of the pipeline are effectively covered. Combining the high-temperature resistant encapsulation design and anti-corrosion structure ensures the long-term stability of the sensors in extreme environments and significantly improves the reliability of feature parameter extraction. The collected original signals (i.e., the first sensing information) are processed by the data processing module to extract characteristic safety parameters such as the stress concentration coefficient, strain rate, temperature gradient, and vibration frequency. The processed data is transmitted to the edge computing module through a multi-hop network for local data processing and standardization to generate pipeline operation information that can be directly used for state evaluation. Finally, continuous monitoring and abnormal warning are realized through the server. Compared with the traditional wired scheme, it not only reduces the wiring complexity but also enhances the adaptability to complex industrial environments. The fault self-diagnosis module realizes the rapid positioning and isolation of abnormal devices through cross-validation of heterogeneous sensor data, link error rate threshold judgment, and power supply deviation detection. Combining with the deep learning analysis model on the server side, it can identify the trend of pipeline creep damage in advance and provide technical support for preventive maintenance, providing a reliable technical means to ensure the safe operation of industrial pipelines.

[0084] In some embodiments, the fault self-diagnosis module is used to achieve fault location and isolation of abnormal devices through cross-validation and link error rate analysis, including:

[0085] Perform cross-validation on multiple pieces of the first sensing information to obtain several first stress calculation values, and synchronously obtain the actual first stress value associated with the first sensing information corresponding to each first stress calculation value. Determine whether the deviation value between the first stress calculation value and the actual first stress value is within the range of a first preset threshold. If not, it indicates that the heterogeneous sensor corresponding to the actual first stress value is in abnormal acquisition;

[0086] In addition, monitor the communication link status of the wireless communication module and calculate the link error rate, and determine whether the link error rate is within the range of a second preset threshold. If not, it indicates that the wireless communication module is in abnormal communication;

[0087] In addition, monitor the power access value of each heterogeneous sensor and obtain the calibrated power value of the heterogeneous sensor;

[0088] When the power access value of the heterogeneous sensor is lower than the calibrated power value of the heterogeneous sensor, calculate the deviation between the power access value and the calibrated power value, denoted as the first deviation value;

[0089] Determine whether the first deviation value is within the range of the third preset threshold. If not, it indicates that the heterogeneous sensor corresponding to the power access value is abnormally powered.

[0090] If the abnormal device is a heterogeneous sensor, isolate the heterogeneous sensor from the fault and enable the backup acquisition mode associated with the heterogeneous sensor.

[0091] If the abnormal device is a wireless communication module, switch to the backup communication link, or adjust the backup communication mode of the wireless communication module.

[0092] In this embodiment, cross-verifying multiple first sensing information includes:

[0093] Obtain multiple first sensing information, divide the multiple first sensing information into a first verification group and a first operation group according to a preset verification mode, denote the first sensing information of the first verification group as first verification information, and denote the first sensing information of the first operation group as first operation information.

[0094] Perform data fusion on the first operation information to obtain a first stress calculation value, and perform conversion on the first verification information to obtain a first actual stress value.

[0095] Determine whether the deviation value between the first stress calculation value and the first actual stress value is within the range of the first preset threshold.

[0096] If not, it indicates that the heterogeneous sensor corresponding to the first actual stress value is abnormally acquired.

[0097] If so, it indicates that the heterogeneous sensor corresponding to the first actual stress value is operating normally.

[0098] The preset verification mode is a dynamic rotation mechanism. Preferably, 30% of the first sensing information of the sensors is randomly selected as the first verification group each time for verification, and the remaining 70% is used as the first operation group. The rotation period is synchronized with the data acquisition period. Optionally, data fusion is achieved by the weighted average method. Weight coefficients are assigned to the data of heterogeneous sensors in the same monitoring area within the operation group (such as a strain gauge weight of 0.6 and an optical fiber sensor weight of 0.4), and finally, the first stress calculation value is generated by fusion. The first actual stress value is directly obtained by converting the sensor data (i.e., the first verification information) within the first verification group through a calibration formula, and the calibration formula is a strain-voltage linear relationship model established when the sensor leaves the factory. The first preset threshold is set according to engineering experience, and when it exceeds the first preset threshold, it is determined that the corresponding heterogeneous sensor is abnormally acquired.

[0099] The link error rate is the ratio of the number of error bits to the total number of transmitted bits during the communication process. For the abnormal detection of the communication link, the second preset threshold is set based on the communication protocol standard. The fault self-diagnosis module continuously counts the error rate of the wireless communication module. When the average error rate exceeds the threshold within 3 consecutive communication cycles, a communication anomaly alarm is triggered.

[0100] It is judged whether the power supply is abnormal by the power access value and the calibrated power value of the heterogeneous sensor. Specifically, the third preset threshold for power supply anomaly judgment is defined as the deviation range of ±10% of the calibrated power value, and the calibrated power value is determined by the output voltage of the energy acquisition module under the rated working condition. When it is monitored that the power access value of a certain heterogeneous sensor is continuously lower than 10% of the calibrated power value for 5 minutes, it is determined that the power supply is abnormal. When the heterogeneous sensor is determined to have a power supply anomaly, the fault self-diagnosis module performs the following steps:

[0101] Start the backup power supply, and the backup power supply is a pre-configured redundant energy acquisition module or energy storage device;

[0102] If the backup power supply is unavailable, adjust the working mode of the heterogeneous sensor, including reducing the sampling frequency or entering the low-power mode;

[0103] Record the power supply anomaly event and transmit the energy shortage signal to the edge computing module and the server.

[0104] In the power supply anomaly handling, preferably, the backup power supply adopts a supercapacitor energy storage device, and its capacity is designed to maintain the low-power operation of the abnormal node for 12 hours. If the backup power supply is unavailable, an instruction is sent to the adjacent node through the edge computing module to reduce the sampling frequency of the abnormal node from 1Hz to 0.2Hz.

[0105] For each heterogeneous sensor, there is a corresponding backup acquisition mode. The backup acquisition mode is realized by activating multiple adjacent redundant heterogeneous sensors around the target position. When the heterogeneous sensor is abnormal, multiple heterogeneous sensors around the heterogeneous sensor are divided to detect the information of this position together, and the acquisition frequency is increased. For specific content, please refer to the following description.

