An optical fiber leak detection system suspended in a water supply line and a monitoring method

Abstract

The application discloses a kind of optical fiber leak detection system and monitoring method suspended in water supply pipeline, the optical fiber leak detection system includes sensing optical fiber: suspended in water supply pipeline, hydrolysis-resistant, food-grade optical fiber;Sweeping frequency laser host computer: transmit laser and receive scattered light, sensing optical fiber is connected by optical switch, real-time detection sound signal that spreads in the axial of water supply pipeline;Algorithm server: by analyzing the characteristics of various sound signals collected by sensing optical fiber in water supply pipeline, realize the detection and positioning of pipeline leakage;Display PC: algorithm server is connected using BS architecture, and the detection and positioning result is operated and displayed.The application uses distributed optical fiber sensing pDAS system, is not influenced by intensity amplitude fluctuation, can greatly improve the leakage detection accuracy and leakage positioning accuracy of system.

Description

Technical Field

[0001] This invention relates to a pipeline leakage monitoring system and method, and more particularly to a fiber optic leakage detection system and monitoring method suspended in a water supply pipeline. Background Technology

[0002] Leaks in urban underground water supply pipes may not show any signs on the surface. Even if water seeps from the surface, the seepage point may not be the leak point; this is especially true when there is a layer of cement or asphalt pavement, making it even harder to detect. Leaks in water supply pipes not only waste water resources but can sometimes develop from small leaks into pipe bursts, causing severe flooding on the road, seriously affecting traffic, and causing water to enter roadside shops and residential buildings. Therefore, timely identification and repair of leaks in water supply pipes is one of the most effective methods for conserving water resources.

[0003] There are currently six commonly used methods for detecting leaks in water supply pipelines: area metering method, sound listening method (acoustic vibration method), infrared method, ground penetrating radar method, correlation leak detection method, and DAS (Distributed Acoustic Sensing) fiber optic sensing method.

[0004] 1. Area Metering Method: Metering discrepancies caused by leaks. This method involves comparing the flow meters entering and leaving a specific area of ​​the water supply network. The difference represents the unmetered losses within that area. If there are no other unmetered consumptions, then the leakage losses in that area can be identified, providing managers with a clear understanding of the situation. The denser the metering and the clearer the segmentation, the clearer the understanding of leakage in each segment. However, due to limitations in power supply, communication, and installation location of the flow meters, the metering cannot be too dense. Furthermore, this method cannot pinpoint the exact location of the leak, and therefore cannot be used as a basis for specific repairs or road excavation.

[0005] 2. Sound Listening Method (Sound Vibration Method): Leakage causes vibration and sound effects. The sound listening method involves using a sound-transmitting tool to listen to the sound of leaking water and determining the location of the leak based on the volume and sound quality. From simple mechanical leak-listening rods to various sound-detecting instruments, this method should essentially be called the "sound vibration method." Disadvantages: It requires holding the tool above the testing point, which is very time-consuming and labor-intensive, and relies entirely on personal experience.

[0006] 3. Infrared Method: Leaks cause localized changes in infrared radiation (temperature effect). Infrared thermal imaging detection uses photoelectric technology to detect specific bands of infrared radiation emitted by an object, converting this signal into images and graphics that can be discerned by the human eye. Infrared scanning measurements are performed in the pipeline area. When underground water leaks, a temperature difference occurs between the local area and the surrounding area, resulting in different infrared radiation patterns. The infrared image reflects this difference, allowing the leak point to be located. Alternatively, using drones or satellite remote sensing technology, high-altitude thermal images are captured and analyzed against a water supply pipeline distribution map to locate the leak. This method requires a large leak volume to cause localized ground cooling. However, due to poor underground drainage, low-lying areas with accumulated water, and other non-leakage factors, false alarms are common. Furthermore, concrete roads, asphalt roads, elevated roads, and vehicles on the road can prevent detection, leading to missed detections. Therefore, the application of this method is limited. Disadvantages include poor positioning accuracy, the need for a large leak volume for detection, a relatively high false alarm rate, and an extremely high missed detection rate. The advantage is that only one high-altitude thermal image is required.

[0007] 4. Ground Penetrating Radar (GPR): This method uses electromagnetic waves to scan the underground environment and observe the distribution of underground objects from the reflected signals. If it can provide a clear and immediate view, it is both clear and accurate. However, because the underground medium is different from air and has a highly complex and layered structure, the penetration of electromagnetic waves is limited. This is especially true around water pipes where there is already water accumulation and the nozzles are pointing downwards, making it even more difficult to see clearly. In addition, these instruments are currently expensive and have not yet reached the level of practical application.

[0008] 5. Related Leak Detection Method: In principle, this is a transplanted technique from the acoustic vibration method. The vibration caused by a leak propagates along the pipe to both sides. Sensors placed at different distances on both sides receive the sound waves emitted by the leak at a certain moment, resulting in a time difference. This time difference is determined by the sound velocity in the pipe and the location of the leak. Its main advantage is that it utilizes the good sound transmission of the pipe, allowing direct measurement on the pipe and instrument calculation to pinpoint the leak, eliminating the influence of human experience and avoiding the need for the tester to hold tools above the test point. Its practical difficulties lie in the constraints: it requires two points of direct contact with the pipe where sensors are placed, and a thorough understanding of the pipe's condition is essential, including its routing, bends, pipe diameter, sound velocity in different pipes, and good sound transmission conditions. Another factor is its high cost and the need for operators to have certain computer application skills. Disadvantages: Many practical constraints and high cost.

