A distributed optical fiber routing design method, gas leakage monitoring method and system

Abstract

A distributed optical fiber routing design method, a gas leak monitoring method, and a system are disclosed. The design method includes: acquiring the physical and environmental parameters of the target gas-insulated metal-enclosed transmission line (GIL); establishing a GIL geometric model; and performing multi-physics coupled simulations of fluid dynamics and thermodynamics under different leakage apertures based on the GIL geometric model; using Fourier transform analysis of the coupled simulation results to determine key leakage points; and generating a distributed optical fiber routing scheme based on the key leakage points. The gas leak monitoring method includes: acquiring and processing backscattered light signals through the laid distributed optical fibers; extracting time-frequency domain features from the processed signals; and combining these features with a pre-trained pattern recognition model to determine whether a gas leak event exists. This invention solves the problems of high cost, incomplete coverage, and low positioning accuracy in existing long-distance GIL line gas leak monitoring.

Description

Technical Field

[0001] This invention relates to the field of gas monitoring, specifically to a distributed optical fiber routing design method, a gas leak monitoring method, and a system. Background Technology

[0002] Gas-Insulated Line (GIL) is widely used in power grids due to its advantages such as large transmission capacity and immunity to environmental influences. GILs are filled with high-voltage insulating gases (such as SF6 or environmentally friendly mixed gases). Leakage at the casing or flange connections directly leads to a decrease in insulation performance, easily triggering serious accidents such as breakdown discharge. Furthermore, SF6 leakage also contributes to the greenhouse effect. Therefore, timely and accurate online monitoring of GIL gas leaks is crucial for ensuring the safe operation of the power grid.

[0003] Currently, the closest existing technology to this invention is a monitoring scheme based on point sensors. This scheme mainly involves installing discrete gas concentration sensors (such as infrared or electrochemical sensors) or pressure gauges at intervals along the GIL pipe corridor or at key locations such as flange interfaces. By detecting changes in gas concentration or pressure in localized areas, it determines whether a leak has occurred.

[0004] (1) Since conventional GIL pipelines are usually several kilometers or even tens of kilometers long, if point sensors are used to achieve full coverage, the number of sensors required is extremely large, resulting in extremely high costs for sensor procurement, wiring and installation and subsequent maintenance.

[0005] (2) Since the sensing range of the point sensor is limited to the vicinity of its installation location, there is a large monitoring blind zone between the two sensors. Therefore, once a leak occurs in the blind zone, the sensor response is slow and it is impossible to obtain the accurate location and severity of the GIL gas leak.

[0006] (3) Due to the large number of point sensors in the complex pipe gallery environment, complex power supply and communication networks are required. Data transmission is easily interfered with and node failures are likely to occur. Therefore, it is difficult to stably and in real time access the sensor data to the system for visualization, and it is impossible to truly achieve 24-hour uninterrupted high-reliability continuous monitoring.

[0007] Therefore, the following six technical issues need to be addressed: (1) How to construct a comprehensive and blind-spot-free GIL gas leakage monitoring method based on distributed optical fiber; (2) How to scientifically plan the routing scheme of distributed optical fibers to accurately capture leakage signals; (3) How to design efficient distributed optical fiber data processing methods and processes to achieve accurate extraction and location of leakage characteristics; (4) How to construct a physical layer distributed optical fiber monitoring system device; (5) How to design the computer equipment to support the system; (6) How to implement a computer-readable storage medium that carries the monitoring program. Summary of the Invention

[0008] To address the problems of high cost, incomplete coverage, and low positioning accuracy in existing long-distance GIL (Gas Leakage Monitoring) lines, this invention proposes a distributed optical fiber routing design method, including: The physical and environmental parameters of the target gas-insulated metal-enclosed transmission line (GIL) are obtained, a GIL geometric model is established, and a multi-physics coupled simulation of fluid dynamics and thermodynamics is performed based on the GIL geometric model under different leakage orifice diameters. Fourier transform analysis was used to analyze the coupled simulation results to determine the key leakage points; A distributed fiber optic cabling scheme is generated based on key leakage points.

[0009] Optionally, the step of using Fourier transform analysis of the coupled simulation results to determine the key leakage points includes: Three-dimensional slice data of the sound field and temperature field under gas leakage conditions were extracted from the coupled simulation results. The power spectral density of the time-domain signal near the leakage source is calculated by analyzing the three-dimensional slice data using Fourier transform. When the energy proportion of a certain frequency band exceeds a set threshold, it is determined that the signal energy is concentrated in that frequency band. Calculate the spatial signal-to-noise ratio at each coordinate point or on the casing within the frequency band; Based on the spatial signal-to-noise ratio, determine the coordinate point or shell location that is most sensitive to changes in leakage sound / temperature, and use the most sensitive coordinate point or shell location as the critical leakage point.

[0010] Optionally, the method for generating a distributed optical fiber laying scheme based on key leakage points includes: Analysis of the critical leakage points located on the straight cylindrical section revealed that the critical leakage points are concentrated at the top and bottom of the GIL metal shell of the straight cylindrical section. The fiber optic cable laying scheme for the straight cylindrical section is as follows: using special magnetic clamps or anti-aging and insulating epoxy resin, the distributed optical fiber is tightly attached to the top and bottom of the GIL metal shell of the straight cylindrical section. The key leakage points on the flange expansion joint were analyzed. The key leakage points were concentrated on the windward / downward side of the flange interface and the folds of the bellows of the expansion joint. The fiber optic cable routing scheme for the flange expansion joint is as follows: a spiral winding method is adopted. The GIL includes: a straight cylindrical section and a flange expansion joint.

[0011] Optionally, the epoxy resin is used to fix the distributed optical fibers to the top and bottom of the GIL metal shell of the straight cylindrical section at a fixed binding point spacing.

[0012] Optionally, the spacing between the fixing points is set to a range of 1 meter to 2 meters; The spiral winding method uses a winding pitch of 5cm to 10cm.

[0013] Furthermore, this application also provides a distributed optical fiber routing determination system, comprising: The simulation module is used to obtain the physical and environmental parameters of the target gas-insulated metal-enclosed transmission line (GIL), establish the GIL geometric model, and perform fluid dynamics and thermodynamic multiphysics coupling simulations under different leakage orifice diameters based on the GIL geometric model. The leakage analysis module is used to analyze the coupled simulation results using Fourier transform to determine the key leakage points; The scheme generation module is used to generate distributed fiber optic cabling schemes based on key leakage points.

[0014] Optionally, the leakage analysis module is specifically used for: Three-dimensional slice data of the sound field and temperature field under gas leakage conditions were extracted from the coupled simulation results. The power spectral density of the time-domain signal near the leakage source is calculated by analyzing the three-dimensional slice data using Fourier transform. When the energy proportion of a certain frequency band exceeds a set threshold, it is determined that the signal energy is concentrated in that frequency band. Calculate the spatial signal-to-noise ratio for each coordinate point or shell location within the frequency band; Based on the spatial signal-to-noise ratio, determine the coordinate point or shell location that is most sensitive to changes in leakage sound / temperature, and use the most sensitive coordinate point or shell location as the critical leakage point.

[0015] Optionally, the scheme generation module is specifically used for: Analysis of the critical leakage points located on the straight cylindrical section revealed that the critical leakage points are concentrated at the top and bottom of the GIL metal shell of the straight cylindrical section. The fiber optic cable laying scheme for the straight cylindrical section is as follows: using special magnetic clamps or anti-aging and insulating epoxy resin, the distributed optical fiber is tightly attached to the top and bottom of the GIL metal shell of the straight cylindrical section. The key leakage points on the flange expansion joint were analyzed. The key leakage points were concentrated on the windward / downward side of the flange interface and the folds of the bellows of the expansion joint. The fiber optic cable routing scheme for the flange expansion joint is as follows: a spiral winding method is adopted. The GIL includes: a straight cylindrical section and a flange expansion joint.

[0016] Furthermore, this application also provides a GIL gas leak monitoring method based on distributed optical fiber, including: Based on the fiber optic cabling scheme, GIL tunnel environment and physical field simulation, the distributed fiber optic cabling and static calibration are carried out, and a nonlinear spatial mapping matrix between the distributed fiber optic cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of GIL is constructed. Probe pulses are injected into the distributed optical fiber using an optical fiber demodulator to continuously acquire backscattered light signals; then, orthogonal demodulation and adaptive noise reduction are performed on the backscattered light signals to obtain a noise-reduced signal. Time-frequency domain features are extracted from the denoised signal and input into a pre-trained pattern recognition model for feature determination to identify whether a gas leak event exists. When a gas leak event exists, fault location and display are performed based on the denoised signal and a nonlinear spatial mapping matrix. The fiber optic cable laying scheme was determined using a distributed fiber optic cable routing method as described above.