[0106] The backup communication link can be understood as a separate communication hardware configuration. Starting the backup communication link means selecting another communication hardware for communication. For example, switching the backup communication link to a LoRa communication module in the 433 MHz band forms heterogeneous redundancy with the 2.4 GHz ZigBee module of the main communication link. The backup communication mode refers to the backup software communication mode in the current communication hardware configuration state, that is, selecting another communication method for communication without replacing the communication hardware. All status change records generated during the exception handling process are used by the server to generate exception logs, which are automatically synchronized to the central database after communication is restored for subsequent review and invocation. It should be noted that all exception statuses are immediately reported to the server after being determined. The backup mode is only temporarily enabled during the duration of the exception and automatically resumes the original working mode after the fault is eliminated.

[0107] The fault self-diagnosis method provided in this embodiment significantly improves the reliability of the high-temperature steam pipeline monitoring system through multi-dimensional anomaly detection and dynamic fault tolerance mechanisms. Based on the dynamic rotation mechanism, heterogeneous sensors are divided into a first verification group and a first operation group. The data of the operation group is fused through the weighted average method to generate a first stress calculation value, which is compared with the first verification group to accurately identify abnormal sensor acquisitions. At the same time, the bit error rate of the communication link is monitored, and instantaneous interference is excluded by combining the overrun determination condition lasting for 5 minutes. For heterogeneous sensor anomalies, a backup acquisition mode is adopted; for wireless communication module anomalies, a backup communication link or a backup communication mode is adopted to achieve fault isolation and function substitution. The abnormal status is reported to the server in real time and a log is generated, and the initial working mode is automatically restored after the fault is repaired. Through the dynamic rotation verification mechanism and the fusion of heterogeneous sensor data, the integrity and consistency of the monitoring data in the key area are ensured, and data loss caused by the failure of a single sensor is avoided; the communication link anomaly can be accurately identified through the second preset threshold and the continuous determination period, and the data transmission continuity is ensured by combining the main and backup communication module switching; based on the power supply deviation detection mechanism calibrated with the rated power value, combined with the backup power supply of the super capacitor and the adaptive adjustment of the sampling frequency, the basic monitoring ability of the abnormal node is maintained; the collaborative acquisition mode of adjacent redundant nodes compensates for the data in the abnormal area through spatial distribution to ensure that there is no blind area in the monitoring coverage. All backup mechanisms are used as temporary transition solutions to reduce resource occupancy while maintaining the system function. The real-time synchronization and automatic recovery mechanism of the exception log provide traceability support for operation and maintenance, and ultimately achieve high-robustness continuous monitoring of the system in extreme environments.

[0108] Please refer to Figure 1 , in some embodiments, the backup acquisition mode is configured as:

[0109] S101. Obtain the location information of the heterogeneous sensor in the abnormal state, denoted as the abnormal location information, and denote the heterogeneous sensor in the abnormal state as the abnormal sensor;

[0110] S102. With the abnormal position information as the center and a preset length as the radius, divide a backup acquisition area. Denote the heterogeneous sensors other than the abnormal sensor in the backup acquisition area as backup sensors;

[0111] And, obtain the established acquisition frequency of the abnormal sensor, denoted as the first acquisition frequency;

[0112] S103. Generate a second acquisition frequency according to the first acquisition frequency;

[0113] S104. Control the backup sensors to collect backup sensing information according to the second acquisition frequency;

[0114] S105. Perform data fusion on multiple backup sensing information to obtain a second calculated stress value;

[0115] S106. Take the second calculated stress value as the stress value corresponding to the abnormal position information.

[0116] In steps S101 and S102, the preset length is determined according to the attenuation characteristics of the pipeline stress field, usually set to 3 times the diameter of the pipe section where the abnormal position is located (for example, 0.9 meters for a DN300 pipeline). This range can cover more than 90% of the influence range of the stress concentration area. When dividing the backup acquisition area, preferentially select adjacent nodes in the same circular array as the abnormal sensor to ensure the continuity of spatial distribution.

[0117] In step S103, the second acquisition frequency is generated using the principle of linear compensation. The original first acquisition frequency (such as 1 time per hour) is increased to 1.5 times (i.e., 1.5 times per hour), and at the same time, it is required that the acquisition timestamps of the backup sensor group be synchronized with the original sensor. If the original frequency is , then the second acquisition frequency is set to , and the specific coefficient is dynamically adjusted according to the pipeline working conditions and sent to the backup sensors through the edge computing module.

[0118] In step S105, the inverse distance weighted spatial interpolation method (IDW) can be used for data fusion to calculate the contribution value of the backup sensing information collected by the backup sensors to the abnormal position. For the th backup sensor, its weight coefficient , where is the straight-line distance between the sensor and the abnormal position, is a small constant to prevent division by zero (usually taking ). The second calculated stress value is obtained through the formula , is the measured value of each backup sensor. If the types of backup sensors include strain gauges and fiber optic sensors, the data of the same type of sensors are first averaged within the group and then participate in the cross-type fusion calculation.

[0119] Furthermore, the final determination of the abnormal position stress value needs to be verified in combination with historical data: when the deviation between the second calculated stress value of the standby sensor group and the stress mean value stored in the edge computing module at this position in the recent period (such as within 24 hours) exceeds ±3%, the server-side manual review process is triggered. During the operation of the standby acquisition mode, the original data of the abnormal sensor is still retained at a sampling rate of 10% for data consistency comparison after the fault is repaired. The spatio-temporal tags of all standby sensing information are strictly aligned with the original data stream to ensure the integrity of the server-side time series analysis.

[0120] It should be noted that the number of standby sensors needs to and include at least two types of sensors, otherwise it will be automatically upgraded to the cross-region collaborative acquisition mode; the increase in the second acquisition frequency shall not cause the total energy consumption of the standby sensor group to exceed 80% of the supply capacity of the energy acquisition module; a confidence evaluation report is submitted to the server once every hour for the data fusion result, and when the confidence level is lower than 85%, the standby communication link retransmission mechanism is activated.

