Chemical pipeline flange loosening leakage early warning method and system based on distributed optical fiber sensing

Abstract

The invention provides a chemical pipeline flange loosening leakage early warning method and system based on distributed optical fiber sensing, and relates to the technical field of chemical safety monitoring. According to the system, a double-parameter distributed optical fiber sensor is arranged on the outer circle of a flange in a spiral winding mode, temperature and vibration signals of the flange are collected in real time and preprocessed, and the real-time change value of the temperature difference between a leakage point and the environment and the vibration energy value of the 20-200 Hz frequency band are obtained. Establishing a multi-modal signal fusion criterion model which comprises a vibration criterion (the sudden increase of vibration energy is greater than or equal to 3dB) and lasts for more than 10 seconds, and calculating the looseness degree of the bolt according to the vibration criterion; the real-time change of the temperature criterion is greater than or equal to + / -0.5 DEG C and lasts for more than 10 seconds; and a gradient early warning mechanism is implemented through the data model. The technology fills the blank of flange looseness monitoring and leakage early warning, realizes the sensitivity and reliability of flange looseness and leakage monitoring, and ensures the safety of chemical production.

Description

Technical Field

[0001] The present invention relates to the technical field of chemical safety monitoring, and particularly to a method and system for early warning of flange loosening and leakage of chemical pipelines based on distributed optical fiber sensing. Background Art

[0002] During the chemical production process, the reliability of the pipeline flange connection is directly related to the safety of hazardous chemical transportation. Currently, the period of manual inspection is usually more than 24 hours. Such a long time interval makes it impractical to monitor micro-loosening in real time and potential problems cannot be detected in a timely manner. The electrical sensors used, such as piezoelectric accelerometers, have the risk of electromagnetic interference, which cannot meet the strict requirements of explosion protection in the chemical industry. The current point-type monitoring devices (such as pressure transmitters) have monitoring blind spots, resulting in a false alarm rate of more than 30%, significantly reducing the effectiveness of safety monitoring.

[0003] The initial characteristic changes of flange leakage are very weak. Conventional point-type sensors are difficult to capture the temperature changes around the flange. For example, the leakage point is located at a position far from the point-type sensor; vibration acceleration sensors need to be installed on all bolts fixing the flange to monitor the loosening of each bolt; moreover, point-type electrical sensors and wireless signal transmission are vulnerable to electromagnetic interference; and the inability to send data in real time online further increases the difficulty of monitoring. The relationship between the pre-tightening force attenuation of bolts and the development of leakage is non-linear, and the lack of an effective prediction model makes the early warning of leakage a challenging task.

[0004] The vibration monitoring solutions proposed in the prior art still rely on discrete sensors and cannot achieve continuous spatial coverage; and although the prior art uses optical fiber sensing, it often only targets a single parameter (temperature or vibration), with a high false alarm rate. Therefore, there is an urgent need for a flange loosening and leakage early warning technology with high sensitivity and high reliability. Summary of the Invention

[0005] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a method and system for early warning of flange loosening and leakage of chemical pipelines based on distributed optical fiber sensing. The present invention adopts distributed optical fiber sensing technology and a signal processing method with high sensitivity, providing a scientific, safe and effective solution for the early warning of flange loosening and leakage of chemical pipelines, and greatly improving the safety of chemical production.

[0006] To achieve the above purpose, the present invention provides the following solutions:

[0007] A method for early warning of flange loosening and leakage of chemical pipelines based on distributed optical fiber sensing, comprising:

[0008] Laying a temperature and vibration dual-parameter distributed optical fiber sensor in a spiral winding manner in the gap between flanges;

[0009] The temperature signal and vibration signal of the flange are collected in real time by a distributed optical fiber demodulator, and the temperature signal and the vibration signal are respectively preprocessed to obtain the real-time change value of the temperature difference between the leakage point and the environment and the vibration energy value in the frequency band of 20 - 200 Hz;

[0010] A multimodal signal fusion criterion model is established according to the real-time change value and the vibration energy value; the multimodal signal fusion criterion model includes a temperature criterion and a vibration criterion; the temperature criterion is that the real-time change value ≥ ±0.5 °C and lasts for 10 seconds; the vibration criterion is that the vibration energy suddenly increases by more than 3 dB and lasts for 10 seconds;

[0011] Calculate the degree of bolt loosening according to the vibration signal;

[0012] Execute a gradient warning mechanism according to the multimodal signal fusion criterion model and the degree of bolt loosening.

[0013] Preferably, the number of turns of the spiral winding method is 4 - 5 turns to achieve full circumferential monitoring of the flange.

[0014] Preferably, the same pitch is maintained between each turn, and after the distributed optical fiber sensor is wound, it is fixed and protected by a thermal conductive silica gel layer with a thickness of 0.2 mm.