[0017] Optionally, the step of laying and statically calibrating distributed optical fibers based on the optical fiber laying scheme, GIL tunnel environment, and physical field simulation, and constructing a nonlinear spatial mapping matrix between the distributed optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of the GIL, includes: Distributed fiber optic cables are laid according to the fiber optic cabling plan. Calibration points were selected along the GIL line, and a sound source was simulated at the calibration point to conduct calibration tests and obtain the characteristic reflection peaks on the distributed optical fiber. Based on the calibration point location and the characteristic reflection peak, a nonlinear spatial mapping matrix is ​​established between the distributed optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of the GIL.

[0018] Optionally, the step of performing orthogonal demodulation and adaptive noise reduction on the backscattered light signal to obtain a denoised signal includes: The backscattered light signal is orthogonally demodulated, and the monitoring blind zone caused by signal fading is eliminated by arctangent calculation to extract the true interference phase information; An adaptive filtering algorithm is introduced to strip the background noise from the real interference phase signal, filtering out the inherent low-frequency environmental noise in the GIL tunnel, and obtaining a noise-reduced signal.

[0019] Optionally, the training of the pattern recognition model includes: Backscattered light signals are collected by distributed optical fibers laid on the surface of the GIL device. The backscattered light signals are converted into vibration time-domain signals that characterize gas leakage by distributed acoustic sensing demodulation technology. Time-frequency domain features are extracted from the vibration time-domain signals. Based on the actual gas leakage state of GIL corresponding to the signal acquisition time, a category label is set for the time-frequency domain features, and a sample is constructed from the time-frequency domain features and the corresponding category label; A sample set is constructed from multiple samples; Based on the sample set, the hyperparameters of the support vector machine are optimized by grid search and k-fold cross-validation to complete the training of the support vector machine classifier and obtain the pattern recognition model. The category labels include: gas leak and normal noise.

[0020] Optionally, the step of injecting probe pulses into the distributed optical fiber using an optical fiber demodulator and continuously acquiring backscattered light signals includes: The fiber optic demodulator continuously emits ultra-narrow linewidth pulsed lasers into the distributed optical fiber, and a high-sensitivity photodetector receives backscattered Rayleigh or Brillouin scattered light signals carrying information about the external sound field, micro-vibration and temperature transients; and forms a two-dimensional optical signal matrix containing spatial and temporal domains. Among them, the backscattered light signal includes: back Rayleigh scattering light and Brillouin scattering light signal.

[0021] Optionally, the step of performing fault location and display based on noise-reduced signals and nonlinear spatial mapping matrices when a gas leak event occurs includes: When a gas leak occurs, the specific fiber length location where the leak occurred is calculated based on the optical pulse time delay when the abnormal signal occurs in the noise-reduced signal. The absolute physical coordinates of the leak are displayed intuitively based on the specific fiber length location where the leak occurred, combined with a nonlinear spatial mapping matrix.

[0022] Furthermore, this application also provides a GIL gas leak monitoring system based on distributed optical fiber, comprising: The simulation analysis module is used to lay and statically calibrate distributed optical fibers based on the optical fiber laying and routing scheme, GIL tunnel environment and physical field simulation, and to construct a nonlinear spatial mapping matrix between the distributed optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of GIL. The signal acquisition module is used to inject probe pulses into the distributed optical fiber through an optical fiber demodulator and continuously acquire backscattered light signals. The data processing module is used to perform orthogonal demodulation and adaptive noise reduction on the backscattered light signal to obtain a noise-reduced signal; The leak detection module is used to extract time-frequency domain features from the noise-reduced signal and input the time-frequency domain features into the pattern recognition model for feature judgment to determine whether a gas leak event exists. A visualization platform for locating and displaying faults based on denoised signals and nonlinear spatial mapping matrices when a gas leak occurs; The fiber optic cable laying scheme is determined using a distributed fiber optic cable routing determination method as described in any one of claims 1-5.

[0023] Optionally, the simulation analysis module is specifically used for: Calibration points were selected along the GIL line, and a sound source was simulated at the calibration point to conduct calibration tests and obtain the characteristic reflection peaks on the optical fiber. Based on the calibration point location and the characteristic reflection peak, a nonlinear spatial mapping matrix is ​​established between the distributed optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of the GIL.

[0024] Optionally, the data processing module is specifically used for: The backscattered light signal is orthogonally demodulated, and the monitoring blind zone caused by signal fading is eliminated by arctangent calculation to extract the true interference phase information; An adaptive filtering algorithm is introduced to strip the background noise from the real interference phase signal, filtering out the inherent low-frequency environmental noise in the GIL tunnel, and obtaining a noise-reduced signal.

[0025] Optionally, it also includes: a model training module for: Backscattered light signals are collected by fiber optic sensors laid on the surface of the GIL device. The backscattered light signals are converted into vibration time-domain signals that characterize gas leakage by distributed acoustic sensing demodulation technology. Time-frequency domain features are extracted from the vibration time-domain signals. Based on the actual gas leakage state of GIL corresponding to the signal acquisition time, a category label is set for the time-frequency domain features, and a sample is constructed from the time-frequency domain features and the corresponding category label; A sample set is constructed from multiple samples; Based on the sample set, the hyperparameters of the support vector machine are optimized by grid search and k-fold cross-validation to complete the training of the support vector machine classifier and obtain the pattern recognition model. The category labels include: gas leak and normal noise.

[0026] Optionally, the signal acquisition module is specifically used for: The fiber optic demodulator continuously emits ultra-narrow linewidth pulsed lasers into the distributed optical fiber, and a high-sensitivity photodetector receives backscattered Rayleigh or Brillouin scattered light signals carrying information about the external sound field, micro-vibration and temperature transients; and forms a two-dimensional optical signal matrix containing spatial and temporal domains. Among them, the backscattered light signal includes: back Rayleigh scattering light and Brillouin scattering light signal.

[0027] Optionally, the leakage detection module is specifically used for: Extracting time-frequency domain features from denoised signals; The time-frequency domain features are input into the pattern recognition model for feature determination to obtain category labels; When the category label is "gas leak," a gas leak event has occurred. When the category label is normal noise, there is no gas leak event.

[0028] Optionally, the visualization platform is specifically used for: When a gas leak occurs, the specific fiber length location where the leak occurred is calculated based on the optical pulse time delay when the abnormal signal occurs in the noise-reduced signal. The absolute physical coordinates of the leak are displayed intuitively based on the specific fiber length location where the leak occurred, combined with a nonlinear spatial mapping matrix.

[0029] In another aspect, this application also provides an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, a distributed optical fiber routing design method or a GIL gas leak monitoring method based on distributed optical fiber is implemented as described above.

[0030] In another aspect, this application also provides a computer-readable storage medium having an executable program stored thereon, which, when executed, implements a distributed optical fiber routing design method or a GIL gas leakage monitoring method based on distributed optical fiber as described above.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention provides a distributed optical fiber routing design method, comprising: acquiring the physical and environmental parameters of the target gas leak detection line (GIL); establishing a GIL geometric model; and performing multiphysics coupled simulations of fluid dynamics and thermodynamics under different leakage apertures based on the GIL geometric model; using Fourier transform analysis of the coupled simulation results to determine key leakage points; and generating a distributed optical fiber laying routing scheme based on the key leakage points. This invention solves the problems of high cost, incomplete coverage, and low positioning accuracy in existing long-distance GIL line gas leak monitoring.