[0121] The standby acquisition mode provided in this embodiment realizes the improvement of the continuity and reliability of pipeline stress monitoring through a systematic abnormal data processing mechanism. Based on the preset length parameter set according to the attenuation characteristics of the pipeline stress field, it is ensured that the standby acquisition area can cover more than 90% of the influence range of the stress concentration area corresponding to the abnormal position. At the same time, by preferentially selecting the spatial distribution strategy of adjacent nodes in the same circular array, the spatial continuity of the sensor network is maintained, and the generation of monitoring blind spots is avoided. The acquisition frequency is dynamically adjusted according to the linear compensation principle, which not only ensures the data acquisition density and the timing consistency of historical monitoring data, but also adapts to different pipeline working conditions through the dynamic adjustment ability of the edge computing module, effectively balancing the monitoring accuracy and equipment energy consumption. Through the multi-source data fusion mechanism of the inverse distance weighted spatial interpolation method, the geometric relationship between the standby sensor and the abnormal position is fully considered, and the stress value is reconstructed with the inverse square of the distance as the weight coefficient, significantly improving the spatial correlation accuracy of the abnormal position stress value calculation. At the same time, the hierarchical fusion strategy set for the differences in heterogeneous sensor types effectively reduces the systematic error brought by different sensing principles through the collaborative processing of within-group averaging and cross-type calculation. In addition, the acquisition timestamp synchronization mechanism and the mathematical processing of preventing zero and tiny constants further ensure the timing alignment and calculation stability of the data fusion process. Through the technical collaboration of the trinity of spatial coverage optimization, acquisition timing enhancement, and data fusion algorithm, this mode realizes the dynamic replacement and compensation of the functions of abnormal sensors without adding new hardware devices, providing multiple guarantees for the robustness of the pipeline stress monitoring system.

[0122] In some embodiments, if the abnormal device is a wireless communication module, then switch to the standby communication link, or the adjustment of the standby communication mode of the wireless communication module includes:

[0123] The backup communication link is a pre-configured redundant communication path;

[0124] If the backup communication link is unavailable, the backup communication mode of the wireless communication module is enabled. The backup communication mode includes reducing the communication rate and switching the communication protocol;

[0125] Moreover, record communication exception events and transmit the communication exception events to the edge computing module and the server.

[0126] In this embodiment, the redundant communication path is a pre-deployed independent communication channel, whose physical link does not overlap with the main communication path. Path redundancy can be achieved through multi-hop relay nodes or different frequency bands, and basic communication capabilities can be maintained when the main link is interrupted.

[0127] Reducing the communication rate can be achieved by adjusting the modulation method or extending the transmission interval, thereby reducing the channel load and enhancing the signal anti-interference ability, enabling the communication module to still maintain the basic data transmission function in a weak signal environment; switching the communication protocol means switching to a communication standard that is more suitable for the current channel conditions under the support of the same hardware. For example, switching from the long-range low-power (LoRa) protocol to the narrowband Internet of Things (NB-IoT) protocol, or switching from the TCP protocol to the UDP protocol, and matching different service quality requirements through protocol stack reconstruction. The former focuses on reducing the transmission delay, and the latter focuses on ensuring the packet arrival rate.

[0128] Transmit the communication exception events to the edge computing module and the server for two-level collaborative processing. Specifically, the edge computing module performs real-time optimization based on event-triggered local policies, including dynamically adjusting the channel scanning period or the number of retransmissions; the server aggregates and analyzes historical exception data to establish a communication link health assessment model, providing a decision-making basis for subsequent maintenance. The record of the communication exception event includes the exception type, occurrence time, switching operation log, and communication quality indicators after switching, forming a complete exception handling evidence chain to support fault tracing and reliability verification. If both the backup communication link and the backup communication mode cannot restore the communication function, a hierarchical alarm mechanism is triggered. First, the edge computing module performs local caching and delayed uploading, and data backfilling is performed after the communication is restored to ensure the continuity of the monitoring data.

[0129] This mechanism adopts a three-layer fault tolerance design of path redundancy, mode degradation, and event collaboration. When communication is abnormal, it gives priority to ensuring the reachability of key data. At the same time, through the dynamic adaptation of the protocol and rate, it makes the best use of the existing hardware resources to maintain basic services, avoiding the overall system failure caused by a single module failure.

[0130] In some embodiments, the wireless communication module includes multiple sensor nodes, which are distributed on the pipeline in a second preset manner to form a multi-hop network. Each sensor node is configured to have data receiving and relay forwarding functions. Each sensor node establishes a communication connection with at least one neighbor node, and the neighbor node is configured as another sensor node adjacent to the sensor node.

[0131] In this embodiment, the second preset manner is a node spatial distribution rule designed according to the pipeline geometric characteristics and communication coverage requirements. The sensor nodes are evenly deployed along the axial and circumferential directions of the pipeline through an annular array or spiral arrangement. The spacing between neighbor nodes is set according to the effective transmission distance of the wireless signal, usually 1.5 - 2 times the pipeline diameter, to ensure that the signal strength between adjacent nodes is higher than the communication threshold.

[0132] The construction of the multi-hop network relies on the relay forwarding function between nodes. While each sensor node receives its own collected data, it can decode and verify the data transmitted by the upstream neighbor node and perform signal enhancement processing, and then forward it to the next-hop neighbor node through a time-division multiplexing or frequency-division multiplexing mechanism, and finally converge to the edge computing module.

[0133] A neighbor node is defined as another node that is directly adjacent to the current sensor node in physical space, and its connection relationship is automatically identified by the wireless signal strength threshold or statically bound by a preset topology table.

[0134] In this embodiment, through the spatial layout optimization of the second preset manner and the multi-hop cooperation mechanism, the network coverage range is extended under the constraint of limited power consumption. At the same time, the redundant links between neighbor nodes are used to improve the reliability of data transmission, which is especially suitable for compensating for the signal attenuation and occlusion problems in long-distance pipeline monitoring scenarios, and forms a synergy with the abnormal sensor replacement mechanism in the aforementioned standby acquisition mode to jointly enhance the fault tolerance of the monitoring system.

[0135] Please refer to Figure 2 , in some embodiments, the multi-hop network communication mode is configured to include the following steps:

[0136] S201. When a certain sensor node needs to send data, select the nearest neighbor node as the next-hop node according to the preset routing protocol;

[0137] S202. After the neighbor node receives the data, forward the data to the next relay node according to the routing protocol until the data reaches the edge computing module. The relay node is the neighbor node adjacent to the neighbor node;

[0138] S203. The edge computing module receives the data from multiple sensor nodes and performs summarization and processing;

[0139] S204. Monitor the communication status of each sensor node in the multi-hop network in real time. If a fault is detected, the routing protocol automatically adjusts the network topology and selects an alternative path for data transmission.

[0140] In step S201, the preset routing protocol can be a dynamic source routing based on distance vector (such as AODV) or a gradient routing protocol (such as RPL). The basis for selecting the next-hop neighbor node includes, but is not limited to, the remaining battery power of the neighbor node, the received signal strength indicator (RSSI), or the minimum number of hops to the edge computing module. Among them, the nearest neighbor node can be understood as the node with the fewest path hops to the edge computing module among the nodes that meet the communication link quality threshold.