[0015] Preferably, the calculation formula for the degree of bolt loosening is:

[0016]

[0017] where F is the degree of bolt loosening, K0 is the initial stiffness eigenvalue, K t is the real-time stiffness eigenvalue, and α is the flange material coefficient; the extraction formula for the real-time stiffness eigenvalue K t is: K t = ∫f(t)·e -t / τ dt; where τ is the characteristic time constant of the flange structure, and f(t) is the temperature change amount on the flange surface or around the bolt.

[0018] Preferably, the flange material coefficient α is calibrated through experiments, specifically including the fitting analysis of the elastic modulus, Poisson's ratio of the flange material and the bolt torque - pre-tightening force relationship.

[0019] Preferably, the thermal conductivity of the thermal conductive silica gel layer ≥ 2.5 W / (m·K), and the thermal conductive silica gel layer is used to enhance the coupling effect between the energy of flange leakage and the temperature of the distributed optical fiber sensor laid between the flange gaps.

[0020] Preferably, executing the gradient warning mechanism according to the multimodal signal fusion criterion model and the degree of bolt loosening includes:

[0021] When only the temperature criterion or the vibration criterion is satisfied and the time lasts for 10 seconds, a first-level early warning is triggered;

[0022] When the temperature criterion and the vibration criterion are both satisfied, the time lasts for 20 seconds, and the degree of bolt loosening decreases by 10%-20%, a second-level early warning is triggered;

[0023] When the temperature criterion and the vibration criterion are both satisfied, the time lasts for 30 seconds, and the degree of bolt loosening decreases by >20%, a third-level early warning is triggered and the maintenance and repair plan is automatically associated.

[0024] Preferably, after the first-level early warning is triggered, the following operations are performed:

[0025] a. Display the flange number, location, yellow alarm logo and abnormal parameter type on the operation interface of the monitoring terminal;

[0026] b. Record the abnormal signal data and generate a preliminary diagnosis report;

[0027] c. Send an alarm through the acoustic-optic alarm device;

[0028] After the second-level early warning is triggered, the following operations are performed:

[0029] a. Display the flange number, location, orange alarm logo and the percentage of bolt loosening reduction on the monitoring terminal;

[0030] b. Send a warning notice containing the flange number and location coordinates to the mobile terminal of the designated maintenance personnel;

[0031] c. Activate the standby monitoring device to enhance the data acquisition frequency;

[0032] After the third-level early warning is triggered, the following operations are performed:

[0033] a. Display the flange number, location, red alarm logo and the leakage level assessment result on the monitoring terminal;

[0034] b. Automatically close the valve at the front end of the leaking flange;

[0035] c. Send a work order containing emergency operation instructions, leaking valve number, three-dimensional coordinates and safety protection requirements to the maintenance and repair terminal;

[0036] d. Synchronize the alarm information to the superior supervision system through the industrial Internet of Things platform.

[0037] A chemical pipeline flange loosening and leakage early warning system based on distributed optical fiber sensing, comprising:

[0038] An optical fiber winding unit is used to lay a temperature and vibration dual-parameter distributed optical fiber sensor in a spiral winding manner in the gap between flanges;

[0039] A signal preprocessing unit is used to collect the temperature signal and vibration signal of the flange in real time through a distributed optical fiber demodulator, and preprocess the temperature signal and the vibration signal respectively to obtain the real-time change value of the temperature difference between the leakage point and the environment and the vibration energy value in the 20 - 200 Hz frequency band;

[0040] A criterion construction unit is used to establish a multi-modal signal fusion criterion model according to the real-time change value and the vibration energy value; the multi-modal signal fusion criterion model includes a temperature criterion and a vibration criterion; the temperature criterion is that the real-time change value ≥ ±0.5 °C and lasts for 10 seconds; the vibration criterion is that the vibration energy suddenly increases by more than 3 dB and lasts for 10 seconds;

[0041] A loosening calculation unit is used to calculate the degree of bolt loosening according to the vibration signal;

[0042] An early warning unit is used to execute a gradient early warning mechanism according to the multi-modal signal fusion criterion model and the degree of bolt loosening.

[0043] According to the specific embodiments provided by the present invention, the following technical effects are disclosed by the present invention:

[0044] The present invention provides a method and system for early warning of loosening and leakage of chemical pipeline flanges based on distributed optical fiber sensing, including: laying a temperature and vibration dual-parameter distributed optical fiber sensor in a spiral winding manner in the gap between flanges; collecting the temperature signal and vibration signal of the flange in real time through a distributed optical fiber demodulator, and preprocessing the temperature signal and the vibration signal respectively to obtain the real-time change value of the temperature difference between the leakage point and the environment and the vibration energy value in the 20 - 200 Hz frequency band; establishing a multi-modal signal fusion criterion model according to the real-time change value and the vibration energy value; the multi-modal signal fusion criterion model includes a temperature criterion and a vibration criterion; the temperature criterion is that the real-time change value ≥ ±0.5 °C and lasts for 10 seconds; the vibration criterion is that the vibration energy suddenly increases by more than 3 dB and lasts for 10 seconds; calculating the degree of bolt loosening according to the vibration signal; executing a gradient early warning mechanism according to the multi-modal signal fusion criterion model and the degree of bolt loosening. The present invention adopts distributed optical fiber sensing technology and a high-sensitivity signal processing method, provides a scientific, safe and effective solution for early warning of loosening and leakage of chemical pipeline flanges, and greatly improves the safety of chemical production. Description of the Drawings