[0009] 6. DAS Fiber Optic Sensing Method: This method uses optical fibers to detect vibrations and sound effects caused by leaks. DAS is a distributed fiber optic sound sensing technology based on a location detection algorithm using sound wave signals. Optical fibers are tightly wrapped around or attached to the outer wall of the water supply pipe. The sensitivity of the fiber optic cable is extremely high; it can detect even the slightest vibrations, such as sound waves or seismic waves. The frequency of the audio signal detected by the fiber optic cable is analyzed to determine if it is a leak, the intensity of the received audio signal is used to determine the amount of leakage, and the location of the leak is determined by the echo delay. Disadvantages: For new pipelines, the leakage rate is extremely low within 20 years, so there is no immediate need for it; for existing pipelines, fiber optic installation requires extensive road excavation, which is constrained by cost and construction conditions.

[0010] How to solve the problem of detecting and locating leaks in water supply pipelines? Through analysis and summary of the six commonly used leak detection technologies and methods, it is concluded that distributed fiber optic sensing technology is most suitable for signal detection in long-distance pipelines, for the reasons described below, but the difficulty of construction needs to be addressed.

[0011] Distributed fiber optic sensing technology uses optical fibers as both sensors and communication channels. Requiring no power supply, it is suitable for signal detection in long-distance pipelines. It can continuously sense signal changes at any location along an entire optical cable tens of kilometers long. Slight changes in the external environment, such as temperature, strain, vibration, and sound, will affect the optical signals transmitted within the fiber. Various interference signals have different characteristics; through pattern recognition, these signals are detected and distinguished, identifying the type and precise location of the event, enabling long-distance real-time monitoring of various events. With continuous breakthroughs in AI technology, time-division multiplexing laser scanning technology can be used within every meter of optical fiber. Utilizing different optical scattering methods such as Raman, Brillouin, and Rayleigh scattering, and employing different demodulation methods and AI algorithms, it is possible to excite 20 temperature sensors, 20 strain sensors, 20 vibration sensors, 20 acoustic sensors, and so on. This is equivalent to integrating at least 20 of the most commonly used sensor types per meter of optical cable. Furthermore, fiber optic sensors have a strong ability to integrate into the field, requiring no additional power supply or communication channel; they are intrinsically stable, resistant to electromagnetic interference and lightning strikes; and can detect minute changes with extremely high sensitivity.

[0012] If optical fibers can be deployed inside pipes, and a simple jellyfish-shaped robot can drag the fibers along the water flow, the construction difficulties will be easily solved. However, two technical challenges need to be addressed: 1. The optical fiber needs to be able to float in the water pipe, detect all the sound characteristic signals emitted inside the pipe while floating, and transmit them back without loss. 2. Through spectrum analysis and pattern recognition, it can correctly distinguish between leakage water sound, background noise, and system noise, and accurately identify the characteristic frequency of the leakage signal.

[0013] The propagation of sound within pipes is highly complex, influenced by various factors such as pipe material, pipe diameter, fluid properties, and flow velocity. This is demonstrated through simulation and experimentation. 1. Both the sound propagation environment and the transmission medium can affect changes in sound characteristics; 2. The factors affecting sound propagation by pipe structure include axial propagation, reflection, and scattering.

[0014] (1) Axial Propagation Mode: Sound propagates along the axial direction of the pipe medium. This propagation mode is mainly affected by the properties of the fluid inside the pipe, including the fluid density, viscosity, and flow velocity. When a leak occurs in the pipe, high-frequency sound signals are generated at the leak point. These signals propagate along the axial direction of the pipe, such as... Figure 1 As shown.

[0015] (2) Reflection and Scattering: When sound propagates through the medium in a pipe, it will be reflected and scattered when it encounters the inner wall of the pipe, pipe bends, joints, tees, crosses or other structural changes. These reflected and scattered signals will affect the propagation path of the sound, weaken the signal strength, and increase the complexity of signal analysis. Therefore, it is necessary to use appropriate techniques to keep the signal strength unchanged. Summary of the Invention

[0016] The technical problem to be solved by the present invention is to provide an optical fiber leak detection system and monitoring method that is suspended in a water supply pipeline, which is not affected by intensity amplitude fluctuations and can significantly improve the accuracy of leak detection and leak location.

[0017] The technical solution adopted by this invention to solve the above-mentioned technical problems is to provide a fiber optic leak detection system suspended in a water supply pipeline, including: a sensing fiber optic cable: suspended in the water supply pipeline, a hydrolysis-resistant, food-grade fiber optic cable; a frequency-sweeping laser host: emitting laser light and receiving scattered light, connected to the sensing fiber optic cable through an optical switch to detect axially propagating sound signals in the water supply pipeline in real time; an algorithm server: analyzing the characteristics of various sound signals collected by the sensing fiber optic cable in the water supply pipeline to realize the detection and location of pipeline leaks; and a demonstration PC: using a B / S architecture to connect to the algorithm server, operate and display the detection and location results.