[0141] In step S202, the relay node can be understood as the next-level forwarding node that meets the preferred conditions of the routing protocol in the set of adjacent nodes of the current neighbor node. For example, in gradient routing, a node with a smaller level number is preferentially selected. The data forwarding process follows a hop-by-hop acknowledgment mechanism, that is, an acknowledgment signal from the next node needs to be received after each hop of forwarding. If the acknowledgment times out, retransmission or path switching is triggered.

[0142] In step S203, the edge computing module receives data from multiple sensor nodes and performs aggregation and processing, including deduplication verification of multi-source data, timestamp alignment, and spatial interpolation fusion of stress values. For example, the data of multiple sensor nodes in the same circular array are superimposed and calculated according to preset weights.

[0143] In step S204, when the sensor node fails to respond to the heartbeat detection for a preset number of consecutive times or the packet cyclic redundancy check (CRC) error rate exceeds the threshold (such as 10%), it indicates that a communication fault has occurred. The network topology is the logical structure composed of all nodes and their communication connection relationships in the multi-hop network. The alternative path refers to an alternative transmission link that bypasses the faulty node, and its construction depends on the candidate neighbor node list maintained by the routing protocol. For example, when a certain sensor node fails, its upstream neighbor node forwards the data to the sub-optimal neighbor node according to the routing table and notifies the whole network of the topology change through flooding update messages to ensure that subsequent data is transmitted along the new path.

[0144] The multi-hop network communication mode provided in this embodiment significantly improves the communication reliability of the pipeline monitoring system through dynamic path optimization and fault tolerance mechanisms. Based on the path selection strategy of the preset routing protocol, by comprehensively evaluating the remaining power, signal reception strength, and path hop count of neighbor nodes, it ensures that data forwarding always proceeds in the direction with the optimal energy efficiency and stable link, avoiding communication interruptions caused by overloading of a single node or signal attenuation. The hop-by-hop confirmation mechanism is combined with the hierarchical preference rule of relay nodes to dynamically adjust the forwarding priority through hierarchical numbers in gradient routing. Together with the timeout retransmission and path switching mechanisms, it effectively reduces the data transmission packet loss rate. The edge computing module performs deduplication verification and timestamp alignment processing on multi-source data, eliminating data redundancy or timing disorders that may be caused by multi-hop transmission. The spatial interpolation fusion of stress values further ensures the spatial continuity of monitoring data. The fault perception mechanism accurately identifies communication abnormal nodes through the dual threshold determination of the heartbeat detection response failure count and the cyclic redundancy check error rate, triggers the routing protocol to automatically reconstruct the network topology, quickly establishes an alternative path bypassing the faulty node using the candidate neighbor node list, and achieves full network topology synchronization through flooding update, ensuring that the data transmission link can still maintain end-to-end connectivity in the scenario of node failure. Through the coordinated action of routing dynamic optimization, data integrity guarantee, and network self-healing ability, this mode constructs a low-power and highly robust closed-loop communication architecture, providing a stable multi-hop transmission foundation for long-distance pipeline monitoring.

[0145] Please refer to Figure 3 , in some embodiments, the energy harvesting module includes a vibration energy harvester, a wind speed energy harvester, a power reserve unit, and a power management circuit. The energy harvesting module is used to harvest energy from the pipeline surface and convert it into electrical energy for transmission to heterogeneous sensors, including:

[0146] S301. Set a vibration energy harvester and a wind speed energy harvester at key positions of the pipeline. The key positions include the pipeline elbow area, the weld area, the gas flow port, and the high stress concentration area. The vibration energy harvester is used to convert the mechanical vibration energy on the pipeline surface into electrical energy, and the wind speed energy harvester is used to convert the air flow velocity around the pipeline into electrical energy;

[0147] S302. Store the electrical energy generated by the vibration energy harvester and the wind speed energy harvester in the power reserve unit;

[0148] S303. The power management circuit monitors the reserved electrical energy value of the power reserve unit and the electrical energy access value of the heterogeneous sensors in real time;

[0149] S304. When the electrical energy access value of the heterogeneous sensor is lower than the calibrated electrical energy value of the heterogeneous sensor, calculate the deviation between the electrical energy access value and the calibrated electrical energy value, denoted as the second deviation value;

[0150] S305. Determine whether the reserved electric energy value is greater than the second deviation value;

[0151] S306. If so, connect the power reserve unit to the heterogeneous sensor;

[0152] S307. If not, do not connect.

[0153] In step S301, vibration energy collectors are deployed at high stress concentration positions such as the pipe elbow area and the weld area. The periodic deformation caused by the mechanical vibration of the pipe triggers the piezoelectric material or the electromagnetic induction device to generate electricity; a wind speed energy collector is set in the area with significant air flow disturbance such as the gas flow port, and the kinetic energy of the air flow is converted into electric energy through a micro turbine or a piezoelectric cantilever beam, so that the energy collection density is positively correlated with the intensity of the pipe working condition, and the energy capture efficiency is improved.

[0154] In step S302, the power reserve unit can be an energy storage device with a high cycle life, such as a super capacitor or a lithium thionyl chloride battery pack. Its function is to temporarily store and smoothly output the intermittent electric energy generated by the vibration energy collector and the wind speed energy collector, and avoid power supply interruption caused by environmental energy fluctuations.

[0155] In step S303, the power management circuit is composed of a voltage monitoring chip, a DC-DC conversion module and a microcontroller. The terminal voltage and output current of the power reserve unit are periodically sampled through an analog-to-digital converter (ADC), and the real-time reserved electric energy value is calculated in combination with the preset reserved electric energy-voltage mapping relationship; at the same time, the product of the voltage and current at the power supply port of the heterogeneous sensor is monitored, and its electric energy access value is dynamically calculated and compared with the calibrated electric energy value.

[0156] In steps S305 to S307, when the available energy of the power reserve unit is sufficient to cover the second deviation value (i.e., the current energy consumption gap of the sensor), the power management circuit closes the relay to charge the sensor; if the reserved electric energy is lower than the demand threshold, the disconnected state is maintained to prevent over-discharge damage of the energy storage device. This mechanism ensures the minimum working energy consumption of the sensor while giving priority to maintaining the basic energy storage level of the power reserve unit, and avoids the complete shutdown of the system caused by intermittent environmental energy.

