Fusion type multi-parameter distributed optical fiber sensing system with low polarization fading noise

Abstract

The invention discloses a low polarization fading noise fusion type multi-parameter distributed optical fiber sensing system. The broadband light source module generates a modulated light signal and transmits the modulated light signal to the optical fiber circulator, and the 2 * 2 optical fiber coupler is cascaded with the 2 * 1 optical fiber coupler and the depolarization module to form an optical fiber interferometer. The output port of the optical fiber circulator is connected with the incident end of the optical fiber interferometer; and the output end of the 2 * 1 optical fiber coupler is connected with a section of long-distance transmission sensing optical fiber. Downlink output light of the 2 * 2 optical fiber coupler and output light of the optical fiber circulator enter the 2 * 2 optical fiber coupler and are connected with a signal receiving and processing module through a photoelectric converter for double-signal synchronous detection, optical information is demodulated from the two parts of intensity and frequency, real-time detection of physical parameters such as vibration, temperature and stress is achieved, and real-time and uninterrupted detection is guaranteed. According to the invention, the defects of single demodulation mode, high noise, complex structure and the like of a traditional distributed optical fiber sensing system are overcome, and multi-parameter sensing application is realized while the polarization fading noise of an interference type optical fiber sensing system is effectively reduced.

Description

Technical Field

[0001] This invention relates to the field of fiber optic sensing technology, and in particular to a fusion-type multi-parameter distributed fiber optic sensing system with low polarization fading noise. Background Technology

[0002] Fiber optic sensing is a technology that uses optical fiber as a medium to acquire external physical quantities (such as temperature, vibration, magnetic field, current, angular velocity, etc.) modulated in optical signals (such as intensity, wavelength, phase, etc.) and demodulate them to extract effective information.

[0003] Currently, interferometric fiber optic sensors represent an important branch of fiber optic sensor research and application. In interferometric fiber optic sensing systems using single-mode fiber as the sensing medium, when light propagates through the fiber, it is not an ideally symmetrical circular cross-section with a straight axis and uniform refractive index distribution. Besides these factors, external environmental conditions such as temperature, pressure, and bending can cause random fluctuations in the polarization state, resulting in random fluctuations in the output light intensity during interference. This phenomenon is called polarization-induced signal fading. During the detection of backscattered signals, polarization fading leads to an overall reduction in the output signal of the backscattering curve, decreasing the signal-to-noise ratio and sensitivity of the fiber optic sensing unit. Simultaneously, fluctuations in the interference light intensity are also one of the main sources of noise in fiber optic sensing systems.

[0004] Polarization equivalent time-domain and frequency-domain descriptions: In a Gaussian beam, the time width is... τ (full width at half maximum), spectral width υ=0.44 / τ, assuming the polarization state changes by at least 2π within the full width at half maximum (FWHM), the fiber length must be greater than c / ( υδn), where δn is the fiber birefringence. At this length, the pulse will be separated by 2.7 times the pulse width, and 82% of the light will be uncorrelated.

[0005] For backscattered fiber optic sensing systems, both the incident and backscattered light require depolarization, necessitating the combined action of three fiber segments. The polarization degree P of a two-segment depolarization structure can be expressed as: (1) Where θp is the angle between the polarization direction of the incident light and the X-axis of the coordinate system, and θ is the angle between the polarization-maintaining fiber and the fusion splice. From the above equation (1), it can be seen that the polarization degree of the outgoing light is 0 only when θ = 45°.

[0006] Furthermore, distributed fiber optic sensors possess significant advantages such as resistance to electromagnetic interference, high sensitivity, low manufacturing cost, and wide monitoring range. They enable effective structural safety monitoring and are widely used in fundamental structural health monitoring fields such as machinery manufacturing, bridge transportation, and oil and gas pipelines. These sensors are used to predict and effectively prevent potential threats from infrastructure emergencies or accidents. Meanwhile, the development of high signal-to-noise ratio and multi-parameter sensors, driven by miniaturization and multi-functional sensor solutions, as well as the need to measure multiple physical quantities in confined operating environments, is propelling distributed fiber optic devices towards integration and multi-functionality.

[0007] In summary, existing distributed fiber optic sensing technologies have the following limitations: (1) Strong noise. At present, distributed fiber optic sensing technology mainly uses narrow linewidth fiber lasers as light sources to excite nonlinear scattering effects to achieve sensing applications, which brings strong nonlinear scattering signals and polarization-induced signal attenuation, greatly reducing the system signal-to-noise ratio and spatial resolution.

[0008] (2) Single detection parameter. Currently, distributed fiber optic sensing technology can only monitor a single physical quantity such as vibration, temperature, or stress. However, in specific engineering applications, there are often multiple physical parameters changing simultaneously and the need to monitor different parameters at the same time. The single scattering mechanism limits the number of parameters that can be measured and cannot accurately and comprehensively reflect the true state of the object being measured.