[0018] Furthermore, the core of the sensing optical fiber is a hydrolysis-resistant food-grade core, comprising the following components: silicon dioxide, phosphate, borates and metal oxides, wherein the metal oxides are titanium oxide or zinc oxide, and silicon dioxide accounts for more than 98%.

[0019] Furthermore, the core of the sensing optical fiber is covered with a hydrolysis-resistant, food-grade cladding layer, which is a fluoride coating layer.

[0020] Furthermore, the frequency-sweeping laser host includes a laser source. The emitted light generated by the laser source is modulated by a frequency-sweeping device, so that the frequency of the laser signal changes continuously within a frequency band. Then, it is sent to an optical signal amplifier via an optical system. The optical signal amplifier enhances the intensity of the optical signal by adding a small amount of erbium. The laser is then sent to a circulator via an isolator, and then emitted by an optical switch. The laser signal scattered back by the sensing fiber is then sent to a photodetector via the circulator.

[0021] To address the aforementioned technical problems, this invention also provides a fiber optic leak detection method suspended in a water supply pipeline. Employing the aforementioned fiber optic leak detection system, the monitoring method includes the following steps: S1) Leakage sound feature extraction: High-frequency components with concentrated energy in the sound signal are obtained using a sensing fiber optic cable suspended in the water supply pipeline, which manifests as short-duration burst signals in the time domain; S2) Leakage signal detection: The laser host employs a frequency sweeping mechanism to generate a continuously varying laser signal within a frequency band, overcoming the diffraction interference present in a single light source, effectively enhancing the emitted light signal at the same emission power; the laser host incorporates a laser calibration sensor to calibrate optical calibration parameters, ensuring consistent emitted light and reducing systematic errors; a polarization spectral filtering algorithm is used to filter polarization noise, and DSP noise reduction technology is employed to reduce system noise; and an NPU is integrated. Neural network iterative filters and cyclic filters are used to filter out background noise and perform signal preprocessing; a signal transmission algorithm based on phase-sensitive optical time-domain reflection and phase generation carrier demodulation is used to maintain the signal-to-noise ratio during transmission; a time-frequency feature extraction method is used to extract signal features and distinguish the leakage signal from the background noise; S3) Leakage localization: a collaborative algorithm that integrates echo time difference algorithm and sensor array mode is used, and a multi-element acoustic wave detector is configured in the signal processing algorithm module of the algorithm server to effectively eliminate obvious abnormal data in the array signal.

[0022] Furthermore, in step S2, the laser host employs a polarization acoustic spectrum filtering algorithm to filter out polarization noise and DSP noise reduction technology to reduce system noise. It also integrates an NPU neural network iterative filter and a cyclic filter to filter out background noise and perform signal preprocessing.

[0023] Furthermore, the signal preprocessing in step S2 includes reconstructing and filling in the missing data and abnormal data by using the correlation between the preceding and following data.

[0024] Furthermore, the multi-element acoustic wave detector configured in step S3 comprehensively analyzes the signal sets transmitted back from the abnormal point and its neighboring points in different time domains and at different frequencies as multiple array data, thereby eliminating obvious abnormal data in the array signal.

[0025] Furthermore, in step S2, multiple acoustic monitoring point signals near the leak point are selected to synthesize a sound source, thereby enhancing the strength of the effective signal. At the same time, beamforming technology is used to weight the array element output through delay compensation. When the focusing direction coincides with the actual signal source direction, the maximum output can be formed. In step S3, the leak location is completed by searching for the output peak point and inversely deducing the wave arrival direction.

[0026] Further, in step S3, the distance information S of the leak point is calculated according to the following formula, and the specific location of the leak point is given by means of the water supply pipeline distribution map and Beidou positioning. S = Δt × C, where C is the speed of light; △t = (t2 - t1) / 2, where t1 is the emission time of the laser beam and t2 is the reception time of the scattered wave from the same beam.

[0027] Compared with the prior art, the present invention has the following advantages: the fiber optic leak detection system suspended in the water supply pipeline provided by the present invention is not affected by the intensity amplitude fluctuation, and can greatly improve the leak detection accuracy and leak location accuracy of the system. Attached Figure Description

[0028] Figure 1 A schematic diagram showing the propagation of a high-frequency signal generated at the leak point along the axial direction of the pipeline. Figure 2 This is a schematic diagram of the fiber optic leak detection system suspended in the water supply pipeline of the present invention; Figure 3 This is a block diagram of the sweeping laser host circuit of the present invention; Figure 4 This is a diagram of the original signal acquired by this invention; Figure 5 This is an enhanced signal diagram after beamforming technology is applied in this invention; Figure 6 This is a schematic diagram illustrating how the present invention simultaneously identifies and locates multiple leaks on a pipeline. Figure 7 This is a comparison diagram showing water being released from the same point in the water supply pipeline at different distances, as per the present invention. Detailed Implementation

[0029] The present invention will now be further described with reference to the accompanying drawings and embodiments.