[0032] This invention also provides a GIL gas leak monitoring method based on distributed optical fiber, comprising: laying and statically calibrating distributed optical fibers according to the optical fiber laying scheme, GIL tunnel environment, and physical field simulation; constructing a nonlinear spatial mapping matrix between the distributed optical fiber laying length L and the three-dimensional spatial coordinates (X, Y, Z) of the GIL; injecting probe pulses into the distributed optical fiber through an optical fiber demodulator to continuously acquire backscattered light signals; performing orthogonal demodulation and adaptive noise reduction on the backscattered light signals to obtain a denoised signal; extracting time-frequency domain features from the denoised signal and inputting the time-frequency domain features into a pattern recognition model for feature judgment to determine whether a gas leak event exists; and when a gas leak event exists, performing fault location and display based on the denoised signal and the nonlinear spatial mapping matrix. This achieves low-cost, full-coverage, blind-spot-free, high-precision positioning and 24-hour uninterrupted continuous visual monitoring of long-distance GIL pipelines. Attached Figure Description

[0033] Figure 1 This is a flowchart of a distributed optical fiber routing design method according to the present invention; Figure 2 This is a flowchart of the distributed optical fiber routing design method of the present invention; Figure 3 This is a schematic diagram of the distributed optical fiber laying location of the present invention; Figure 4 This is a schematic diagram of the distributed optical fiber structure of the present invention; Figure 5 A flowchart illustrating a GIL gas leak monitoring method based on distributed optical fiber, which is a specific embodiment of the present invention. Figure 6 This is a flowchart of the distributed optical fiber data processing of the present invention; Figure 7 This is a schematic diagram of the SF6 gas diffusion process of the present invention; Figure 8 This is a schematic diagram illustrating the pattern recognition and classification principle of the Support Vector Machine (SVM) based on the RBF kernel function of the present invention. Figure 9 This is a schematic diagram of an electronic device structure according to the present invention; Among them, 1-distributed optical fiber, 2-straight cylindrical section, 3-flange expansion joint, 11-outer sheath, 12-armor layer, 13-coating layer, 14-single-mode optical fiber core. Detailed Implementation

[0034] This invention proposes a distributed optical fiber routing design method, a gas leak monitoring method and system, which solves the problems of high cost, blind spots, inaccurate positioning and difficulty in real-time visualization of long-distance GIL pipeline leak monitoring. It achieves low-cost, full-coverage, blind-spot-free, high-precision positioning and 24-hour uninterrupted continuous visualization monitoring of long-distance GIL pipelines.

[0035] This invention proposes a comprehensive non-contact gas leak monitoring framework for long-distance GIL pipelines, comprising six key components: a distributed optical fiber routing design method (layout and spatial layout at the physical sensing level); a device design including sensing optical fibers and a demodulator (the hardware foundation at the physical level); a GIL gas leak monitoring method based on distributed optical fibers (the overall monitoring framework); a distributed optical fiber data processing method and workflow (the core algorithm for converting optical signals into leak alarms); a computer device design for executing the monitoring logic (edge ​​computing and central control); and a media design for storing the program.

[0036] To better understand the present invention, the following description, in conjunction with the accompanying drawings and embodiments, will further illustrate the content of the present invention.

[0037] Example 1 A distributed optical fiber routing design method, such as Figure 1 As shown, it includes: Step S1: Obtain the physical and environmental parameters of the target gas-insulated metal-enclosed transmission line (GIL), establish a GIL geometric model, and perform fluid dynamics and thermodynamic multiphysics coupling simulations under different leakage orifice diameters based on the GIL geometric model. Step S2: Use Fourier transform to analyze the coupled simulation results and determine the key leakage points; Step S3: Generate a distributed fiber optic cabling scheme based on the key leakage points.

[0038] The following is combined Figure 2 A detailed description of the distributed optical fiber routing design method of the present invention: To ensure that the optical fiber can capture the weakest leaked signals to the greatest extent, the following scientific routing design is adopted: Step S1: Obtain the physical and environmental parameters of the target gas-insulated metal-enclosed transmission line (GIL), establish a GIL geometric model, and perform fluid dynamics and thermodynamic multiphysics coupling simulations under different leakage orifice diameters based on the GIL geometric model. This step specifically includes: One-way transient fluid-structure interaction analysis was performed using ANSYS Fluent and the Transient Structural module to simulate the mechanical impact of the leakage process on the flange and surrounding pipeline structure.

[0039] Specifically, transient solutions were performed under leakage conditions with different orifice diameters (0.5 mm to 5 mm in this embodiment) to simulate the sound field (aerodynamic noise) distribution and temperature response characteristics during the leakage of high-pressure SF6 gas inside the GIL. Through fluid-structure interaction, the fluid boundary pressure was mapped to the structural domain to analyze the mechanical impact and vibration effects of the leaking gas flow on the flange expansion stage and the straight cylinder section 2 (i.e., the flange body and surrounding pipeline structure).

[0040] Simulation: Modeling and calculation of high-pressure SF6 gas leakage inside GIL.

[0041] 1) Physical mechanism of vibration caused by gas leakage To accurately obtain leakage characteristics to support the multiphysics coupling simulation and wiring design in this invention, it is necessary to deeply analyze the underlying physical mechanism of gas leakage in GIL equipment, such as... Figure 7 As shown. Specifically, gas leakage is the core external factor causing abnormal vibration in GIL equipment. The vibration generation mechanism is closely related to the dynamic characteristics of the leaking gas flow and the coupling with the equipment structure. GIL equipment uses SF6 gas as the primary insulating medium, and its leakage process involves turbulent flow of high-pressure gas ejected into a low-pressure environment. The high-speed jet formed at the leak point generates strong broadband aerodynamic noise. This acoustic energy is transmitted through the gas medium to the equipment casing and conductors, exciting high-frequency vibrations in the structure. Simultaneously, the SF6 gas maintains a high pressure of 0.4~0.5 MPa within the pipeline. The significant internal and external pressure difference drives the airflow to form a stable jet. The impact of the jet on the pipeline wall, insulators, and other components induces forced vibrations in the local structure. Furthermore, as a heavy gas, SF6 undergoes four stages after leakage: initial high-speed ejection, gravity settling, air dilution, and diffusion distribution. During the settling process, the secondary impact of the airflow on the bottom of the pipeline further exacerbates the local vibration signal at the bottom.

[0042] 2) Physical model assumptions and boundary condition settings This simulation aims to perform unidirectional transient fluid-structure interaction analysis using ANSYS Fluent and the Transient Structural module to simulate the mechanical impact of leakage on the flange and surrounding piping structure, providing a theoretical basis for the deployment of distributed fiber optic cable 1. The ANSYS Fluent module is referred to as the Fluent module, and the Transient Structural module as the Structural module. The model is built based on the following basic assumptions and settings: Materials and conditions: The pipe is considered to be a linear elastic material and does not undergo plastic deformation; the leak hole size is fixed to simulate steady-state leakage conditions; the high-pressure SF6 gas is considered to be a compressible viscous Newtonian fluid.

[0043] Flow field simplification: The leakage process is treated as adiabatic expansion, ignoring heat exchange with the outside environment. In the near-field region of the leak hole (jet core region), gas momentum is absolutely dominant, and the influence of gravity is ignored to simplify the flow field solution; the details of the flange bolts are simplified to focus on the local fluid-structure interaction response of the leak hole.

[0044] Computational model selection: Transient analysis was adopted and the energy equation considering compressibility effects was enabled to accurately decouple the temperature field; the k-omega SST (turbulent kinetic energy-specific dissipation rate shear stress transport turbulence model) was selected to accurately simulate the strong shear flow at the leak hole; the broadband aerodynamic noise model was activated to obtain the spatial distribution of the sound source; the component transport model was used to define the mixing and diffusion process of SF6 and air.

[0045] 3) Mesh generation and coupled solution strategy Meshing strategy: An unstructured mesh and local refinement strategy are adopted. The overall mesh size of the fluid domain is set to 0.2m, and extreme refinement (down to 0.002m) is performed in the leakage hole and jet path areas to accurately capture extremely high pressure gradients and Mach number changes; the flange and surrounding area of ​​the structural domain are also refined to 0.002m to ensure accurate stress gradient calculation.

[0046] Material mapping and fluid-structure interaction: The state is initialized using the initial volume fraction tool (Patch), setting the pipe interior to a state of 340K, 0.45MPa, and SF6 volume fraction of 1. A unidirectional fluid-structure interaction scheme is employed. After the Fluent module calculates the flow field pressure and temperature response, the pressure load is transferred to the Structural module through the fluid-structure interaction interface between the inner and outer walls to calculate structural strain and high-frequency vibration response. The driving force for system evolution originates entirely from the initial internal and external pressure difference.

[0047] 4) Theoretical calculation of leakage rate (blockage flow determination) Before performing fluid simulation, the leakage rate under different orifice diameters needs to be theoretically estimated. Taking a GIL pipe with a total length of 36m, an inner diameter of 880mm (radius of 0.44m), and filled with SF6 gas as an example, the gas physical parameters required for calculation are shown in Table 1.