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0046] Figure 1 The flowchart of the method provided by the embodiment of the present invention;

[0047] Figure 2 The schematic diagram of the system structure provided by the embodiment of the present invention. Detailed implementation manners

[0048] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0049] The purpose of the present invention is to provide a method and system for early warning of chemical pipeline flange loosening and leakage based on distributed optical fiber sensing. By using distributed optical fiber sensing technology and a high-sensitivity signal processing method, a scientific, safe, and effective solution is provided for early warning of chemical pipeline flange loosening and leakage, greatly improving the safety of chemical production.

[0050] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the drawings and specific implementation manners.

[0051] Figure 1 The flowchart of the method provided by the embodiment of the present invention, as Figure 1 shown, the present invention provides a method for early warning of chemical pipeline flange loosening and leakage based on distributed optical fiber sensing, including:

[0052] Step 100: Dispose the temperature and vibration dual-parameter distributed optical fiber sensor in a spiral winding manner in the gap between the flanges;

[0053] Step 200: Real-time collect the temperature signal and vibration signal of the flange through a distributed optical fiber demodulator, and respectively preprocess the temperature signal and vibration signal to obtain the real-time change value of the temperature difference between the leakage point and the environment and the vibration energy value in the frequency band of 20 - 200 Hz;

[0054] Step 300: Establish a multimodal signal fusion criterion model based on the real-time change value and the vibration energy value; the multimodal signal fusion criterion model includes a temperature criterion and a vibration criterion; the temperature criterion is that the real-time change value ≥ ±0.5°C and lasts for 10 seconds; the vibration criterion is that the vibration energy suddenly increases by more than 3 dB and lasts for 10 seconds;

[0055] Step 400: Calculate the degree of bolt loosening according to the vibration signal;

[0056] Step 500: Execute a gradient warning mechanism according to the multimodal signal fusion criterion model and the degree of bolt loosening.

[0057] Preferably, the number of layers of the spiral winding method is 4 - 5 turns, the total winding length of each layer is 2.5 m / flange, and the adjacent layers are cross-covered to achieve circumferential global monitoring of the flange. Preferably, the pitch of the spiral winding method is 3 - 5 mm, and a heat-conducting silica gel layer with a filling thickness of 0.2 mm is provided between each layer. The thermal conductivity of the heat-conducting silica gel layer ≥ 2.5 W / (m·K), and the heat-conducting silica gel layer is used to enhance the coupling sensitivity of the temperature signal and the flange surface.

[0058] Specifically, in this embodiment, a numerically controlled winding device is used to perform precise winding operations on the outer circumferential surface of the flange. The Φ0.9 mm armored distributed optical fiber is uniformly arranged along the circumference of the flange in a spiral path (pitch 3 - 5 mm). After each layer of winding is completed, the winding start angle is adjusted (usually offset by 120°) to form a cross-covered structure for the optical fibers in adjacent layers. Each flange is wound 4 - 5 turns with a total length of 2.5 m to ensure that there is no monitoring blind area in the entire flange area. During the winding process, the pitch accuracy is calibrated in real time by a laser rangefinder (error ±0.2 mm), and a flexible fixture is used to fix the end of the optical fiber to avoid signal distortion caused by uneven tension. Demonstratively, the installation position of the temperature and vibration dual-parameter distributed optical fiber sensor in this embodiment is around the flange gap for one week, and it is wound 4 - 5 turns according to the flange specification model.

[0059] Optionally, the same pitch is maintained between each turn. After the distributed optical fiber sensor is wound, it is fixed and protected by a thermal conductive silicone layer with a thickness of 0.2 mm. After each layer of optical fiber winding is completed, a high-precision dispensing machine is used to uniformly inject thermal conductive silicone (thermal conductivity ≥ 2.5 W / (m·K)) along the winding gap. The thickness of the glue layer is controlled at 0.2 mm ± 0.05 mm by a micron-level scraper. After the silicone is cured, its flexible characteristics can not only closely fit the flange surface but also adapt to the thermal expansion and contraction deformation of the pipeline. To verify the thermal conductivity, the transient plane source method (TPS) is used to test the thermal conductivity of the silicone layer to ensure that it meets the design requirement of ≥ 2.5 W / (m·K). The finally formed "optical fiber - silicone" composite structure can increase the temperature signal transmission efficiency by 40%, significantly enhancing the detection sensitivity of micro-leakage (ΔT ≥ 0.8 °C). Exemplarily, the thermal conductive silicone layer is used to enhance the coupling effect between the energy of flange leakage and the temperature of the distributed optical fiber sensor laid between the flange gaps.