[0030] With the continuous iteration of AI technology and the emergence of laser frequency scanning technology, a distributed sound sensing technology (pDAS) based on phase generation carrier demodulation algorithm has emerged. It has a unique multi-dimensional vibration decomposition algorithm, which greatly improves the analysis dimension of vibration data. At the same time, based on phase-sensitive optical time-domain reflectometry (Φ-OTDR) and phase generation carrier (PGC) demodulation algorithm, the phase demodulated signal is resolved at the front end. It is not limited by the acousto-optic modulation frequency, does not lose rich phase information, is non-intensity-based, and is not affected by intensity amplitude fluctuations. It overcomes the problem of severe intensity fluctuations in Rayleigh scattering signals in both time and space scales, and can accurately reconstruct the vibration intensity and frequency information at various locations of the optical fiber in real time. The pDAS system designed in this invention exhibits superior performance in real-time signal processing and Rayleigh polarization-induced fading suppression due to its relatively low data requirements and polarization-independent structure. It suppresses coherent fading noise, achieving RIN noise (relative intensity noise) suppression of over 125 dB, significantly improving sensitivity and coherence. The system employs a neural network NPU processing unit and laser-encoded frequency sweeping technology to overcome the optical diffraction interference inherent in single-point laser sources. An integrated laser calibration sensor enables automated calibration of optical parameters, ensuring consistency of emitted light parameters and reducing system errors. Polarization noise is filtered out using a polarization acoustic spectrum filtering algorithm, and DSP noise reduction technology is employed to reduce system noise and balance the signal-to-noise ratio. Furthermore, an NPU neural network iterative filter and a cyclic filter are integrated to filter background noise and perform signal preprocessing, maintaining long-term stability of the signal-to-noise ratio. The training of a 10TB-level acoustic signal sample library, combined with neural network deep learning algorithms, significantly improves the pattern recognition rate of Rayleigh scattering signals. The feature change sensing technology employs a combination of time-domain statistical analysis and amplitude spectrum analysis using fast Fourier transform to detect abrupt changes and feature persistence in signal amplitude and frequency, thereby improving the sound feature pattern recognition rate. Data preprocessing technology adds anomaly data reconstruction techniques to conventional data cleaning techniques, reconstructing and filling in missing and abnormal data parts through correlation between preceding and subsequent data, ensuring the integrity of data analysis. A unique multi-element acoustic wave detector algorithm is added, significantly improving the system's leak detection accuracy and leak location precision. Beamforming technology, based on the pDAS system's multi-point synchronous monitoring capability, synthesizes signals from multiple acoustic wave monitoring points near the leak point (i.e., points on the same sensing fiber optic cable selected at equal intervals as fiber optic sound sensors) into a single sound source, enhancing the strength of the effective signal. Simultaneously, beamforming technology suppresses irrelevant noise, improving anti-noise interference capabilities and enhancing the reliability and accuracy of leak detection and location.

[0031] Please see Figure 2The pDAS system of this invention adopts a B / S architecture and mainly consists of a swept-frequency laser host, an optical switch (multi-channel expansion), a sensing fiber, an algorithm server, and a demonstration PC. The swept-frequency laser host detects Rayleigh scattered light carrying the characteristics of axially propagating sound signals inside the pipe by emitting laser light; the algorithm server detects and locates pipe leaks by analyzing changes in the characteristics of the sound signals propagating inside the pipe.

[0032] The pDAS system uses a swept-frequency laser generator, which is a laser device that generates continuous frequencies using swept-frequency technology. Its main components are as follows: Laser source: The core component of a frequency-sweeping laser generator is a laser diode, which is used to generate a single-frequency laser beam. Excited by an electric current, the laser diode emits nearly coherent light.

[0033] Frequency sweeping device: In order to scan within a frequency band, the laser host contains a frequency sweeping device. This device can quickly change the output frequency of the laser by changing the operating current or temperature of the laser diode. The change in current or temperature will cause the emission wavelength of the laser to change, thereby achieving the frequency change.

[0034] Optical system: The laser beam emitted by the laser source passes through a series of optical components, such as lenses and mirrors. These components are used to guide, focus, and shape the beam to ensure the stability and directionality of the laser beam during propagation.

[0035] Optical signal amplifier: It is an amplifier of optical signals. It enhances the intensity of optical signals by doping with a small amount (about 1%) of erbium. It is one of the key technologies for realizing long-distance optical signal transmission.

[0036] Control System: The sweep laser host is equipped with an advanced control system that can precisely control parameters such as laser start-up, stop, scanning speed, and frequency range, ensuring the consistency and repeatability of laser output, realizing system self-diagnosis, fiber breakage fault alarm and accurate location, etc.