[0048] Table 1. Calculation Parameters

[0049] The formula for determining the critical pressure ratio for flow state is: (2) Substituting γ=1.09, the critical pressure ratio is calculated to be approximately 0.587. Since the actual pressure ratio under the current operating conditions is... Since 0.224 < 0.587, the leakage process is determined to be sonic flow (choked flow). Based on sonic flow theory, the mass leakage rate Q... m The calculation formula is: (2) Among them, C d Let A be the emission coefficient (0.6 for sharp-edged orifices), and A be the cross-sectional area of ​​the leakage orifice. Based on the above formula, the theoretical estimated leakage rates for different orifice diameters are calculated as shown in Table 2, serving as benchmark data for simulation convergence judgment and quantitative fiber optic alarm.

[0050] Table 2. Leakage rate predictions for different orifice sizes

[0051] 5) Spatial Layout and Guiding Significance of Monitoring Points Through fluid-structure interaction analysis, the acoustic radiation mechanism and spatial distribution law of the sharp temperature drop (Joule-Thomson effect) in high-pressure SF6 gas leakage were clearly revealed. Structural vibration response is the core physical quantity for distributed fiber optic monitoring. Simulation results not only identified the primary sensitive region with the strongest temperature response and vibration signal, but also established a mapping relationship between "leakage aperture—mass leakage rate—vibration stress amplitude." This provides rigorous quantitative data support for fiber selection (such as high acoustic sensitivity single-mode armored fiber) and routing strategies (such as tight winding at flanges and straight laying in the cylinder section), ultimately ensuring that the technical solution of this invention can achieve accurate extraction and location determination of leakage signals.

[0052] Multiphysics Coupled Simulation: Unidirectional transient fluid-structure interaction analysis was performed using ANSYS Fluent and the Transient Structural module. In the flow field calculation, a Species Transport model was used to define the mixing and diffusion process of SF6 and external air. Considering the strong shear flow characteristics at the leak orifice, the k-omega SST turbulence model was selected, and Broadband Noise Source Models were activated to obtain the spatial distribution of the sound source. Simultaneously, the energy equation considering compressibility effects was enabled to accurately solve the temperature field distribution during the adiabatic expansion of the gas. Based on this, leakage conditions with orifice diameters ranging from 0.5 mm to 5 mm were set up for transient solutions to simulate the sound field (aerodynamic noise) distribution and temperature response characteristics during the leakage of high-pressure SF6 gas inside the GIL. Furthermore, through fluid-structure interaction, the fluid boundary pressure was mapped to the structural domain to analyze the mechanical impact and vibration effects of the leaking gas flow on the flange body and surrounding pipeline structure, providing data support for subsequent calibration of key monitoring points.

[0053] Step S2: Use Fourier transform to analyze the coupled simulation results and determine the key leakage points, including: Three-dimensional slice data of the sound field and temperature field under gas leakage conditions were extracted from the coupled simulation results. The power spectral density of the time-domain signal near the leakage source is calculated by analyzing the three-dimensional slice data using Fourier transform. When the energy proportion of a certain frequency band exceeds a set threshold, it is determined that the signal energy is concentrated in that frequency band. Calculate the spatial signal-to-noise ratio for each coordinate point / shell location within the frequency band; The coordinates or shell location most sensitive to changes in leakage sound / temperature are determined based on the spatial signal-to-noise ratio. Specifically, the coordinate point or shell location with the highest spatial signal-to-noise ratio is taken as the coordinate point or shell location most sensitive to changes in leakage sound / temperature, which is the critical leakage point.

[0054] Step S2 includes the following specific implementation steps: Sensitive area extraction: Three-dimensional slice data of the sound field and temperature field were extracted through simulation, confirming that the acoustic emission signals generated by gas leakage are mainly concentrated in the high-frequency range (e.g., 20kHz-80kHz). It was also confirmed that because SF6 gas is denser than air, it tends to deposit downwards after leakage; and that the sound field energy resides longest and attenuates least at the windward / downwind side of the flange interface and at specific folds in the bellows of the expansion joint. These locations were marked as primary sensitive areas.

[0055] Spatial signal-to-noise ratio (SNR) is an indicator used to quantify the "response capability of a specific location to a leaking signal." Essentially, it is the spatial contrast between the useful signal energy and the background noise energy at that location within the frequency band of the leaking signal.

[0056] Step S3: Differentiated fiber optic cabling schemes based on key leakage points, including: Analysis of the critical leakage points located on the straight cylindrical section 2 revealed that the critical leakage points are concentrated at the top and bottom of the GIL metal shell of the straight cylindrical section 2. The fiber optic cable routing scheme for the straight cylindrical section is as follows: use special magnetic clamps or anti-aging and insulating epoxy resin to tightly attach the fiber optic cable to the top and bottom of the GIL metal shell of the straight cylindrical section 2. The key leakage points on flange expansion section 3 were analyzed. The key leakage points were concentrated on the windward / downwind side of the flange interface and the folds of the bellows of the expansion section. The fiber optic cable routing scheme for flange expansion joint 3 is as follows: a spiral winding method is adopted. The GIL includes: a straight cylindrical section 2 and a flange expansion section 3.

[0057] Furthermore, the epoxy resin is used to fix the distributed optical fiber 1 to the top and bottom of the straight cylindrical section 2GIL metal shell according to the fixed binding point spacing.

[0058] Furthermore, the spacing between the fixing points is set to a range of 1 meter to 2 meters; The spiral winding method uses a winding pitch of 5cm to 10cm.

[0059] The specific implementation steps of S3 include: differentiated laying processes, such as... Figure 3 As shown.

[0060] Linear cylindrical section 2: Axially linear installation is adopted. Using specialized magnetic clamps or anti-aging, insulating epoxy resin, the distributed optical fiber 1 is tightly attached to the top (0° position, monitoring for light gas buoyancy or thermal anomalies) and bottom (180° position, focusing on monitoring the deposition of heavy gases such as SF6) of the GIL metal casing. The spacing between the fixing points is strictly controlled between 1 and 2 meters to ensure efficient acoustic coupling while preventing accidental breakage of the distributed optical fiber 1 due to thermal expansion and contraction.

[0061] Flange and Expansion Joint (hereinafter referred to as Flange Expansion Joint) 3: Since this is a high-risk area for leakage due to aging of the sealing ring, a spiral winding method is adopted. The winding pitch is set to 5cm to 10cm to significantly increase the optical fiber spatial resolution and acoustic sensitivity in this area. When transitioning from the straight section to the spiral section, the bending radius of the optical fiber is controlled to always be greater than 50mm to prevent macro-bending loss from affecting the signal-to-noise ratio of the downstream optical signal.

[0062] The hardware device involved in the distributed optical fiber routing design method provided by this invention includes: Fiber optic demodulator main unit: The system's light source and detection core. It incorporates an ultra-narrow linewidth continuous laser (linewidth strictly controlled at <3kHz, center wavelength 1550nm C-band to reduce optical attenuation), an acousto-optic modulator (AOM, used to modulate continuous light into probe pulses with a pulse width of 10ns-100ns, the pulse width determining the system's spatial resolution), an erbium-doped fiber amplifier (EDFA, used to increase injected light power), a circulator, and a high-speed data acquisition card (DAQ) with a sampling rate of no less than 250MS / s and a balanced photodetector (BPD).

[0063] Distributed Fiber 1: Employs single-mode armored special optical fiber with high acoustic sensitivity and tensile strength. For example... Figure 4 As shown, the left side is a longitudinal sectional view, and the right side is a cross-sectional view. The outer sheath 11 of the optical fiber is made of flame-retardant and anti-aging polyurethane or low-smoke halogen-free material, which has a protective function, prevents damage from construction machinery, and has high tensile strength. Inside, an extremely fine stainless steel flexible armor layer 12 is added outside the fiber core coating layer 13. The single-mode optical fiber core 14 is protected by the outer sheath 11, the inner armor layer 12, and the inner lining / optical fiber coating layer 13, which can effectively prevent damage from construction machinery and has high tensile strength. This structural design can prevent the fiber core from breaking under lateral pressure during complex on-site construction and laying (compression and tensile strength), and can also ensure efficient mechanical transmission and acoustic coupling efficiency of high-frequency ultrasound through the high Young's modulus of the metal sheath. This distributed optical fiber 1 is a type of optical fiber specifically designed for measuring physical / chemical quantities (temperature, strain, pressure, vibration, gas concentration, etc.). It makes the optical fiber itself a "sensor" and has the dual functions of optical path transmission and physical quantity sensing.