[0060] Specifically, step 200 of this embodiment includes:

[0061] The distributed optical fiber demodulator injects into the distributed optical fiber sensor through a pulsed modulation light source (wavelength 1550 nm, pulse width 10 ns) and receives the backscattered light signal. The coherent OTDR (optical time domain reflectometry) technology is used to analyze temperature and vibration information: the temperature signal is calculated by the Raman scattering light intensity ratio (anti-Stokes / Stokes), with a spatial resolution of 0.5 m; the vibration signal is demodulated by the phase-sensitive optical time domain reflectometry (Φ-OTDR) technology, with a sampling rate of 200 Hz. The demodulator is built-in with a dual-channel processor to synchronously output the temperature change value (temperature resolution ±0.1 °C, temperature accuracy of 0.5 °C) and the vibration original waveform to ensure the alignment of space-time data.

[0062] After the temperature signal eliminates environmental noise through moving average filtering (window width 5 s), the temperature difference ΔT between the leakage point and the environmental reference point is calculated; the vibration signal extracts the energy in the 20 - 200 Hz frequency band through the FFT transform weighted by the Hanning window, and the baseline calibration algorithm (current energy / historical mean logarithmic ratio) is used to determine the threshold for a sudden increase of 3 dB. The preprocessed dual-parameters are transmitted to the edge computing unit through the Modbus TCP protocol, and the time delay is controlled within 50 ms to meet the real-time requirement.

[0063] Optionally, step 300 of this embodiment includes:

[0064] This embodiment uses a timestamp alignment mechanism to ensure synchronous analysis of temperature and vibration signals, and establishes a joint analysis window (a 10-second sliding window with a step of 1 second) through the edge computing unit. Then, this embodiment monitors the change curve of ΔT in real time, and triggers the condition when 10 consecutive sampling points (sampling interval of 1 second) all satisfy ΔT≥±0.5°C; Implementation of the vibration criterion: Calculate the decibel difference between the vibration energy in the current 20 - 200 Hz frequency band and the baseline energy (taking the moving average of the previous 30 minutes), and determine it to be effective when there are more than 3 sudden increases of ≥3 dB within 10 seconds. The dual-parameter signals are comprehensively judged through a weighted decision fusion algorithm (temperature weight 0.6, vibration weight 0.4), and the weight coefficients are optimized and determined through on-site calibration tests.

[0065] For different pipe diameters and medium characteristics, the model is built with an adaptive adjustment module: when the ambient temperature fluctuates greatly, the temperature criterion threshold is automatically increased to 2.0°C (dynamically adjusted by analyzing the historical temperature standard deviation); wavelet noise reduction processing is applied to the vibration signal to eliminate broadband interference such as mechanical vibration. The system automatically updates the vibration energy baseline value every 24 hours to prevent misjudgment caused by long-term drift. All criterion trigger events record the original waveform segments (30 seconds before to 10 seconds after the trigger) for subsequent model optimization and accident traceability. The criterion logic is accelerated by FPGA hardware to ensure that the entire analysis process is completed within 200 ms.

[0066] Exemplarily, the adaptive adjustment module of this embodiment adopts a double-layer LSTM neural network architecture (128 hidden units in each layer). The input layer receives the time series feature matrix, including: temperature signal (ΔT and its first derivative), vibration energy value (20 - 200 Hz frequency band), and ambient temperature fluctuation standard deviation (data of the past 1 hour). The training data is sourced from the historical leakage event dataset (including more than 200 cases of different pipe diameters and media), and the weighted cross-entropy loss function is minimized through the Adam optimizer (learning rate 0.001). The model output is the dynamically adjusted parameters: the temperature criterion threshold (adjustable range of ±0.5 - 2.5°C) and the vibration energy weight coefficient (in the range of 0.3 - 0.7), and online incremental learning is performed every 8 hours to update the network parameters.

[0067] During deployment, the LSTM model is optimized by the TensorRT engine to achieve real-time inference with a latency of 8 ms on the edge computing unit (such as Jetson Xavier NX). The specific workflow is as follows:

[0068] (1) Standardize the input signal (Z-score) and segment it with a sliding window (window length 60 seconds, step 1 second);

[0069] (2) According to the current inference result, if the standard deviation of the ambient temperature exceeds the set threshold (such as ±3°C), the temperature criterion threshold is automatically increased to 2.0°C, and at the same time, the vibration criterion weight is decreased to 0.3;

[0070] (3) Compare the warning result with the actual maintenance record, and the error cases are automatically added to the retraining data set to continuously optimize the model. An anomaly detector (Isolation Forest algorithm) is built into the module, and when the input features deviate from the training distribution, the model reset alarm is triggered.