[0037] An isolator is an optical element used to prevent reverse-transmitting optical signals from interfering with the system, ensuring that optical signals can only be transmitted in one direction. A circulator is a device that creates a loop in the transmission path of an optical signal, allowing the signal to propagate back and forth within the optical fiber to interact with and be detected by sound waves. This loop structure typically includes one or more fiber couplers, optical switches, and fiber loops, effectively managing the direction and path of the optical signal propagation. A coupler is an optical element that couples light emitted from a light source into the sensing fiber, or couples light signals scattered back from the sensing fiber into the detection device.

[0038] A photodetector is a sensor that converts light signals into electrical signals to detect subtle changes in light signals caused by sound waves. These changes may include variations in light intensity, phase, or frequency. The output signal of the photodetector can be further processed and analyzed to extract sound wave information.

[0039] The emission frequency of the laser light source in this invention is determined by the energy levels of the material used in the laser light source. The greater the energy level difference, the higher the center frequency of the emitted monochromatic light. After the emitted light is modulated by a frequency sweeping mechanism, the laser signal continuously changes from high to low (or from low to high) within a frequency band. Then, it is sent to an optical signal amplifier via an optical system, including calibration. The optical signal amplifier enhances the intensity of the optical signal by doping with about 1% erbium. The laser is then sent to a circulator via an isolator, to an optical switch extension, and to a multi-channel sensing fiber, thereby enabling a single laser host to monitor multiple water supply pipelines. The laser signal scattered back by the sensing fiber is then sent to a photodetector via a circulator.

[0040] The optical switch of this invention cycles through optical channels in a time sequence to expand the number of optical loops accessed by the sensing fiber. There are 4 / 8 / 16 optical switches available. The optical switches can be cascaded, and can be cascaded in two, three, or multiple stages. Since online water leakage monitoring does not have high real-time requirements, the optical channels can be expanded and reserved as much as possible in order to save costs.

[0041] The sensing optical fiber of the pDAS system of this invention is a food-grade, hydrolysis-resistant special optical fiber with a certain mechanical strength. The main structure and technical principle of this type of optical fiber are similar to its application in ordinary optical fiber communication, that is, it uses the principle of total internal reflection of light to transmit signals in the optical fiber. However, it needs to meet the requirements of food-grade optical fiber, such as biocompatibility, high temperature resistance, chemical resistance, and the ability to maintain stable performance under specific environments (high temperature, humidity, or strong light), to meet strict food safety standards and ensure that it will not cause harm to human health when exposed to tap water.

[0042] The special optical fiber of this invention uses a silicon-based material, namely quartz glass (SiO2). This material has good chemical stability and biocompatibility, and will not release harmful substances or pollute tap water upon contact. Quartz glass optical fibers are already widely used in the food industry and medical devices (gastroscopes, colonoscopes, etc.) because they can withstand the high temperatures and chemical conditions during food processing and maintain stable performance under various environmental conditions, including changes in temperature, humidity, and light. Preferably, the special optical fiber of this invention is composed as follows: Silicon dioxide (SiO2): As the main component, it provides optical properties and chemical stability.

[0043] Phosphates: Used to improve the high temperature resistance and mechanical strength of optical fibers.

[0044] Borates: enhance the chemical resistance and transparency of optical fibers.

[0045] Small amounts of metal oxides, such as titanium oxide or zinc oxide, are used to improve the mechanical and optical properties of optical fibers.

[0046] The addition of phosphates, borates, and metal oxides is typically very small to avoid adversely affecting the optical properties of the optical fiber. Precise stoichiometry and process control are used during fiber fabrication to ensure product quality and safety. Generally, silica accounts for approximately 98%, with the remaining components totaling about 2%. Preferably, the components and their proportions are as follows: Silicon dioxide (SiO2) 98%; Phosphate 0.8%; 0.7% borate; Titanium oxide 0.5%.

[0047] The above ratios can actually be adjusted according to the length of the deployed optical fiber. The longer the distance, 20km to 30km, the better the total internal reflection performance of the optical fiber is required, the purer the SiO2 should be, and the higher its proportion should be. For optical fibers with a distance of less than 10km, the proportion of SiO2 can be appropriately reduced, and the proportion of other components can be appropriately increased in the same proportion.

[0048] The optical fiber needs to be wrapped with a protective layer, namely the outer jacket, to further enhance its durability and safety. The materials of these protective layers also need to meet food safety standards and take into account the sound transmission effect. Therefore, the optical fiber jacket is made of a material with good sound absorption and transmission, light weight, and extremely stable chemical properties under extreme conditions. It can absorb sound wave signals in water to the greatest extent, float in water, resist hydrolysis, and have a certain mechanical strength to withstand the mechanical stress during the installation process.

[0049] Special optical fibers exposed to water for extended periods face hydrolysis issues. While quartz glass exhibits good chemical stability, it can still undergo hydrolysis under extreme humidity conditions, particularly in high-temperature and high-humidity environments. This can lead to performance degradation and even compromise the fiber's structural integrity.

[0050] To address the issue of hydrolysis resistance, the present invention employs the following measures during the optical fiber production process: 1. Material selection: Choose materials with higher hydrolysis resistance, or add specific chemical components to quartz glass, such as borosilicate glass or toughened glass. These materials usually have better high temperature resistance and hydrolysis resistance.