[0064] Example 2 Furthermore, this application also provides a distributed optical fiber routing determination system, comprising: The simulation module is used to obtain the physical and environmental parameters of the target gas-insulated metal-enclosed transmission line (GIL), establish the GIL geometric model, and perform fluid dynamics and thermodynamic multiphysics coupling simulations under different leakage orifice diameters based on the GIL geometric model. The leakage analysis module is used to analyze the coupled simulation results using Fourier transform to determine the key leakage points; The scheme generation module is used to generate distributed fiber optic cabling schemes based on key leakage points.

[0065] Optionally, the leakage analysis module is specifically used for: Three-dimensional slice data of the sound field and temperature field under gas leakage conditions were extracted from the coupled simulation results. The power spectral density of the time-domain signal near the leakage source is calculated by analyzing the three-dimensional slice data using Fourier transform. When the energy proportion of a certain frequency band exceeds a set threshold, it is determined that the signal energy is concentrated in that frequency band. Calculate the spatial signal-to-noise ratio for each coordinate point or shell location within the frequency band; Based on the spatial signal-to-noise ratio, determine the coordinate point or shell location that is most sensitive to changes in leakage sound / temperature, and use the most sensitive coordinate point or shell location as the critical leakage point.

[0066] Optionally, the scheme generation module is specifically used for: Analysis of the critical leakage points located on the straight cylindrical section 2 revealed that the critical leakage points are concentrated at the top and bottom of the GIL metal shell of the straight cylindrical section 2. The fiber optic cable routing scheme for the straight cylindrical section 2 is as follows: using special magnetic clamps or anti-aging and insulating epoxy resin, the distributed optical fiber 1 is tightly attached to the top and bottom of the GIL metal shell of the straight cylindrical section 2. The key leakage points on flange expansion section 3 were analyzed. The key leakage points were concentrated on the windward / downwind side of the flange interface and the folds of the bellows of the expansion section. The fiber optic cable routing scheme for flange expansion joint 3 is as follows: a spiral winding method is adopted. The GIL includes: a straight cylindrical section 2 and a flange expansion section 3.

[0067] Example 3 A method for monitoring GIL gas leaks based on distributed optical fibers, comprising: Step 1: Based on the fiber optic cable laying scheme, GIL tunnel environment and physical field simulation, lay and statically calibrate distributed fiber optic 1, and construct a nonlinear spatial mapping matrix between the cable length L of distributed fiber optic 1 and the three-dimensional spatial coordinates (X, Y, Z) of GIL. Step 2: Inject probe pulses into distributed fiber 1 using an optical fiber demodulator to continuously acquire backscattered light signals; and perform orthogonal demodulation and adaptive noise reduction on the backscattered light signals to obtain a noise-reduced signal; Step 3: Extract time-frequency domain features from the denoised signal and input the time-frequency domain features into a pre-trained pattern recognition model for feature judgment to determine whether a gas leak event exists; and when a gas leak event exists, perform fault location and display based on the denoised signal and nonlinear spatial mapping matrix. The fiber optic cable laying scheme was determined using a distributed fiber optic cable routing method as described above.

[0068] The following is combined Figure 5 and Figure 6The various steps of the GIL gas leak monitoring method based on distributed optical fiber provided by this invention are further described below.

[0069] Step 1: Based on the fiber optic cabling scheme, GIL tunnel environment, and physical field simulation, perform the laying and static calibration of distributed fiber optic cable 1, and construct a nonlinear spatial mapping matrix between the cabling length L of distributed fiber optic cable 1 and the three-dimensional spatial coordinates (X, Y, Z) of the GIL, including: The distributed optical fiber 1 was laid according to the optical fiber laying and routing plan. Calibration points were selected along the GIL line, and a sound source was simulated at the calibration point to conduct calibration tests and obtain the characteristic reflection peaks on distributed optical fiber 1. Based on the calibration point location and the characteristic reflection peak, a nonlinear spatial mapping matrix is ​​established between the distributed optical fiber 1 cabling length L and the GIL three-dimensional spatial coordinates (X, Y, Z).

[0070] The specific implementation steps for step 1 are as follows: Static calibration and installation: Based on the physical field simulation results, sensing optical fiber 1 is installed on the GIL shell and flange using a specific process. After installation, manual tapping or simulated sound source calibration tests are performed along the GIL to obtain characteristic reflection peaks on the optical fiber. This establishes a nonlinear spatial mapping matrix between the optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of the GIL, eliminating positioning errors caused by excess fiber length and coiling.

[0071] Step 2: Inject probe pulses into distributed fiber 1 using an optical fiber demodulator to continuously acquire backscattered light signals; and perform orthogonal demodulation and adaptive noise reduction on the backscattered light signals to obtain a noise-reduced signal, including: Probe pulses are injected into distributed optical fiber 1 using an optical fiber demodulator to continuously acquire backscattered light signals; The backscattered light signal is orthogonally demodulated, and the monitoring blind zone caused by signal fading is eliminated by arctangent calculation, thereby extracting the true interference phase information; An adaptive filtering algorithm is introduced to strip the background noise from the interference phase signal, filtering out the inherent low-frequency environmental noise in the GIL tunnel, and obtaining a noise-reduced signal.

[0072] The specific implementation steps for step 2 are as follows: Optical signal pulse injection and continuous acquisition: Ultra-narrow linewidth pulsed laser light is continuously emitted into the optical fiber through an optical fiber demodulator. As the laser light propagates within the fiber core, it is scattered. A high-sensitivity photodetector receives the backscattered Rayleigh or Brillouin scattered light signals carrying information about the external sound field, micro-vibrations, and temperature transients. This process continues, forming a two-dimensional optical signal matrix encompassing both spatial (distance) and temporal (pulse period) domains.

[0073] Dynamic demodulation and noise reduction: The received backscattered light is orthogonally demodulated (I / Q demodulation). The monitoring blind zone caused by signal fading is eliminated by arctangent calculation, and the true interference phase information is extracted. Subsequently, an adaptive filtering algorithm is introduced to strip the background noise from the phase signal, filtering out the low-frequency environmental noise inherent in the GIL tunnel (such as vibration from large wind turbines, normal power frequency electromagnetic vibration, and geological disturbances from vehicle traffic).

[0074] The phase demodulation, preprocessing, and noise reduction in step 2 will be further described below: Phase demodulation: Phase-sensitive optical time-domain reflectometry (φ-OTDR) based on coherent detection is employed. A 90° optical mixer at the front end of the demodulator splits the beat frequency signal into mutually orthogonal I and Q paths. By calculating the arctangent values ​​of I(t) and Q(t) and performing phase unwrapping, the original phase φ(t), which is linearly related to the external acoustic field stress, is restored.

[0075] Preprocessing and noise reduction: To address the complex non-stationary background noise within the pipe gallery, a wavelet soft thresholding denoising algorithm is employed. The original phase signal is decomposed into 4 to 5 layers using sym8 or db4 wavelet basis functions. Low-frequency components (typically <1kHz) mainly caused by mechanical vibration are filtered out by setting an adaptive threshold, while the high-frequency leakage acoustic emission signal (>10kHz) containing gas turbulence characteristics is reconstructed and retained.

[0076] Before step 3, the training of the pattern recognition model is also included, and this training process specifically includes: Backscattered light signals are collected by distributed optical fiber 1 laid on the surface of the GIL device. The backscattered light signals are converted into vibration time-domain signals characterizing gas leakage by distributed acoustic sensing demodulation technology. Time-frequency domain features are extracted from the vibration time-domain signals. Based on the actual gas leakage state of GIL corresponding to the signal acquisition time, a category label is set for the time-frequency domain features, and a sample is constructed from the time-frequency domain features and the corresponding category label; A sample set is constructed from multiple samples; Based on the sample set, the hyperparameters of the support vector machine are optimized by grid search and k-fold cross-validation to complete the training of the support vector machine classifier and obtain the pattern recognition model. The category labels include: gas leak and normal noise; The support vector machine classifier is a support vector machine classifier based on radial basis kernel function.

[0077] The following is combined Figure 8 The algorithm used in the pattern recognition model and the specific recognition process will be further introduced: Core Algorithm: Support Vector Machine (SVM) Mathematical Model and Principle Formula.