[0071] The core of the LSTM module in this embodiment lies in transforming the traditional fixed threshold criterion into a dynamically adjustable parameter system. Through temporal feature learning, the model can identify the differences between environmental disturbances (such as day-night temperature differences) and real leakage signals: for example, when a slow temperature gradient (caused by solar radiation) is detected, even if ΔT reaches ±0.5°C, no warning is triggered; while high sensitivity is maintained for rapid temperature rises (such as medium leakage). The dynamic adjustment of the vibration weight solves the problem of large differences in vibration energy conduction for different media (liquid / gas). The hardware acceleration design ensures the stable operation of the algorithm on industrial field devices with limited resources, and its adaptive performance can reduce the false alarm rate by 42% after testing (compared with the static model).

[0072] Preferably, the calculation formula for the degree of bolt loosening is:

[0073]

[0074] where F is the degree of bolt loosening, K0 is the initial stiffness eigenvalue, K t is the real-time stiffness eigenvalue, and α is the flange material coefficient; the extraction formula for the real-time stiffness eigenvalue K t is: K t = ∫f(t)·e -t / τ dt; where τ is the characteristic time constant of the flange structure, and f(t) is the temperature change amount on the flange surface or around the bolt.

[0075] Exemplarily, the temperature change amount on the flange surface or around the bolt is obtained through distributed fiber optic sensing technology. This technology utilizes the temperature measurement ability of optical fibers. By installing fiber optic sensors on the flange or bolt, the temperature change is monitored in real time. The fiber optic sensor can sense the thermal expansion of the material and temperature fluctuations, and then convert them into electrical signals. By analyzing these signals, the temperature data of the flange and its surrounding area at different time points can be obtained. These temperature data are collected and recorded, reflecting the influence of factors such as the working environment and fluid temperature change on the flange connection state, thereby providing a basis for judging whether the flange is loose. By continuously monitoring, the dynamic change of temperature can be effectively tracked, helping the warning system to identify potential leakage risks in a timely manner.

[0076] Preferably, the flange material coefficient α is calibrated through experiments, specifically including the fitting analysis of the elastic modulus, Poisson's ratio of the flange material, and the bolt torque-preload force relationship.

[0077] Specifically, in this embodiment, non-contact monitoring of the bolt preload force is achieved through vibration signal feature extraction. First, at the initial stage of flange installation (when the bolts are not loose), the reference vibration signal is collected, and the initial stiffness eigenvalue K0 is extracted through frequency response function analysis. This value reflects the overall rigidity of the flange-bolt connection structure. During the real-time monitoring stage, the system continuously analyzes the time-domain decay characteristics of the vibration signal, and the exponential decay fitting method is used to extract the current stiffness eigenvalue K t , where the characteristic time constant τ is calibrated through impact hammer testing and represents the energy dissipation rate of a specific flange structure. The calibration of the flange material coefficient α is completed through laboratory destructive tests: the bolt elongation (strain gauge) and vibration signals are synchronously measured at different torque levels (20%-100% of the rated torque), a three-dimensional relationship surface of "torque-preload force-stiffness" is established, and finally the α value is obtained through least squares fitting. This system can achieve a preload force evaluation accuracy of ±8%, and there is no need to disassemble the flange or interrupt production.

[0078] Specifically, in this embodiment, the establishment of the multi-modal signal fusion criterion model adopts a comprehensive statistical method based on time series analysis. First, the distributed optical fiber demodulator is used to collect the temperature change value ΔT and the vibration signal energy value E to form a time series data stream. Subsequently, dynamic windows are set for these two parameters respectively, and the moving average method is used to filter the data to eliminate short-term noise interference. Within a 10-second moving window, if the absolute value of ΔT continuously exceeds ±0.5°C and there are 3 sudden increases (≥3 dB), that is, the conditions of both the temperature criterion and the vibration criterion are met, the model determines that there is a risk of flange loosening or potential leakage.

[0079] To further improve the accuracy of the criterion model, this embodiment adopts a weighted decision fusion algorithm, where the weight of the temperature data is set to 0.6 and the weight of the vibration data is 0.4. The weight coefficients are optimized through on-site measured data to ensure that the model can adapt to changes under different working conditions. When the combination of temperature change and vibration sudden increase occurs, the command signal will be issued to trigger the subsequent early warning mechanism.

[0080] Preferably, a gradient early warning mechanism is executed according to the multi-modal signal fusion criterion model and the degree of bolt loosening, including:

[0081] When only the temperature criterion or the vibration criterion is satisfied and the time lasts for 10 seconds, a first-level early warning is triggered;

[0082] When both the temperature criterion and the vibration criterion are satisfied, the time lasts for 20 seconds, and the degree of bolt loosening decreases by 10%-20%, a second-level early warning is triggered;

[0083] When the temperature criterion and the vibration criterion are simultaneously satisfied and the time lasts for 30 seconds, and the degree of bolt loosening decreases by more than 20%, a third-level warning is triggered, and the maintenance and emergency repair plan is automatically associated.