[0051] 2. Outer layer protection: A hydrolysis-resistant outer layer, such as a fluoride outer layer, is coated on the surface of the optical fiber. This material exhibits very stable chemical properties under extreme conditions and can effectively prevent moisture from affecting the optical fiber it is wrapped in.

[0052] 3. Structural design: Improve the structural design of optical fibers, such as by using reinforcing ribs or fiber-reinforced structures, to improve the mechanical strength and hydrolysis resistance of optical fibers.

[0053] 4. Manufacturing process: Special processing techniques, such as high-temperature annealing or chemical strengthening, are used in the production of optical fibers to improve their resistance to hydrolysis.

[0054] 5. Quality Control: Strict control of conditions such as temperature, humidity and chemical composition during the production process to ensure that the hydrolysis resistance of optical fibers meets food safety standards.

[0055] 6. Food Safety Certification: Ensure that the fiber optic cable passes the relevant food safety certifications, such as FDA (U.S. Food and Drug Administration) or EU (European Union) food safety certifications, to prove that it meets food safety requirements.

[0056] The process of pipeline leakage monitoring implemented by the system of this invention includes the following three steps: leakage sound feature extraction, leakage signal detection, and leakage location.

[0057] I. Extraction of Leakage Sound Features When a pipe leaks, the leak point generates an acoustic signal with a specific frequency and intensity. These signals propagate along the pipe, superimposed on the background noise, and are received by an array of fiber optic sensors suspended within the pipe. The changes in frequency components, amplitude, and phase of the sound caused by the leak are recorded. These signals also contain time delay and intensity information from the sound source to each receiver. Rayleigh scattering echoes are a phenomenon where, when laser light propagates through an optical fiber, imperfections (such as minor irregularities and defects) cause the light to scatter back. One type of scattering is Rayleigh scattering. The changes in the Rayleigh-scattered optical signal caused by the sound wave (acousto-optic modulation) can be analyzed to identify the leak signal and distinguish it from the background noise. The acoustic signal at a leak point typically has the following characteristics: a. High-frequency components: The sound signal generated by a leak usually has high-frequency components. The water supply pressure in the pipe causes fluid to spray out at the leak point, generating high-frequency vibrations. b. Concentrated energy: The energy of the leakage signal is relatively concentrated. Compared with the background noise under normal operating conditions, the leakage signal has stronger energy, thus distinguishing whether the water is being used normally at a distant point or leaking at a nearby point. c. Time-domain characteristics: The leakage signal manifests as a short-duration burst signal in the time domain, with its amplitude and frequency changing significantly over time; II. Leakage Signal Detection Feature extraction is a key step in signal detection. Signal detection requires feature extraction and analysis of the signal, combined with appropriate discrimination criteria, to obtain the final signal pattern recognition result.

[0058] a. When acquiring signals, try to increase the strength of the leaked signal (delay, intensity, phase, frequency) and reduce background noise and system white noise; I. The sensing fiber uses a special fiber with good sound absorption and transmission; II. The laser host adopts a frequency sweeping mechanism, which overcomes the defects of optical diffraction interference and enhances optical efficiency; III. The laser host adopts a low data requirement and polarization-independent structure, and also shows its superior performance in real-time signal processing and Rayleigh polarization-induced fading suppression. It suppresses coherent fading noise, and the RIN noise (Relative Intensity Noise) suppression reaches more than 125dB, which greatly improves sensitivity and coherence. IV. The laser host has a built-in laser calibration sensor to achieve automated calibration of optical calibration parameters, ensuring the consistency of emitted light parameters and reducing system errors; V. The laser host uses a polarization acoustic spectrum filtering algorithm to filter out polarization noise and DSP noise reduction technology to reduce system noise and balance the signal-to-noise ratio. It also integrates an NPU neural network iterative filter and a cyclic filter to filter out background noise and perform signal preprocessing to maintain a stable signal-to-noise ratio over a long period of time. b. Maintain a constant signal-to-noise ratio during signal transmission; The signal transmission is based on the phase-sensitive optical time-domain reflectometry (Φ-OTDR) and phase-generated carrier (PGC) demodulation algorithm. The phase demodulated signal is resolved at the front end, which is not limited by the acousto-optic modulation frequency, does not lose rich phase information, is non-intensity-based, and is not affected by intensity amplitude fluctuations. It overcomes the problem of severe intensity fluctuations in Rayleigh scattering signals in both time and space scales, and can accurately restore the vibration intensity and frequency information at various locations of the optical fiber in real time.

[0059] c. The time-frequency feature extraction method was used to extract the features of the signal and distinguish the leakage signal from the background noise.

[0060] I. Time-domain feature extraction: Analyzing the characteristics of the signal in the time domain. This typically requires combining parameters such as peak value, mean, and variance of the signal in the time domain to identify the sudden changes in the signal's time-domain characteristics at the time of a leak, thus enabling signal detection.