[0009] (3) Multi-physical parameter systems are complex. At present, distributed fiber optic sensing technology has made breakthroughs in realizing dual-parameter sensing of vibration and temperature by multiplexing dual systems, including nonlinear Brillouin and Raman scattering systems. However, the system integration is low and cannot be effectively integrated.

[0010] Therefore, by introducing a self-developed high-power broadband light source, integrating linear and nonlinear scattering effects, and utilizing a depolarization structure to effectively reduce system scattering noise, the technical bottlenecks of low signal-to-noise ratio and inability to achieve multi-parameter collaborative detection in existing distributed fiber optic sensing technologies can be solved. Summary of the Invention

[0011] To address the shortcomings of existing technologies, the present invention aims to provide a low polarization fading noise fused multi-parameter distributed optical fiber sensing system, which solves the problems of high polarization fading noise, poor structural compactness, single detection parameter, and low system integration in existing distributed optical fiber sensing systems.

[0012] This invention overcomes polarization fading in distributed fiber optic sensing systems by integrating linear and nonlinear scattering effects to achieve multi-parameter collaborative sensing applications. The invention features a compact structure and ease of packaging. By introducing this depolarization structure into the fiber optic sensing link, the signal-to-noise ratio of the backscattered signal is improved, increasing the sensitivity of the sensing system. By integrating linear and nonlinear scattering effects, it overcomes the limitations of monitoring single vibration parameters, enabling synchronous, real-time monitoring of multiple parameters such as vibration, temperature, and stress. It relates to the removal of polarization fading noise and multi-parameter sensing applications.

[0013] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: To achieve the above objectives, this invention provides a low-polarization fading noise fused multi-parameter distributed fiber optic sensing system, comprising a broadband light source module, a fiber optic circulator, a first 2×2 fiber optic coupler, a fiber depolarization structure, a 2×1 fiber optic coupler, a second 2×2 fiber optic coupler, a first photodetector, a second photodetector, a signal receiving and processing module, a comprehensive signal processing module, a linear signal processing module, and a nonlinear signal processing module. The broadband light source module generates a modulated optical signal, which is then transmitted to the input end of the fiber optic circulator. The first 2×2 fiber optic coupler, the second 2×2 fiber optic coupler, and the 2×1 fiber optic coupler are cascaded to form a fiber optic interferometer. The two output ports of the fiber optic circulator are respectively connected to the incident end of the fiber optic interferometer, and the output end of the 2×1 fiber optic coupler is connected to... A sensing fiber is connected; a fiber depolarization structure is set between the first 2×2 fiber coupler and the 2×1 fiber coupler; the downlink output light of one set of the second 2×2 fiber couplers and the output light of the fiber circulator enter the 2×2 fiber coupler together; the output light of the other two sets of the second 2×2 fiber couplers is transmitted to the signal receiving and processing module via a photoelectric converter composed of a first photodetector and a second photodetector; the signal of the signal receiving and processing module is demodulated by a linear signal processing module and a nonlinear signal processing module to realize the real-time detection of physical parameters; the backscattered signal after differential processing by the signal receiving and processing module is transmitted to the integrated signal processing module and fed back to the broadband light source module for controlling the broadband light source module, the signal receiving and processing module, and the preprocessing of the transmitted data.

[0014] Preferably, the broadband light source module of the present invention is a broadband light source with a center wavelength of 1550 nm and a bandwidth of 5 nm.

[0015] Preferably, the broadband light source module of the present invention can output pulse-modulated optical signals with a peak pulse power of 130mW, which can effectively suppress scattering noise while generating linear and nonlinear scattering in the sensing optical fiber.

[0016] Preferably, the optical fiber depolarization structure of the present invention uses three panda-type polarization-maintaining optical fibers of specific lengths, which are fused together at 45° with the fast (slow) axis, thereby reducing polarization fading noise and improving the system signal-to-noise ratio.

[0017] Preferably, under the drive of the broadband light source module, the fiber depolarization structure of the present invention separates the two orthogonally polarized beams in time and outputs two light signals with equal intensity values, making the backscattering curve of the interferometric fiber optic sensing system smoother and reducing random fluctuations, which is used for signal differential and amplification processing.

[0018] Preferably, the signal receiving and processing module of the present invention integrates linear and nonlinear scattering for dual-signal synchronous detection, and demodulates the output signal from both intensity and frequency aspects, including time domain, frequency domain and spectral dimensions, which can realize multi-parameter sensing and detection of vibration, temperature and stress in real time without interruption.

[0019] Preferably, the integrated signal processing module of the present invention performs sampling, filtering, periodic heterodyne, periodic intradyne, and data compression processing on the differential backscattered signal. Combined with the deep learning pattern recognition algorithm on the host computer, it constructs a deep convolutional neural network multimodal optical signal feature decoupling model, which intelligently realizes the demodulation, recognition, and prediction of multi-parameter signals of vibration, temperature, and stress.