[0084] Preferably, after the first-level warning is triggered, the following operations are performed:

[0085] a. Display the flange number, position, yellow alarm flag, and abnormal parameter type on the operation interface of the monitoring terminal;

[0086] b. Record the abnormal signal data and generate a preliminary diagnosis report;

[0087] c. Send an alarm through the audible and visual alarm device;

[0088] After the second-level warning is triggered, the following operations are performed:

[0089] a. Display the flange number, position, orange alarm flag, and the percentage of bolt loosening decrease on the monitoring terminal;

[0090] b. Send a warning notice containing the flange number and position coordinates to the mobile terminal of the designated maintenance personnel;

[0091] c. Activate the standby monitoring device to enhance the data acquisition frequency;

[0092] After the third-level warning is triggered, the following operations are performed:

[0093] a. Display the flange number, position, red alarm flag, and the leakage level assessment result on the monitoring terminal;

[0094] b. Automatically close the valve in front of the leaking flange;

[0095] c. Send a work order containing emergency operation instructions, the leaking valve number, three-dimensional coordinates, and safety protection requirements to the maintenance and emergency repair terminal;

[0096] d. Synchronize the alarm information to the superior supervision system through the industrial Internet of Things platform.

[0097] Specifically, when a single parameter (temperature or vibration) is detected as abnormal in this embodiment, the timer module is activated to continuously monitor whether the abnormal signal is stably maintained for 30 seconds. After confirmation, the HMI interface of the monitoring terminal calls the preset alarm template to display a yellow identifier and the type of abnormality (such as "Temperature abnormality: ΔT = 1.8°C"). At the same time, the SQL database recording function is triggered to store the original signal segment (including 30 seconds of data before and after the trigger) and generate a diagnostic report (including timestamp, flange number, and preliminary fault speculation). The audible and visual alarm device receives instructions through the GPIO interface to drive the low-frequency buzzer (frequency 500Hz) and the LED yellow light to flash, avoiding high-frequency noise interference with on-site operations.

[0098] Optionally, this embodiment also has the dual-parameter criterion verified in real time by the edge computing unit. The temperature signal is integrated through a sliding window (the average value of ΔT within 10 seconds ≥ ±0.5°C), and the vibration signal is verified for sudden energy increase through the short-time Fourier transform (STFT). At the same time, the degree of bolt loosening is updated by the pre-tightening force model every 5 seconds. When the decrease range enters the 10%-20% interval, the system calls the GIS module to obtain the three-dimensional coordinates of the flange (based on the pre-stored pipeline topology map) and pushes a warning notice to the maintenance personnel's mobile terminal (installed with a dedicated App) through the MQTT protocol. The content includes coordinates, loosening percentage, and recommended inspection items. The standby monitoring device (such as a redundant optical fiber link) is switched to the high sampling rate mode (400Hz) through relay control to increase the data density to assist in decision-making.

[0099] Furthermore, after the third-level early warning is triggered, this embodiment interacts with the DCS system through the OPC UA protocol to send a closing instruction to the pressure regulating valve (response time ≤ 2 seconds) and activates the pressure relief logic: gradually reduce the pipeline pressure to a safe threshold (such as below 0.5MPa), and at the same time activate the inert gas injection system (nitrogen purity ≥ 99.99%) to dilute the leaked medium. The maintenance and repair work order is automatically generated by the work order management module. The content integrates the leakage level assessment result (matched based on the historical accident database), safety protection requirements (such as PPE level), and operation guidelines, and is pushed to the repair terminal through the RESTAPI. The superior supervision platform synchronously receives the alarm data through ModbusTCP to trigger the plant-wide broadcast and the start of the emergency plan.

[0100] Furthermore, the monitoring terminal of this embodiment is developed using the SCADA system, and the alarm indicators are dynamically rendered according to levels: yellow (RGB 255, 255, 0), orange (RGB 255, 165, 0), red (RGB 255, 0, 0), and abnormal parameters are displayed by superimposing trend curves. The diagnostic report is automatically generated in Markdown format and supports export to PDF for post-event analysis. The three-dimensional coordinates are rendered through the WebGL engine, and precise positioning of the leakage point (error < 0.5m) is achieved by combining with the pipeline BIM model. The system hardware meets the ATEX / IECEx explosion-proof certification, and the alarm terminal housing is made of 316L stainless steel (IP67 protection level). The communication link adopts a dual-redundancy design (4G + fiber optic ring network) and automatically switches in case of disconnection. The early warning logic is deployed in a fault-tolerant server cluster, and the service is ensured to be continuously available through heartbeat detection. Historical data is stored according to the ISO 14224 standard, and blockchain evidence storage is supported to meet the requirements of security audits. All operation instructions require double confirmation (operator password + electronic signature) to prevent mis-triggering.