[0157] The energy harvesting module provided in this embodiment realizes self-powered supply for heterogeneous sensors through environmental energy capture and intelligent power distribution mechanism. By deploying vibration energy harvesters in high stress concentration areas such as pipe elbows and welds, piezoelectric materials or electromagnetic induction devices are triggered by the mechanical vibration of the pipeline to generate electricity; wind speed energy harvesters are set in areas with significant airflows such as gas flow ports, and the kinetic energy of the airflow is converted into electrical energy through micro turbines or piezoelectric cantilever beams, so that the energy harvesting density matches the intensity of the pipeline working conditions, maximizing the utilization rate of environmental energy. The power storage unit uses supercapacitors or lithium thionyl chloride battery packs to store intermittent electrical energy, and realizes energy temporary storage and smooth output through high cycle life characteristics, avoiding power supply interruption caused by vibration or wind speed fluctuations. The power management circuit monitors the reserved power value and the sensor power access value in real time through a voltage monitoring chip and a microcontroller. When the sensor access value is lower than the calibrated value, the second deviation value is calculated and a power supply decision is triggered: if the reserved value is greater than the deviation value, the relay is closed for power replenishment; if it is insufficient, it remains open to prevent over-discharge damage to the power storage unit. This mechanism preferentially maintains the basic energy storage level through dynamic threshold control, while ensuring the minimum energy consumption requirements of the sensors, and avoiding the complete shutdown of the system due to intermittent energy. Through the collaborative design of key position directional acquisition, multi-modal energy storage and dynamic power distribution control, this embodiment converts the self-vibration of the pipeline and the surrounding air flow into a stable power supply, significantly reducing the dependence on external batteries, and is especially suitable for long-term monitoring scenarios in unattended or harsh environments. It forms an energy-data transmission collaborative optimization system with the aforementioned multi-hop network communication mode, jointly extending the overall life cycle of the monitoring system.

[0158] In some embodiments, the data processing module is used to receive and process the first sensing information and extract characteristic safety parameters, including:

[0159] Generate a data processing task according to the first sensing information, decompose the data processing task into multiple processing subtasks, and allocate the multiple processing subtasks to a multi-core CPU for parallel computing;

[0160] The multi-core CPU parallel computing includes the following steps:

[0161] Use wavelet transform or Fourier transform to compress and denoise the first sensing information to reduce the data volume of the first sensing information;

[0162] Use the Kalman filter algorithm to eliminate the sensor noise and environmental interference noise in the first sensing information to obtain the second sensing information;

[0163] Use a data fusion algorithm to fuse the second sensing information to obtain characteristic safety parameters;

[0164] Send the characteristic safety parameters to the edge computing module.

[0165] In this embodiment, the data processing task is decomposed into multiple processing subtasks and allocated to the multi-core CPU for parallel computing, which can achieve load balancing of computing resources, improve processing efficiency, shorten the overall processing latency, and meet the real-time requirements in high-sampling-rate pipeline monitoring scenarios.

[0166] Wavelet transform or Fourier transform is used to compress and denoise the first sensing information. By performing time-frequency domain conversion, high-frequency noise components and redundant data are removed, reducing the computational load and storage pressure for subsequent processing.

[0167] The Kalman filtering algorithm, through the state space model and the recursive prediction-correction mechanism, dynamically estimates the true measurement value of the sensor and suppresses noise interference. It can distinguish between the inherent noise of the sensor (such as thermal noise, quantization error) and environmental interference noise (such as electromagnetic radiation, mechanical vibration crosstalk), and adaptively adjusts the filtering intensity through the covariance matrix, improving the signal-to-noise ratio while retaining the dynamic characteristics of the signal.

[0168] The data fusion algorithm fuses the second sensing information through weighted aggregation of multi-source heterogeneous data, and assigns different weight coefficients according to the sensor type, spatial position, and historical reliability. For example, for the data of multiple vibration sensors within the same annular array, first generate a local stress distribution map through spatial interpolation, and then linearly superimpose it with the thermal expansion compensation value of the temperature sensor, and finally output the normalized characteristic safety parameter. If data fusion of fiber optic sensors and strain gauges is involved, principal component analysis (PCA) is used to extract the common characteristic vectors, eliminating the dimension difference and coupling interference between cross-type sensors.

[0169] Through the progressive processing of task parallelization, signal purification, and multi-source fusion, this embodiment accurately extracts high-dimensional sensing information into key safety indicators with limited computing resources, providing low-noise, compact, and physically meaningful input data for the decision-making analysis of the edge computing module, forming a closed-loop of computational-energy consumption co-optimization with the sustainable power supply ability of the aforementioned energy harvesting module.

[0170] In some embodiments, the edge computing module is used to process the characteristic safety parameter to obtain the first processing information, including:

[0171] The characteristic safety parameter is processed by an incremental computing method, and the data processing includes the following steps:

[0172] Perform Huffman coding on the characteristic safety parameter, and record the processed characteristic safety parameter as the first packed data;

[0173] And construct the operation state characteristic vector of the pipeline according to the characteristic safety parameter;

[0174] Perform state evaluation on the operation status feature vector to obtain an evaluation result, where the evaluation result includes a warning state, a maintenance state, an overhaul state, and a normal state;

[0175] Perform Huffman coding on the evaluation result, and record the processed evaluation result as the second packaged data;

[0176] Integrate the second packaged data with the first packaged data to form the first processed information, and transmit the first processed information to the server;

[0177] In addition, store the feature safety parameters and the evaluation result in the local database.

[0178] In this embodiment, the incremental calculation method refers to only performing local processing on the newly added or updated feature safety parameters, rather than recalculating all data. For example, only perform the encoding and vector construction operations on the feature safety parameters of the latest acquisition cycle each time, so as to reduce the consumption of computing resources and improve the real-time response speed.

[0179] Huffman coding is a variable-length lossless compression algorithm based on the frequency of character occurrences. By assigning short code elements to high-frequency feature values and long code elements to low-frequency values, the feature safety parameters are transformed into compact first packaged data. For example, if the axial stress gradient value has a high repetition rate in historical data, its corresponding coding length is significantly shorter than that of outliers that appear sparsely, achieving data volume reduction.