[0061] II. Frequency Domain Feature Extraction: Analyzing the characteristics of the signal in the frequency domain. Commonly used spectrum analysis methods include Fourier transform and wavelet transform, which convert the time-domain signal to the frequency domain. By extracting and filtering the frequency components of the signal, the frequency characteristic changes of the signal when leakage occurs can be identified, thus realizing frequency domain feature identification of the leakage signal.

[0062] III. By extracting the characteristic changes of the sensed signal in the time and frequency domain and performing pattern recognition of multi-dimensional signals, the influence of pipeline topology or external interference can be reduced, thereby improving the accuracy of leak signal detection.

[0063] III. Leakage Location Based on the signal features extracted during signal detection, the system uses advanced positioning algorithms to locate leak points. The leak location algorithm integrates a collaborative algorithm for sensor array mode and an echo time difference (Δt) algorithm. Considering the unique array-type sensor characteristics of the distributed system, an algorithm for a multi-element acoustic wave detector is added to the background algorithm. Big data is used to check for false positives, significantly improving the system's leak detection accuracy. S = Δt × C, Δt = (t2 - t1) / 2, where t1 is the laser beam emission time, t2 is the reception time of the scattered wave from the same beam, and C is the speed of light. The distance to the leak point is calculated, and the specific location is given using a water supply pipeline distribution map and BeiDou positioning. The positioning accuracy for a 30km long pipeline is within 2 meters, and for a 1km long pipeline, the positioning accuracy is within 0.5 meters.

[0064] a. Spectrum analysis and acoustic signal processing Spectrum analysis identifies the characteristic frequencies of leaked signals by analyzing the frequency components of sound signals. It primarily employs time-frequency conversion techniques, signal enhancement processing techniques, and active feature extraction techniques to extract leaked signals from complex environmental noise. Specific steps include: I. Signal Processing: Noise is reduced through filtering and noise reduction techniques; effective signals are enhanced through beamforming techniques, thereby improving leakage signal characteristics and detection accuracy.

[0065] II. Spectrum Analysis: Apply Fourier transform to convert the time-domain signal to the frequency domain and analyze the frequency components of the signal.

[0066] III. Characteristic Frequency Identification: Through active feature extraction technology, the characteristic variables of the signal are automatically captured and spectrum analysis is performed to identify the characteristic frequencies of the leaked signal, which are distinguished from the background noise interference frequencies, providing more accurate results for leakage feature extraction.

[0067] IV. Data preprocessing technology: In addition to conventional data cleaning techniques, anomaly data reconstruction technology is added. For missing and abnormal data parts, the abnormal parts are reconstructed and filled by the correlation between the preceding and following data, ensuring the integrity of data analysis. b. Multi-element acoustic wave detector By configuring a multi-element acoustic detector in the signal processing algorithm module, the data of multiple array elements (the set of signals scattered back from anomaly points and their adjacent points on the same optical fiber at different time domains and frequencies) can be analyzed simultaneously. This effectively eliminates obvious abnormal data (caused by interference signals) in the array signal, which can effectively increase the correct output of signal feature detection results, prevent false alarms, and improve the reliability of leak detection.

[0068] c. Beamforming technology Beamforming algorithms, including adaptive beamforming and conventional beamforming, are commonly used in array signal processing techniques. They can be used for spatial filtering and direction-of-arrival (DOA) estimation. Essentially, they weight the array element outputs to enhance the desired signal and suppress interference signals. Beamforming technology weights the array element outputs through delay compensation. Maximum output is achieved when the focusing direction coincides with the actual source direction. Localization is completed by searching for the output peak point and inversely deducing the DOA. The original signal is as follows: Figure 4 As shown; the enhanced signal (using beamforming technology) is as follows Figure 5 As shown.

[0069] d. The training of the 10TB-level acoustic signal sample library combined with neural network deep learning algorithms greatly improved the pattern recognition rate of Rayleigh scattering signals. e. Fine-tuning techniques for AI model training: Through feedback, the model is guided to adjust, thereby optimizing the model's decisions and outputs to better meet expectations. The trained model can then correctly identify whether there is normal water use in a distant location or a leak in a nearby pipeline.

[0070] The pDAS system of this invention, monitoring results of different leakage orifice diameters under different pressures in a steel pipe pressure water transmission pipeline with a diameter of 90cm and a wall thickness of 2cm are as follows: 1. The maximum detection distance at a single end is ≤30km; when the detection distance is ≤1km, the positioning accuracy is within 0.5m; when 1km≤detection distance≤10km, the positioning accuracy is within 1m; when 10km≤detection distance≤30km, the positioning accuracy is within 2m. 2. It can simultaneously identify and locate multiple leaks in a pipe, such as... Figure 6 As shown; 3. Pipe bends, tees, and crosses have almost no impact, indicating that changes in flow direction are negligible; 4. It can distinguish between normal water usage from a distant source and a pipe leak. Simulated testing: At a leak point, a 30m hose is connected. When water is released from a distance, a disturbance signal is detected, indicating a change in water pressure. Removing the hose causes the pipe to leak directly, and the disturbance immediately increases. By setting an amplitude threshold, disturbances caused by distant water usage can be eliminated. Furthermore, the signal from distant water usage is not continuous 24 hours a day; comparison with data from nighttime detection makes the distinction easy. Figure 7 As shown; 5. Optical fibers suspended in buried water pipes are protected by the pipe walls and buffered by the water medium, so the vibration interference from the ground surface and the interference from groundwater are minimal. 6. Important pipelines, such as cross-river pipelines and main water supply pipelines, can be monitored online in real time using pDAS. In complex old urban areas, the meters installed in each area can be used to initially identify areas with serious leaks, and targeted investigations can be carried out, using pDAS as a detection tool to find the leak points.