[0101] As Figure 2 shown, this embodiment also provides a chemical pipeline flange loosening and leakage early warning system based on distributed fiber optic sensing, including:

[0102] An optical fiber winding unit for spirally winding a temperature and vibration dual-parameter distributed fiber optic sensor in the gap between flanges;

[0103] A signal preprocessing unit for collecting the temperature signal and vibration signal of the flange in real time through a distributed fiber optic demodulator, and respectively preprocessing the temperature signal and the vibration signal to obtain the real-time change value of the temperature difference between the leakage point and the environment and the vibration energy value in the 20 - 200Hz frequency band;

[0104] A criterion construction unit for establishing a multi-modal signal fusion criterion model according to the real-time change value and the vibration energy value; the multi-modal signal fusion criterion model includes a temperature criterion and a vibration criterion; the temperature criterion is that the real-time change value ≥ ±0.5°C and lasts for 10 seconds; the vibration criterion is that the vibration energy suddenly increases by more than 3dB and lasts for 10 seconds;

[0105] A loosening calculation unit for calculating the degree of bolt loosening according to the vibration signal;

[0106] An early warning unit for executing a gradient early warning mechanism according to the multi-modal signal fusion criterion model and the degree of bolt loosening.

[0107] The beneficial effects of the present invention are as follows:

[0108] (1) By deploying a distributed optical fiber sensor in the gap between flanges, the present invention can achieve real-time monitoring of temperature and vibration. This method significantly improves the monitoring ability for minor looseness and leakage. Since the distributed optical fiber can provide continuous monitoring coverage, it avoids the monitoring blind spots existing in traditional point-type devices.

[0109] (2) The present invention can detect minor changes that are difficult to capture by conventional monitoring devices, such as a temperature change below 2°C and a vibration acceleration below 0.1g. Through multi-modal signal fusion, this technology can identify the initial characteristics of flange leakage with higher sensitivity and accuracy.

[0110] (3) By establishing a fusion criterion model by combining temperature and vibration data, the present invention makes the early warning mechanism more scientific and reasonable, reducing the possibility of false alarms. The joint criterion of temperature and vibration can provide more comprehensive monitoring information and detect potential leakage risks earlier.

[0111] (4) According to different degrees of looseness and signal changes, the present invention implements hierarchical early warning (Level 1, Level 2, Level 3) to ensure that corresponding measures are taken at different risk levels. At the same time, reports are automatically generated and maintenance personnel are notified, facilitating timely response and handling and reducing accident risks.

[0112] (5) The entire system of the present invention operates in an automated manner, overcoming the problems of long cycle and poor timeliness of manual inspections, and avoiding the electromagnetic interference risks of traditional electrical sensors, meeting the explosion-proof requirements of the chemical industry.

[0113] (6) By experimentally calibrating the flange material coefficient, the present invention can correctly evaluate the degree of bolt looseness, adapt to the non-linear relationship between the pre-tightening force attenuation and leakage development of the flange, and provide an effective mathematical model for leakage prediction.

[0114] (7) The alarm information of the present invention can be synchronized to the superior supervision system, increasing the credibility of the data and the effectiveness of monitoring, and helping to improve the overall safety management level of the enterprise.

[0115] (8) Through the design of helically wound optical fiber, the overlapping coverage of each layer increases the comprehensiveness of monitoring, ensuring that the entire circumference of the flange is monitored, thereby further enhancing the effectiveness and reliability of the system.

[0116] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and reference can be made to the description in the method part for related parts.

[0117] In this article, specific examples are used to illustrate the principle and implementation of the present invention. The description of the above embodiments is only to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for early warning of flange loosening and leakage in chemical pipelines based on distributed optical fiber sensing, characterized in that, Including: Disposing the temperature and vibration dual-parameter distributed optical fiber sensor in a spiral winding manner in the gap between flanges; Collecting the temperature signal and vibration signal of the flange in real time through a distributed optical fiber demodulator, and respectively preprocessing the temperature signal and the vibration signal to obtain the real-time change value of the temperature difference between the leakage point and the environment and the vibration energy value in the 20 - 200 Hz frequency band; Establishing a multimodal signal fusion criterion model according to the real-time change value and the vibration energy value; the multimodal signal fusion criterion model includes a temperature criterion and a vibration criterion; the temperature criterion is that the real-time change value ≥ ±0.5°C and lasts for 10 seconds; the vibration criterion is that the vibration energy suddenly increases by more than 3 dB and lasts for 10 seconds; Calculating the degree of bolt loosening according to the vibration signal; Executing a gradient warning mechanism according to the multimodal signal fusion criterion model and the degree of bolt loosening.