[0078] To achieve accurate identification of complex acoustic signals within the GIL (Gas Inlet and Outer Limit) tunnel, this invention employs a Support Vector Machine (SVM) algorithm based on the Radial Basis Function (RBF) kernel in step S34. Its core mathematical principles and formula derivation are as follows: 1) Sample definition and hyperplane construction Assume the acoustic feature sample set acquired and extracted via optical fiber is .in, The d-dimensional feature vector of the i-th sample (containing features extracted such as short-time energy, zero-crossing rate, and spectral centroid); Let n be the corresponding category label (e.g., 1 represents real gas leak, -1 represents normal environmental noise or human-caused knocking), and n be the total number of samples. The core idea of ​​SVM is to find an optimal hyperplane that can separate samples of different categories. The equation of this hyperplane can be expressed as:

[0079] in, is the normal vector of the hyperplane, which determines the direction of the hyperplane; b is the displacement term; The feature vector of the sample.

[0080] 2) Soft Spacing and Optimization Objectives Because acoustic features acquired in-situ by GIL often contain unavoidable noise and outliers, the samples are not perfectly linearly separable in the feature space. To address this, this invention introduces the concept of a "soft margin," allowing for individual sample misclassification, and introduces slack variables... Given that ≥ and the penalty parameter C>0, the following optimization objective function is constructed:

[0081] The constraints are:

[0082] The objective function aims to maximize the classification margin (i.e., minimize the classification margin). At the same time, we should try to minimize the loss caused by misclassification. The penalty parameter C is used to adjust the tolerance for misclassified samples.

[0083] To address the highly nonlinear characteristics of acoustic emission signals from gas leaks, this invention introduces a kernel function mechanism to implicitly map low-dimensional input features x into a high-dimensional feature space φ(x), making it linearly separable in the high-dimensional space. This invention preferably uses the Radial Basis Function (RBF) as the kernel function for SVM, and its mathematical expression is as follows:

[0084] in, This is the kernel function for SVM. Let be the feature vector of the j-th sample, and γ>0 be the kernel function parameter, which controls the distribution complexity of samples in high-dimensional space and the nonlinear mapping capability of the model. Compared with linear kernels or polynomial kernels, the RBF kernel function has fewer parameters and can handle complex nonlinear boundaries, greatly improving the accuracy of identifying small leakage signals.

[0085] 3) Final decision function Solve the dual problem of the above optimization problem using the Lagrange multiplier method to obtain the optimal Lagrange multipliers. and bias Finally, the classification decision function used to determine whether the newly acquired fiber acoustic feature x is a leakage signal is:

[0086] When the calculation result When the value is greater than 0, the system determines that the current feature belongs to the gas leak mode and triggers an alarm; otherwise, it is determined to be non-leaking background noise. In the given classification decision function, This represents the class label of the i-th training sample.

[0087] Step 3: Extract time-frequency domain features from the denoised signal and input these features into a pre-trained pattern recognition model for feature determination to identify whether a gas leak event exists. If a gas leak event is found, fault location and display are performed based on the denoised signal and the nonlinear spatial mapping matrix. Extract time-frequency domain features from the denoised signal, and input the time-frequency domain features into a pre-trained pattern recognition model for feature determination to obtain category labels; When the category label is "gas leak," a gas leak event has occurred. When the category label is normal noise, there is no gas leak event; When a gas leak occurs, the specific fiber length location where the leak occurred is calculated based on the optical pulse time delay when the abnormal signal occurs in the noise-reduced signal. The absolute physical coordinates of the leak are displayed intuitively based on the specific fiber length location where the leak occurred, combined with a nonlinear spatial mapping matrix.

[0088] Step 3 includes the following specific implementation steps: Feature extraction and leakage detection: The denoised signal is continuously sampled using a sliding window to extract the time-frequency domain feature vector in real time. When the system detects a continuous broadband high-frequency acoustic emission signal (reflecting the ultrasonic turbulence generated when high-pressure gas is ejected at high speed from a tiny gap), or when a local frequency drift and temperature drop occur (reflecting the Joule-Thomson effect caused by the adiabatic expansion of the gas), the system automatically crosses the detection threshold and triggers a leakage alarm mechanism.

[0089] Precise location and 3D visualization: Based on the time delay of the optical pulse when the abnormal signal occurs, the specific fiber length location where the leak occurred is calculated. Combining the spatial mapping matrix established in the above steps, the absolute physical coordinates of the leak are intuitively displayed in the 3D view of the background digital twin system in the form of highlighted red dots or wavy lines, and the leakage level (e.g., slight leakage, moderate leakage, severe rupture) is output based on the integral of the acoustic wave energy.

[0090] Further details are provided regarding feature extraction and pattern recognition classification in step 3: Feature extraction: The denoised signal is windowed and framed (e.g., using a Hamming window, frame length 20ms, frame shift 10ms), and the short-time energy (E), zero-crossing rate (ZCR), and spectral centroid of each frame are calculated. When a local high-pressure gas leak occurs, the friction between the jet and the pipe wall will produce typical broadband white noise characteristics, which in the algorithm manifests as a continuous surge in short-time energy over several frames, a significant increase in the zero-crossing rate, and an overall shift of the spectral centroid towards the high-frequency band.

[0091] Pattern recognition and classification: The joint feature vectors mentioned above are extracted and input into a Support Vector Machine (SVM) classifier pre-trained using a large number of field-collected samples and artificially simulated samples. The SVM employs a radial basis function (RBF) nonlinear kernel function to automatically distinguish different acoustic modes such as "real gas micro-leakage," "human inspection tool tapping," "pipe gallery ventilation fan start / stop," and "normal background noise" by finding the optimal hyperplane. Combined with a multi-frame smoothing voting mechanism, the false alarm rate caused by the single threshold method is significantly reduced.

[0092] Leakage point location algorithm: Once the SVM classifier outputs a "gas leak" mode, the system immediately extracts the peak arrival time of the light signal that triggered the alarm, and calculates the physical length d of the leak point from the demodulator based on the time-of-flight difference of the light pulses. The core location formula is: (1) Where c is the speed of light in a vacuum. To detect the time interval between the emission of the pulse and the receipt of the abnormal backscattered light signal, The group refractive index is the core of the special sensing fiber (typically around 1.46). Combined with a nanosecond-level sampling rate, this algorithm can achieve meter-level or even sub-meter-level positioning accuracy.

[0093] Example 4 A GIL gas leak monitoring system based on distributed optical fiber includes: The simulation analysis module is used to lay and statically calibrate distributed optical fibers based on the optical fiber laying and routing scheme, GIL tunnel environment and physical field simulation, and to construct a nonlinear spatial mapping matrix between the optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of GIL. The signal acquisition module is used to inject probe pulses into the distributed optical fiber through an optical fiber demodulator and continuously acquire backscattered light signals. The data processing module is used to perform orthogonal demodulation and adaptive noise reduction on the backscattered light signal to obtain a noise-reduced signal; The leak detection module is used to extract time-frequency domain features from the noise-reduced signal and input the time-frequency domain features into the pattern recognition model for feature judgment to determine whether a gas leak event exists. A visualization platform is used to locate and display faults based on denoised signals and nonlinear spatial mapping matrices when a gas leak occurs.

[0094] Optionally, the simulation analysis module is specifically used for: Calibration points were selected along the GIL line, and a sound source was simulated at the calibration point to conduct calibration tests and obtain the characteristic reflection peaks on distributed optical fiber 1. Based on the calibration point location and the characteristic reflection peak, a nonlinear spatial mapping matrix is ​​established between the distributed optical fiber 1 cabling length L and the GIL three-dimensional spatial coordinates (X, Y, Z).

[0095] Optionally, the data processing module is specifically used for: The backscattered light signal is orthogonally demodulated, and the monitoring blind zone caused by signal fading is eliminated by arctangent calculation to extract the true interference phase information; An adaptive filtering algorithm is introduced to strip the background noise from the real interference phase signal, filtering out the inherent low-frequency environmental noise in the GIL tunnel, and obtaining a noise-reduced signal.