[0071] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications and improvements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be defined by the claims.

Claims

1. A method for leak detection and monitoring using optical fibers suspended in a water supply pipeline, characterized in that, The fiber optic leak detection system used includes: Sensing fiber optic cable: suspended in the water supply pipe, hydrolysis resistant, food-grade fiber optic cable; Frequency sweeping laser host: emits laser and receives scattered light, and connects to sensing optical fiber through optical switch to detect axially propagating sound signals in water supply pipeline in real time; Algorithm server: By analyzing the characteristics of various sound signals collected by the sensing optical fiber inside the water supply pipeline, the detection and location of pipeline leaks are realized; Demonstration PC: Uses a B / S architecture to connect to the algorithm server, operate and display detection and positioning results; The frequency-sweeping laser host includes a laser source. The light generated by the laser source is modulated by a frequency-sweeping device so that the frequency of the laser signal changes continuously within a frequency band. The light is then sent to an optical signal amplifier via an optical system. The optical signal amplifier enhances the intensity of the optical signal by adding a small amount of erbium. The laser signal is then sent to a circulator via an isolator, and finally emitted by an optical switch. The laser signal, which is scattered back by the sensing fiber, is then sent to a photodetector via the circulator. The monitoring method includes the following steps: S1) Leakage sound feature extraction: High-frequency components with concentrated energy in the sound signal are obtained by using a sensing optical fiber suspended in the water supply pipe, which are manifested as short burst signals in the time domain. S2) Leakage Signal Detection: The laser host adopts a frequency sweeping mechanism to generate a laser signal with continuously varying frequency within a frequency band, overcoming the optical diffraction interference present in a single light source. This is equivalent to enhancing the output signal under the same emission power. The laser host has a built-in laser calibration sensor to achieve optical calibration parameter calibration, ensuring the consistency of the output light and reducing systematic errors. Polarization noise is filtered out using a polarization acoustic spectrum filtering algorithm, and DSP noise reduction technology is used to reduce system noise. An NPU neural network iterative filter and a cyclic filter are integrated to filter out background noise and perform signal preprocessing. Signal transmission is carried out using a phase-sensitive optical time-domain reflectometry method and a phase-generated carrier demodulation algorithm to maintain a constant signal-to-noise ratio during transmission. A time-frequency feature extraction method is used to extract signal features and distinguish the leakage signal from background noise. S3) Leakage location: A collaborative algorithm that integrates echo time difference algorithm and sensor array mode, and configures a multi-element acoustic wave detector in the signal processing algorithm module of the algorithm server to effectively eliminate obvious abnormal data in the array signal.

2. The fiber optic leak detection method suspended in a water supply pipeline as described in claim 1, characterized in that, The core of the sensing optical fiber is a hydrolysis-resistant, food-grade core containing the following components: silicon dioxide, phosphate, borates, and metal oxides, wherein the metal oxides are titanium oxide or zinc oxide, and silicon dioxide accounts for more than 98%.

3. The fiber optic leak detection method suspended in a water supply pipeline as described in claim 1, characterized in that, The core of the sensing optical fiber is covered with a hydrolysis-resistant, food-grade cladding layer, which is a fluoride coating layer.

4. The fiber optic leak detection method suspended in a water supply pipeline as described in claim 1, characterized in that, The signal preprocessing in step S2 includes reconstructing and filling in the missing and abnormal parts of the data by using the correlation between the preceding and following data.

5. The fiber optic leak detection method suspended in a water supply pipeline as described in claim 1, characterized in that, The multi-element acoustic wave detector configured in step S3 analyzes the signal sets transmitted back from the abnormal point and its neighboring points in different time domains and at different frequencies as multiple array data, thereby eliminating obvious abnormal data in the array signal.

6. The fiber optic leak detection method suspended in a water supply pipeline as described in claim 5, characterized in that, Step S2 selects multiple acoustic monitoring points near the leak point to synthesize a sound source, enhancing the strength of the effective signal. At the same time, beamforming technology is used to weight the array element output through delay compensation. When the focusing direction coincides with the actual signal source direction, the maximum output can be formed. Step S3 completes the leak location by searching for the output peak point and inversely deducing the wave arrival direction.

7. The fiber optic leak detection method suspended in a water supply pipeline as described in claim 6, characterized in that, In step S3, the distance information S of the leak point is calculated according to the following formula, and the specific location of the leak point is given by the water supply pipeline distribution map and Beidou positioning. S = Δt × C, where C is the speed of light; △t = (t2 - t1) / 2, where t1 is the emission time of the laser beam and t2 is the reception time of the scattered wave from the same beam.

Images