[0180] Preferably, the operation status feature vector includes the pipeline axial stress gradient, the circumferential strain volatility, the temperature-stress coupling coefficient, the steam pressure change entropy value, and the vibration spectrum distortion degree. The construction of the operation status feature vector is realized by extracting multi-dimensional feature safety parameters, specifically including: the pipeline axial stress gradient is calculated by dividing the stress difference between adjacent monitoring nodes by the axial spacing, the circumferential strain volatility is determined by statistically calculating the standard deviation of the strain values of the nodes in the same annular array, the temperature-stress coupling coefficient is obtained by linearly regressing the correlation coefficient of the temperature and stress data, the steam pressure change entropy value quantifies the fluctuation complexity through the Shannon entropy of the pressure time series, and the vibration spectrum distortion degree is calculated based on the mean square error between the FFT spectrum and the reference spectrum. These parameters together constitute a multi-dimensional vector describing the pipeline health status.

[0181] Preferably, the operation status feature vector can be evaluated through a multi-level state classification model, mapping the operation status feature vector to a warning state, a maintenance state, an overhaul state, or a normal state. The specific content will be described in detail later.

[0182] Furthermore, the warning state corresponds to a single parameter exceeding the first-level threshold but not reaching the dangerous level, the maintenance state is that multiple parameters continuously exceed the limit and the trend deteriorates, and the overhaul state is triggered by sudden parameter anomalies or combined pattern matching of historical fault characteristics.

[0183] The evaluation results are Huffman - encoded to generate second - packaged data, which is integrated with the first - packaged data through message encapsulation to form first - processed information. Its structure includes compressed original features and diagnostic conclusions, meeting both the low - bandwidth transmission requirements and retaining the basic data required for in - depth analysis on the server side. The local database synchronously stores uncompressed feature security parameters and evaluation results to form a complete historical status record chain, supporting offline traceability and model optimization. This process optimizes edge - side computing and communication efficiency while ensuring data validity through the collaborative design of incremental processing and hierarchical compression.

[0184] In some embodiments, the state evaluation of the operating - state feature vector is performed, and the obtained evaluation results include:

[0185] The operating - state feature vector is input into a multi - level state classification model. The operating - state feature vector includes the axial stress gradient of the pipeline, the circumferential strain volatility, the temperature - stress coupling coefficient, the entropy value of steam pressure change, and the vibration spectrum distortion degree.

[0186] The multi - level state classification model performs the following hierarchical judgments:

[0187] First - level judgment: When the temperature - stress coupling coefficient exceeds the preset dynamic threshold and the continuous duration exceeds the preset duration threshold range, a warning state is triggered.

[0188] Second - level judgment: If the product of the axial stress gradient and the circumferential strain volatility exceeds the material yield critical value, a maintenance state is triggered.

[0189] Third - level judgment: When the vibration spectrum distortion degree and the vibration spectrum distortion degree in the preset fault mode are within the range of the same cosine similarity threshold, an overhaul state is triggered.

[0190] Fourth - level judgment: If the operating - state feature vector is within the range of the preset operating threshold, a normal state is triggered.

[0191] A decision - making fusion mechanism based on fuzzy logic is established to perform weighted fusion on the judgment results output by the multi - level state classification model.

[0192] The sliding time - window algorithm is used to perform trend analysis on the preset number of evaluation results in consecutive sampling periods, and the evaluation result output is triggered when the same operating state appears continuously for a preset number of times.

[0193] In addition, the evaluation results are matched with historical fault cases in the local database to correct the credibility level of the evaluation results.

[0194] In this embodiment, the multi-level state classification model refers to a hierarchical judgment logic constructed based on physical failure mechanisms. The first to third levels respectively correspond to thermo-mechanical coupling anomalies, composite stress overrun, and vibration fault mode recognition, and the fourth level is for normal state baseline determination.

[0195] The preset dynamic threshold is determined by regression analysis of the thermal expansion coefficient of the pipeline material and historical operation data. For example, the temperature-stress coupling coefficient threshold for carbon steel pipelines is set to 0.85; preferably, the preset duration threshold is set to 10 minutes through pipeline thermal inertia experiments to avoid false triggering caused by short-term fluctuations.

[0196] The product of the axial stress gradient and the circumferential strain volatility characterizes the composite stress level, and the material yield critical value is obtained through a standard tensile test. For example, the corresponding value for X80 pipeline steel is .

[0197] The vibration spectrum distortion degree refers to the root mean square error between the current spectrum and the reference healthy spectrum. When the vibration spectrum distortion degree and the vibration spectrum distortion degree under the preset fault modes (such as weld cracks and corrosion thinning) are within the range of the same cosine similarity threshold, it indicates that the vibration characteristics highly match the typical faults.

[0198] The preset operation threshold is set by design specifications and historical statistics, including the normal fluctuation range of each characteristic parameter. For example, the entropy value threshold for steam pressure change is [0.2, 1.5].

[0199] The decision fusion mechanism based on fuzzy logic quantifies the credibility of the judgment results at each level through membership functions. The whole process can be carried out through the following steps: First, define triangular or Gaussian membership functions for the warning state, maintenance state, overhaul state, and normal state respectively, and quantify the judgment results of the multi-level classification model into confidence values between 0 and 1; second, set weights according to the state priority (such as the highest weight for the maintenance state), and perform weighted summation on the confidence levels of each state; finally, use the maximum membership principle or the weighted average method to output the final evaluation result.

[0200] The sliding time window algorithm uses a queue with a fixed length (such as 5 sampling periods) to store continuous evaluation results. When the number of occurrences of the same state within the window exceeds the set ratio (such as 3 / 5), the output is triggered to avoid interference from occasional anomalies. The similarity matching between the evaluation result and historical fault cases is achieved through the calculation of the cosine similarity of feature vectors. If the matching degree is higher than 0.85, the credibility level is increased by one level.

[0201] The pipeline operation status evaluation method provided in this embodiment significantly improves the accuracy and reliability of anomaly recognition through a hierarchical diagnosis and multi-dimensional verification mechanism. The multi-level state classification model realizes accurate positioning of fault types through a hierarchical judgment logic. The first-level judgment is based on a combined criterion of a preset dynamic threshold of the temperature-stress coupling coefficient and a preset duration threshold, effectively distinguishing short-term fluctuations from continuous thermal anomalies. The second-level judgment is based on a composite stress index that is the product of the axial stress gradient and the circumferential strain volatility, directly correlating with the critical characteristics of material yield and ensuring the clarity of the physical meaning of the maintenance state trigger. The third-level judgment is based on the matching of the vibration spectrum distortion degree with a cosine similarity threshold of a preset fault mode, realizing the feature recognition of typical faults such as weld cracks. The fuzzy logic decision fusion mechanism quantifies the state confidence by defining triangular / Gaussian membership functions and performs weighted fusion in combination with priority weights to solve the conflict problem of multi-level judgment results and reduce the risk of misjudgment. In addition, the sliding time window algorithm filters out occasional interference signals through continuous state frequency statistics, enhancing the temporal continuity of the evaluation results. The similarity matching of historical fault cases dynamically corrects the credibility level of the current results by comparing the fault feature vectors in the local database, improving the historical interpretability of the diagnostic conclusions. This embodiment forms a closed-loop evaluation system from parameter extraction to decision output by locking fault types through hierarchical criteria, balancing judgment conflicts through fuzzy fusion, filtering false alarms through time windows, and correcting confidence levels through historical cases, ensuring both a rapid response to sudden anomalies and improving diagnostic accuracy through multi-dimensional verification, forming a complete state perception chain with the feature extraction of the aforementioned data processing module.