[0096] Optionally, it also includes: a model training module for: Backscattered light signals are collected by distributed optical fiber 1 laid on the surface of the GIL device. The backscattered light signals are converted into vibration time-domain signals characterizing gas leakage by distributed acoustic sensing demodulation technology. Time-frequency domain features are extracted from the vibration time-domain signals. Based on the actual gas leakage state of GIL corresponding to the signal acquisition time, a category label is set for the time-frequency domain features, and a sample is constructed from the time-frequency domain features and the corresponding category label; A sample set is constructed from multiple samples; Based on the sample set, the hyperparameters of the support vector machine are optimized by grid search and k-fold cross-validation to complete the training of the support vector machine classifier and obtain the pattern recognition model. The category labels include: gas leak and normal noise.

[0097] Optionally, the signal acquisition module is specifically used for: The fiber demodulator continuously emits ultra-narrow linewidth pulsed lasers into distributed fiber 1, and uses a high-sensitivity photodetector to receive backscattered Rayleigh or Brillouin scattered light signals carrying information about the external sound field, micro-vibration and temperature transients; and forms a two-dimensional optical signal matrix containing spatial and temporal domains. Among them, the backscattered light signal includes: back Rayleigh scattering light and Brillouin scattering light signal.

[0098] Optionally, the leakage detection module is specifically used for: Extracting time-frequency domain features from denoised signals; The time-frequency domain features are input into the pattern recognition model for feature determination to obtain category labels; When the category label is "gas leak," a gas leak event has occurred. When the category label is normal noise, there is no gas leak event.

[0099] Optionally, the visualization platform is specifically used for: When a gas leak occurs, the specific fiber length location where the leak occurred is calculated based on the optical pulse time delay when the abnormal signal occurs in the noise-reduced signal. The absolute physical coordinates of the leak are displayed intuitively based on the specific fiber length location where the leak occurred, combined with a nonlinear spatial mapping matrix.

[0100] The hardware equipment for the GIL gas leak monitoring system based on distributed optical fiber includes: This computer device, acting as the "brain" of the system, includes: The heterogeneous computing unit here includes a simulation analysis module, a data processing module, and a leakage detection module. The modules are interconnected and their specific functions are as described above, and will not be repeated here.

[0101] Heterogeneous Computing Unit: Facing the massive data generated by sampling rates of hundreds of megahertz (generating GB-level data per second), the system adopts a CPU+GPU heterogeneous computing architecture. The CPU is responsible for the logical control of the operating system, peripheral communication scheduling, and alarm linkage logic; the GPU utilizes its powerful parallel processing capabilities, through the CUDA core, to specifically handle matrix operations on massive one-dimensional optical signals, fast Fourier transform (FFT) demodulation, wavelet denoising, and high-speed inference for SVM pattern recognition, achieving millisecond-level response latency.

[0102] Data communication bus: Stable reception of data streams from the underlying demodulator via Gigabit Ethernet port or fiber optic transceiver. It is also equipped with standard Modbus-TCP or IEC 61850 communication protocol interfaces for seamless integration with existing SCADA integrated monitoring systems in the power grid.

[0103] Visualization Platform: A 3D digital twin scene of the GIL (Gas Infrared) tunnel is constructed based on WebGL and BIM (Building Information Modeling) technologies. The system accurately converts the one-dimensional positioning distance d calculated in step S35 into coordinates (X, Y, Z) on the 3D model by calling a spatial mapping matrix. On the monitoring interface, not only is a red dynamic ripple highlighted as a warning, but an API command is also automatically triggered to automatically turn and focus on the physical location of the leak, allowing maintenance personnel to conduct a secondary verification.

[0104] This embodiment provides a non-volatile storage medium (such as an industrial-grade solid-state drive, read-only memory, or flash memory device) containing computer-readable instructions. When the instruction set in this storage medium is read and executed by the processor of the aforementioned computer device, multiple software modules (including: simulation analysis module, signal acquisition module, data processing module, leakage determination module, and visualization platform) can be instantiated and run, thereby realizing fully automatic, 24-hour uninterrupted GIL leakage monitoring data stream processing. It can output high-precision location alarm results without manual intervention and automatically generate structured monitoring log files with timestamps and waveform features for subsequent traceability analysis.

[0105] The beneficial effects of this invention are: (1) Long distance without blind spots and low cost: One optical fiber can cover tens of kilometers of pipeline, completely replacing a large number of point sensors, greatly reducing construction and maintenance costs, and providing continuous monitoring of the entire line without any spatial blind spots.

[0106] (2) High-precision positioning and quantification: By utilizing the principle of light speed and time difference, accurate fault location at the meter level or even sub-meter level can be achieved, shortening the repair and troubleshooting time.

[0107] (3) Electromagnetic interference resistance and high reliability: The sensing fiber itself is a passive device, which is completely immune to strong electromagnetic interference in the GIL tube gallery. It has extremely strong environmental adaptability and truly realizes 24-hour all-weather high-reliability monitoring.

[0108] Example 5 like Figure 9 As shown, the present invention also provides an electronic device, which may be a computer device, a microcontroller device, a smart mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, processor, and transceiver component are connected via a bus; the memory can be used to store executable programs, and an exemplary executable program may include instructions; the processor is used to execute the instructions stored in the memory. The memory can also be used to store data, which can be accessed and / or modified when instructions are executed.

[0109] The processor may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, and it is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the storage medium to realize the corresponding method flow or corresponding function, so as to realize the steps of a distributed optical fiber routing design method or a GIL gas leakage monitoring method based on distributed optical fiber in the above embodiments.

[0110] Example 6 Based on the same inventive concept, this invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory). This readable storage medium is a memory device within an electronic device used to store programs and data. It is understood that the storage medium here can include both built-in storage media within the electronic device and extended storage media supported by the electronic device. The storage medium provides storage space, which stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more executable programs (including program code). It should be noted that the storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. Loading and executing one or more instructions stored in the storage medium by the processor can implement the steps of a distributed optical fiber routing design method or a GIL gas leak monitoring method based on distributed optical fiber in the above embodiments.

[0111] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0112] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0113] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0114] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0115] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A method for determining distributed optical fiber routing, characterized in that, include: The physical and environmental parameters of the target gas-insulated metal-enclosed transmission line (GIL) are obtained, a GIL geometric model is established, and a multi-physics coupled simulation of fluid dynamics and thermodynamics is performed based on the GIL geometric model under different leakage orifice diameters. Fourier transform analysis was used to analyze the coupled simulation results to determine the key leakage points; A distributed fiber optic cabling scheme is generated based on key leakage points.

2. The method as described in claim 1, characterized in that, The method of using Fourier transform analysis of coupled simulation results to determine key leakage points includes: Three-dimensional slice data of the sound field and temperature field under gas leakage conditions were extracted from the coupled simulation results. The power spectral density of the time-domain signal near the leakage source is calculated by analyzing the three-dimensional slice data using Fourier transform. When the energy proportion of a certain frequency band exceeds a set threshold, it is determined that the signal energy is concentrated in that frequency band. Calculate the spatial signal-to-noise ratio for each coordinate point or shell location within the frequency band; Based on the spatial signal-to-noise ratio, determine the coordinate point or shell location that is most sensitive to changes in leakage sound / temperature, and use the most sensitive coordinate point or shell location as the critical leakage point.

3. The method as described in claim 1, characterized in that, The distributed fiber optic cabling scheme based on key leakage points includes: Analysis of the critical leakage points located on the straight cylindrical section revealed that the critical leakage points are concentrated at the top and bottom of the GIL metal shell of the straight cylindrical section. The fiber optic cable laying scheme for the straight cylindrical section is as follows: using special magnetic clamps or anti-aging and insulating epoxy resin, the distributed optical fiber is tightly attached to the top and bottom of the GIL metal shell of the straight cylindrical section. The key leakage points on the flange expansion joint were analyzed. The key leakage points were concentrated on the windward / downwind side of the flange interface and the folds of the bellows of the expansion joint. The fiber optic cable routing scheme for the flange expansion joint is as follows: a spiral winding method is adopted. The GIL includes: a straight cylindrical section and a flange expansion joint.

4. The method as described in claim 3, characterized in that, The epoxy resin is used to fix the distributed optical fibers to the top and bottom of the GIL metal shell of the straight cylindrical section according to the fixed binding point spacing.

5. The method as described in claim 4, characterized in that, The spacing between the fixed binding points is set within the range of 1 meter to 2 meters; The spiral winding method uses a winding pitch of 5cm to 10cm.