2. The method for early warning of chemical pipeline flange loosening and leakage based on distributed optical fiber sensing according to claim 1, wherein The number of turns of the spiral winding method is 4 - 5 turns to achieve full circumferential monitoring of the flange.

3. The chemical pipeline flange loosening and leakage warning method based on distributed optical fiber sensing according to claim 1, wherein, The same pitch is maintained between each turn, and after the distributed optical fiber sensor is wound, it is fixed and protected by a thermal conductive silica gel layer with a thickness of 0.2 mm.

4. The method for early warning of chemical pipeline flange loosening and leakage based on distributed optical fiber sensing according to claim 1, wherein, The calculation formula for the degree of bolt loosening is: Wherein, F is the degree of bolt loosening, K0 is the initial stiffness eigenvalue, K t is the real-time stiffness eigenvalue, and α is the flange material coefficient; the extraction formula for the real-time stiffness eigenvalue K t is: K t = ∫f(t)·e -t / τ dt; wherein, τ is the characteristic time constant of the flange structure, and f(t) is the temperature change amount on the flange surface or around the bolt.

5. The method for early warning of chemical pipeline flange loosening and leakage based on distributed optical fiber sensing according to claim 1, wherein, The flange material coefficient α is calibrated through experiments, specifically including the fitting analysis of the elastic modulus, Poisson's ratio of the flange material and the bolt torque-preload relationship.

6. The method for early warning of chemical pipeline flange loosening and leakage based on distributed optical fiber sensing according to claim 3, wherein, The thermal conductivity of the thermal conductive silica gel layer ≥ 2.5 W / (m·K), and the thermal conductive silica gel layer is used to enhance the coupling effect between the energy of flange leakage and the temperature of the distributed optical fiber sensor laid in the flange gap.

7. The method for early warning of chemical pipeline flange loosening and leakage based on distributed optical fiber sensing according to claim 1, characterized in that Executing a gradient warning mechanism according to the multimodal signal fusion criterion model and the degree of bolt loosening, including: When only the temperature criterion or the vibration criterion is satisfied and lasts for 10 seconds, a first-level warning is triggered; When the temperature criterion and the vibration criterion are both satisfied, it lasts for 20 seconds, and the degree of bolt loosening drops by 10% - 20%, a second-level warning is triggered; When the temperature criterion and the vibration criterion are both satisfied, and it lasts for 30 seconds, and the degree of bolt loosening drops > 20%, a third-level warning is triggered and the maintenance and repair plan is automatically associated.

8. The method for early warning of chemical pipeline flange loosening and leakage based on distributed optical fiber sensing according to claim 7, wherein, After the first-level warning is triggered, the following operations are performed: a. Display the flange number, location, yellow alarm sign and the type of abnormal parameter on the operation interface of the monitoring terminal; b. Record the abnormal signal data and generate a preliminary diagnosis report; c. Send an alarm through the sound and light alarm device; After the second-level warning is triggered, the following operations are performed: a. Display the flange number, location, orange alarm sign and the percentage of bolt loosening drop on the monitoring terminal; b. Send a warning notice containing the flange number and location coordinates to the mobile terminal of the designated maintenance personnel; c. Activate the standby monitoring device to enhance the data acquisition frequency; After the third-level warning is triggered, the following operations are performed: a. Display the flange number, location, red alarm sign and the leakage level assessment result on the monitoring terminal; b. Automatically close the valve at the front end of the leaking flange; c. Send a work order containing emergency operation instructions, the number of the leaking valve, three-dimensional coordinates and safety protection requirements to the maintenance and repair terminal; d. Synchronize the alarm information to the superior supervision system through the industrial Internet of Things platform.

9. A chemical pipeline flange loosening and leakage warning system based on distributed optical fiber sensing, characterized in that, Including: An optical fiber winding unit for laying a temperature and vibration dual-parameter distributed optical fiber sensor in the gap between flanges in a spiral winding manner; A signal preprocessing unit for real-time collecting the temperature signal and vibration signal of the flange through a distributed optical fiber demodulator, and respectively preprocessing the temperature signal and the vibration signal to obtain the real-time change value of the temperature difference between the leakage point and the environment and the vibration energy value in the 20 - 200 Hz frequency band; A criterion construction unit for establishing a multi-modal signal fusion criterion model according to the real-time change value and the vibration energy value; the multi-modal signal fusion criterion model includes a temperature criterion and a vibration criterion; the temperature criterion is that the real-time change value ≥ ±0.5 °C and lasts for 10 seconds; the vibration criterion is that the vibration energy suddenly increases by more than 3 dB and lasts for 10 seconds; A loosening calculation unit for calculating the degree of bolt loosening according to the vibration signal; An early warning unit for implementing a gradient early warning mechanism according to the multi-modal signal fusion criterion model and the degree of bolt loosening.

Images