[0202] Different from the prior art, the above technical solution has the following beneficial effects:

[0203] This technical solution provides a monitoring system for the stress and strain state of high-temperature steam pipelines based on characteristic safety parameters, constructs a full-dimensional intelligent monitoring system for the stress and strain state of high-temperature steam pipelines, and significantly improves the accuracy, reliability, and continuous operation ability of the monitoring system through the organic integration of heterogeneous sensor collaborative acquisition, multi-level data processing, and autonomous fault tolerance mechanisms. The system adopts a heterogeneous combination layout of strain gauges, fiber optic sensors, and vibration sensors, combines multi-core parallel processing and data fusion algorithms to eliminate the monitoring blind spots of a single sensing mode, accurately extracts multi-dimensional characteristic parameters such as stress concentration coefficients and strain rates, and realizes a comprehensive perception of the pipeline stress and strain state. The adaptive multi-hop network communication mode, combined with the dynamic routing protocol, while ensuring data transmission efficiency, effectively responds to local communication failures through node relay and path automatic switching mechanisms, and maintains the continuity of monitoring data. The vibration and wind energy composite acquisition system breaks through the traditional power supply limitation, combines intelligent power management strategies to achieve self-power supply of equipment, and provides a stable energy guarantee for long-term monitoring. The fault self-diagnosis module uses cross-validation and link error analysis technologies to accurately locate sensor anomalies, communication interruptions, and power supply failures, and ensures the integrity of the overall system function under local failures through fault tolerance strategies such as reconfiguring the standby acquisition area and switching redundant communication paths. The edge side uses incremental computing and Huffman coding technologies to optimize the data processing process, combines multi-level state classification models and fuzzy decision-making mechanisms to realize real-time intelligent grading evaluation of the pipeline health state, forms a closed-loop monitoring system from data acquisition, transmission, processing to diagnostic decision-making, and provides accurate and reliable technical support for the preventive maintenance of high-temperature steam pipelines.

[0204] Finally, it should be noted that although the above embodiments have been described in the text and drawings of the specification of this application, the patent protection scope of this application cannot be limited thereby. Any technical solutions obtained by equivalent structure or equivalent process substitution or modification based on the essential concept of this application and using the content recorded in the text and drawings of the specification of this application, as well as those directly or indirectly implementing the technical solutions of the above embodiments in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A high-temperature steam pipeline stress and strain state monitoring system based on characteristic safety parameters, characterized in that, Including: A sensor acquisition module, including a plurality of heterogeneous sensors, which are distributed on the pipeline in a first preset manner. The heterogeneous sensors are configured as at least two of a strain gauge, an optical fiber sensor, and a vibration sensor. The sensor acquisition module is used to acquire first sensing information; A data processing module, which is used to receive and process the first sensing information and extract characteristic safety parameters. The characteristic safety parameters include a stress concentration coefficient, a strain rate, a temperature gradient, and a vibration frequency; A wireless communication module, which is configured to use a multi-hop network communication mode to transmit the first sensing information and / or the characteristic safety parameters to an edge computing module; An energy harvesting module, which is arranged on the surface of the pipeline. The energy harvesting module is electrically connected to the sensor acquisition module. The energy harvesting module is used to harvest energy from the surface of the pipeline and convert it into electric energy for transmission to the heterogeneous sensors; An edge computing module, which is used to process the characteristic safety parameters to obtain first processed information and transmit the first processed information to a server. The first processed information is the operation information of the pipeline; A fault self-diagnosis module, which is used to detect an abnormal state. The abnormal state is configured as any one of acquisition abnormality, communication abnormality, and power supply abnormality. The fault self-diagnosis module is used to achieve fault location and isolation of abnormal devices through cross-verification and link error rate analysis, including: Performing cross-verification on a plurality of the first sensing information to obtain a plurality of first stress calculation values, synchronously obtaining the actual first stress value associated with the first sensing information corresponding to each first stress calculation value, and determining whether the deviation value between the first stress calculation value and the actual first stress value is within the range of a first preset threshold. If not, it indicates that the heterogeneous sensor corresponding to the actual first stress value is in an acquisition abnormality; And, monitoring the communication link state of the wireless communication module and calculating the link error rate, and determining whether the link error rate is within the range of a second preset threshold. If not, it indicates that the wireless communication module is in a communication abnormality; And, monitoring the power access value of each heterogeneous sensor and obtaining the calibrated power value of the heterogeneous sensor; When the power access value of the heterogeneous sensor is lower than the calibrated power value of the heterogeneous sensor, calculating the deviation between the power access value and the calibrated power value, denoted as a first deviation value; Determining whether the first deviation value is within the range of a third preset threshold. If not, it indicates that the heterogeneous sensor corresponding to the power access value is in a power supply abnormality; If the abnormal device is a heterogeneous sensor, fault-isolate the heterogeneous sensor and enable a backup acquisition mode associated with the heterogeneous sensor. The backup acquisition mode is configured as: Obtaining the position information of the heterogeneous sensor in the abnormal state, denoted as abnormal position information, and denoting the heterogeneous sensor in the abnormal state as an abnormal sensor; Dividing a backup acquisition area with the abnormal position information as the center and a preset length as the radius, and denoting the heterogeneous sensors other than the abnormal sensor in the backup acquisition area as backup sensors; And, obtain the established acquisition frequency of the abnormal sensor, denoted as the first acquisition frequency; Generate a second acquisition frequency according to the first acquisition frequency; Control the standby sensor to acquire standby sensing information according to the second acquisition frequency; Perform data fusion on multiple pieces of the standby sensing information to obtain a second calculated stress value; Use the second calculated stress value as the stress value corresponding to the abnormal position information; If the abnormal device is a wireless communication module, switch to a standby communication link, or adjust the standby communication mode of the wireless communication module; The server is configured to receive the first sensing information and / or the first processing information and / or the characteristic safety parameter and continuously monitor the working state of the pipeline.