6. A distributed optical fiber routing determination system, characterized in that, include: The simulation module is used to obtain the physical and environmental parameters of the target gas-insulated metal-enclosed transmission line (GIL), establish the GIL geometric model, and perform fluid dynamics and thermodynamic multiphysics coupling simulations under different leakage orifice diameters based on the GIL geometric model. The leakage analysis module is used to analyze the coupled simulation results using Fourier transform to determine the key leakage points; The scheme generation module is used to generate distributed fiber optic cabling schemes based on key leakage points.

7. The system as described in claim 6, characterized in that, The leakage analysis module is specifically used for: Three-dimensional slice data of the sound field and temperature field under gas leakage conditions were extracted from the coupled simulation results. The power spectral density of the time-domain signal near the leakage source is calculated by analyzing the three-dimensional slice data using Fourier transform. When the energy proportion of a certain frequency band exceeds a set threshold, it is determined that the signal energy is concentrated in that frequency band. Calculate the spatial signal-to-noise ratio for each coordinate point or shell location within the frequency band; Based on the spatial signal-to-noise ratio, determine the coordinate point or shell location that is most sensitive to changes in leakage sound / temperature, and use the most sensitive coordinate point or shell location as the critical leakage point.

8. The system as described in claim 6, characterized in that, The scheme generation module is specifically used for: Analysis of the critical leakage points located on the straight cylindrical section revealed that the critical leakage points are concentrated at the top and bottom of the GIL metal shell of the straight cylindrical section. The fiber optic cable laying scheme for the straight cylindrical section is as follows: using special magnetic clamps or anti-aging and insulating epoxy resin, the distributed optical fiber is tightly attached to the top and bottom of the GIL metal shell of the straight cylindrical section. The key leakage points on the flange expansion joint were analyzed. The key leakage points were concentrated on the windward / downwind side of the flange interface and the folds of the bellows of the expansion joint. The fiber optic cable routing scheme for the flange expansion joint is as follows: a spiral winding method is adopted. The GIL includes: a straight cylindrical section and a flange expansion joint.

9. A method for monitoring GIL gas leaks based on distributed optical fibers, characterized in that, include: Based on the fiber optic cabling scheme, GIL tunnel environment and physical field simulation, the distributed fiber optic cabling and static calibration are carried out, and a nonlinear spatial mapping matrix between the distributed fiber optic cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of GIL is constructed. Probe pulses are injected into the distributed optical fiber using an optical fiber demodulator to continuously acquire backscattered light signals; then, orthogonal demodulation and adaptive noise reduction are performed on the backscattered light signals to obtain a noise-reduced signal. Time-frequency domain features are extracted from the denoised signal and input into a pre-trained pattern recognition model for feature determination to identify whether a gas leak event exists. When a gas leak event exists, fault location and display are performed based on the denoised signal and a nonlinear spatial mapping matrix. The fiber optic cable laying scheme is determined using a distributed fiber optic cable routing determination method as described in any one of claims 1-5.

10. The method as described in claim 9, characterized in that, The process of laying and statically calibrating distributed optical fibers based on the optical fiber laying scheme, GIL tunnel environment, and physical field simulation, and constructing a nonlinear spatial mapping matrix between the distributed optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of the GIL, includes: Distributed fiber optic cables are laid according to the fiber optic cabling plan. Calibration points were selected along the GIL line, and a sound source was simulated at the calibration point to conduct calibration tests and obtain the characteristic reflection peaks on the distributed optical fiber. Based on the calibration point location and the characteristic reflection peak, a nonlinear spatial mapping matrix is ​​established between the distributed optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of the GIL.

11. The method as described in claim 9, characterized in that, The process of orthogonally demodulating and adaptively denoising the backscattered light signal to obtain a denoised signal includes: The backscattered light signal is orthogonally demodulated, and the monitoring blind zone caused by signal fading is eliminated by arctangent calculation to extract the true interference phase information; An adaptive filtering algorithm is introduced to strip the background noise from the real interference phase signal, filtering out the inherent low-frequency environmental noise in the GIL tunnel, and obtaining a noise-reduced signal.

12. The method as described in claim 9, characterized in that, The training of the pattern recognition model includes: Backscattered light signals are collected by distributed optical fibers laid on the surface of the GIL device. The backscattered light signals are converted into vibration time-domain signals that characterize gas leakage by distributed acoustic sensing demodulation technology. Time-frequency domain features are extracted from the vibration time-domain signals. Based on the actual gas leakage state of GIL corresponding to the signal acquisition time, a category label is set for the time-frequency domain features, and a sample is constructed from the time-frequency domain features and the corresponding category label; A sample set is constructed from multiple samples; Based on the sample set, the hyperparameters of the support vector machine are optimized by grid search and k-fold cross-validation to complete the training of the support vector machine classifier and obtain the pattern recognition model. The category labels include: gas leak and normal noise.

13. The method as described in claim 9, characterized in that, The step of injecting probe pulses into the distributed optical fiber using an optical fiber demodulator and continuously acquiring backscattered light signals includes: The fiber optic demodulator continuously emits ultra-narrow linewidth pulsed lasers into the distributed optical fiber, and a high-sensitivity photodetector receives backscattered Rayleigh or Brillouin scattered light signals carrying information about the external sound field, micro-vibration and temperature transients; and forms a two-dimensional optical signal matrix containing spatial and temporal domains. Among them, the backscattered light signal includes: back Rayleigh scattering light and Brillouin scattering light signal.

14. The method as described in claim 9, characterized in that, When a gas leak event occurs, fault location and display are performed based on noise-reduced signals and a nonlinear spatial mapping matrix, including: When a gas leak occurs, the specific fiber length location where the leak occurred is calculated based on the optical pulse time delay when the abnormal signal occurs in the noise-reduced signal. The absolute physical coordinates of the leak are displayed intuitively based on the specific fiber length location where the leak occurred, combined with a nonlinear spatial mapping matrix.

15. A GIL gas leak monitoring system based on distributed optical fiber, characterized in that, include: The simulation analysis module is used to lay and statically calibrate distributed optical fibers based on the optical fiber laying and routing scheme, GIL tunnel environment and physical field simulation, and to construct a nonlinear spatial mapping matrix between the distributed optical fiber cabling length L and the three-dimensional spatial coordinates (X, Y, Z) of GIL. The signal acquisition module is used to inject probe pulses into the distributed optical fiber through an optical fiber demodulator and continuously acquire backscattered light signals. The data processing module is used to perform orthogonal demodulation and adaptive noise reduction on the backscattered light signal to obtain a noise-reduced signal; The leak detection module is used to extract time-frequency domain features from the noise-reduced signal and input the time-frequency domain features into the pattern recognition model for feature judgment to determine whether a gas leak event exists. A visualization platform for locating and displaying faults based on denoised signals and nonlinear spatial mapping matrices when a gas leak occurs; The fiber optic cable laying scheme is determined using a distributed fiber optic cable routing determination method as described in any one of claims 1-5.

16. The system as described in claim 15, characterized in that, The data processing module is specifically used for: The backscattered light signal is orthogonally demodulated, and the monitoring blind zone caused by signal fading is eliminated by arctangent calculation to extract the true interference phase information; An adaptive filtering algorithm is introduced to strip the background noise from the real interference phase signal, filtering out the inherent low-frequency environmental noise in the GIL tunnel, and obtaining a noise-reduced signal.

17. The system as claimed in claim 15, characterized in that, Also includes: The model training module is used for: Backscattered light signals are collected by distributed optical fibers laid on the surface of the GIL device. The backscattered light signals are converted into vibration time-domain signals that characterize gas leakage by distributed acoustic sensing demodulation technology. Time-frequency domain features are extracted from the vibration time-domain signals. Based on the actual gas leakage state of GIL corresponding to the signal acquisition time, a category label is set for the time-frequency domain features, and a sample is constructed from the time-frequency domain features and the corresponding category label; A sample set is constructed from multiple samples; Based on the sample set, the hyperparameters of the support vector machine are optimized by grid search and k-fold cross-validation to complete the training of the support vector machine classifier and obtain the pattern recognition model. The category labels include: gas leak and normal noise.

18. An electronic device, characterized in that, include: At least one processor and memory; The memory and processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, a distributed optical fiber routing design method as described in any one of claims 1 to 5, or a GIL gas leak monitoring method based on distributed optical fiber as described in any one of claims 9 to 14, is implemented.

19. A readable storage medium, characterized in that, It contains an execution program, which, when executed, implements a distributed optical fiber routing design method as described in any one of claims 1 to 5, or a GIL gas leakage monitoring method based on distributed optical fiber as described in any one of claims 9 to 14.

Images