2. The high-temperature steam pipeline stress and strain state monitoring system for characteristic safety parameters as described in claim 1, wherein If the abnormal device is a wireless communication module, switching to a standby communication link, or adjusting the standby communication mode of the wireless communication module includes: The standby communication link is a pre-configured redundant communication path; If the standby communication link is unavailable, enable the standby communication mode of the wireless communication module, and the standby communication mode includes reducing the communication rate and switching the communication protocol; And, record the communication abnormal event and transmit the communication abnormal event to the edge computing module and the server.

3. The high-temperature steam pipeline stress and strain state monitoring system for characteristic safety parameters as described in claim 1, characterized in that, The wireless communication module includes a plurality of sensor nodes, and is distributed on the pipeline in a second preset manner to form a multi-hop network. Each sensor node is configured to have data receiving and relay forwarding functions. Each sensor node establishes a communication connection with at least one neighbor node, and the neighbor node is configured as other sensor nodes adjacent to the sensor node.

4. The high-temperature steam pipeline stress and strain state monitoring system for characteristic safety parameters as described in claim 3, characterized in that, The multi-hop network communication mode is configured to include the following steps: When a certain sensor node needs to send data, select the nearest neighbor node as the next-hop node according to a preset routing protocol; After the neighbor node receives the data, forward the data to the next relay node according to the routing protocol until the data reaches the edge computing module, and the relay node is the neighbor node adjacent to the neighbor node; The edge computing module receives data from multiple sensor nodes and performs summarization and processing; Real-time monitor the communication status of each sensor node in the multi-hop network. If a fault is detected, the routing protocol automatically adjusts the network topology and selects a standby path for data transmission.

5. The high-temperature steam pipeline stress and strain state monitoring system for characteristic safety parameters as described in claim 1, characterized in that, The energy harvesting module includes a vibration energy harvester, a wind speed energy harvester, a power storage unit, and a power management circuit. The energy harvesting module is used to harvest energy from the pipeline surface and convert it into electrical energy for delivery to the heterogeneous sensors, including: Set a vibration energy harvester and a wind speed energy harvester at key positions of the pipeline. The key positions include pipeline elbow areas, weld areas, gas flow ports, and high stress concentration areas. The vibration energy harvester is used to convert the mechanical vibration energy on the pipeline surface into electrical energy, and the wind speed energy harvester is used to convert the air flow velocity around the pipeline into electrical energy; Store the electrical energy generated by the vibration energy harvester and the wind speed energy harvester in the power storage unit; The power management circuit monitors in real time the reserved electric energy value of the power reserve unit and the electric energy access value of the heterogeneous sensor; When the electric energy access value of the heterogeneous sensor is lower than the calibrated electric energy value of the heterogeneous sensor, calculate the deviation between the electric energy access value and the calibrated electric energy value, denoted as the second deviation value; Judge whether the reserved electric energy value is greater than the second deviation value. If so, connect the power reserve unit to the heterogeneous sensor; If not, do not connect.

6. The high-temperature steam pipeline stress and strain state monitoring system for characteristic safety parameters as described in claim 1, characterized in that, The data processing module is used to receive and process the first sensing information and extract characteristic safety parameters, including: Generate a data processing task according to the first sensing information, decompose the data processing task into multiple processing subtasks, and allocate the multiple processing subtasks to a multi-core CPU for parallel computing; The parallel computing of the multi-core CPU includes the following steps: Use wavelet transform or Fourier transform to compress and denoise the first sensing information to reduce the data volume of the first sensing information; Use the Kalman filter algorithm to eliminate the sensor noise and environmental interference noise in the first sensing information to obtain the second sensing information; Use a data fusion algorithm to fuse the second sensing information to obtain the characteristic safety parameters; Send the characteristic safety parameters to the edge computing module.

7. The high-temperature steam pipeline stress and strain state monitoring system for characteristic safety parameters according to claim 1, characterized in that the edge The calculation module is used to process the characteristic safety parameters to obtain the first processed information, including: Use an incremental calculation method to process the characteristic safety parameters. The data processing includes the following steps: Perform Huffman coding on the characteristic safety parameters, and record the processed characteristic safety parameters as the first packed data; And construct an operating state characteristic vector of the pipeline according to the characteristic safety parameters; Perform a state evaluation on the operating state characteristic vector to obtain an evaluation result, and the evaluation result includes a warning state, a maintenance state, an overhaul state, and a normal state; Perform Huffman coding on the evaluation result, and record the processed evaluation result as the second packed data; Integrate the second packed data with the first packed data to form the first processed information, and transmit the first processed information to the server; And store the characteristic safety parameters and the evaluation result in the local database.

8. The high-temperature steam pipeline stress and strain state monitoring system for characteristic safety parameters as claimed in claim 7, characterized in that, Perform a state evaluation on the operating state characteristic vector to obtain an evaluation result, including: Input the operating state characteristic vector into a multi-level state classification model. The operating state characteristic vector includes a pipeline axial stress gradient, a circumferential strain volatility, a temperature-stress coupling coefficient, a steam pressure change entropy value, and a vibration spectrum distortion degree; The multi-level state classification model performs the following hierarchical judgments: First-level judgment: When the temperature-stress coupling coefficient exceeds a preset dynamic threshold and the continuous duration exceeds the range of a preset duration threshold, trigger a warning state; Second-level judgment: If the product of the axial stress gradient and the circumferential strain volatility exceeds the material yield critical value, trigger a maintenance state; Third-level judgment: When the vibration spectrum distortion degree and the vibration spectrum distortion degree under a preset fault mode are within the range of the same cosine similarity threshold, trigger an overhaul state; Fourth-level judgment: If the operation status feature vector is within the preset operation threshold range, the normal state is triggered; Establish a decision fusion mechanism based on fuzzy logic to perform weighted fusion on the judgment results output by the multi-level state classification model; Adopt a sliding time window algorithm to perform trend analysis on the evaluation results of a preset number of consecutive sampling periods, and trigger the output of the evaluation results when the same operation state appears continuously for a preset number of times; And, perform similarity matching between the evaluation results and historical fault cases in the local database to correct the credibility level of the evaluation results.

Images