[0020] Compared with the prior art, the present invention, employing the above technical solution, has the following technical effects: The fiber depolarization structure of this invention uses polarization-maintaining fiber for depolarization, which is characterized by its small size and ease of packaging. The introduction of the depolarization structure smooths the backscattering curve, which is beneficial for improving the signal-to-noise ratio and sensitivity of the sensing system.

[0021] This invention integrates linear and nonlinear scattering effects, combined with a self-developed broadband light source, to achieve simultaneous sensing of different physical parameters, including vibration, temperature, and stress, from both intensity and frequency dimensions. Attached Figure Description

[0022] Figure 1 This is a diagram of a fusion-type multi-parameter distributed optical fiber sensing system based on low polarization fading noise.

[0023] Figure 2 The PI curve and spectrum of the broadband light source being tested.

[0024] Figure 3 A schematic diagram and a physical diagram of a depolarizer for polarization-maintaining fiber structures.

[0025] Figure 4 The backscattering curve fluctuation diagram of the polarization-maintaining fiber acting on the distributed optical fiber sensing system.

[0026] Figure 5 This is a decoupling model for multimodal optical signal features in deep convolutional neural networks.

[0027] Figure 6 To utilize intensity and frequency to achieve multi-parameter demodulation maps of vibration, temperature, and stress. Detailed Implementation

[0028] The objectives, advantages, and features of this invention will be illustrated and explained through the following non-limiting description of preferred embodiments. These embodiments are merely typical examples of applying the technical solutions of this invention, and all technical solutions formed by equivalent substitutions or equivalent transformations fall within the scope of protection claimed by this invention.

[0029] like Figure 1 As shown, a low polarization fading noise fused multi-parameter distributed fiber optic sensing system includes a broadband light source module 1, a fiber optic circulator 2, a first 2×2 fiber optic coupler 3, a fiber depolarization structure 4, a 2×1 fiber optic coupler 5, a second 2×2 fiber optic coupler 6, a first photodetector 7, a second photodetector 8, a signal receiving and processing module 9, a comprehensive signal processing module 10, a linear signal processing module 11, and a nonlinear signal processing module 12. The broadband light source module 1 generates a modulated optical signal, which is then transmitted to the input end of the fiber optic circulator 2. The first 2×2 fiber optic coupler 3, the second 2×2 fiber optic coupler 6, and the 2×1 fiber optic coupler 5 are cascaded to form a fiber optic interferometer. The two output ports of the fiber optic circulator 2 are respectively connected to the incident end of the fiber optic interferometer, and the output end of the 2×1 fiber optic coupler 5 is connected to a section of... The sensing fiber optic cable includes a fiber depolarization structure 4 between the first 2×2 fiber coupler 3 and the 2×1 fiber coupler 5. The downlink output light from one set of the second 2×2 fiber couplers 6 and the output light from the fiber optic circulator 2 enter the 2×2 fiber coupler together. The output light from the other two sets of the second 2×2 fiber couplers 6 is transmitted to the signal receiving and processing module 9 via a photoelectric converter composed of a first photodetector 7 and a second photodetector 8. The signal from the signal receiving and processing module 9 is demodulated by a linear signal processing module 11 and a nonlinear signal processing module 12 to achieve real-time detection of physical parameters. The differentially processed backscattered signal from the signal receiving and processing module 9 is transmitted to the integrated signal processing module 10 and fed back to the broadband light source module 1 for controlling the broadband light source module, the signal receiving and processing module, and the preprocessing of the transmitted data.

[0030] The broadband light source module 1 of this invention is a broadband light source with a center wavelength of 1550 nm and a bandwidth of 5 nm. The broadband light source module 1 can output pulse-modulated optical signals with a peak pulse power of 130 mW, and can effectively suppress scattering noise while generating linear and nonlinear scattering in the sensing optical fiber.

[0031] The broadband light source module 1 of the present invention is independently developed and manufactured. It modulates the optical signal to achieve optical signal output with a certain pulse width, duty cycle and power, while effectively suppressing scattering noise and improving the signal-to-noise ratio of the system, exciting linear and nonlinear scattering effects.

[0032] The fiber depolarization structure 4 of the present invention uses three panda-type polarization-maintaining fibers of specific lengths, which are fused together at 45° with the fast and slow axes respectively, which can reduce polarization fading noise and improve the system signal-to-noise ratio.

[0033] Driven by the broadband light source module, the fiber depolarization structure 4 of the present invention separates two orthogonally polarized beams in time and outputs two light signals with equal intensity values, making the backscattering curve of the interferometric fiber optic sensing system smoother and reducing random fluctuations, which is used for signal differential and amplification processing.

[0034] The fiber depolarization structure 4 of this invention, for a light source with a certain spectral width, when the polarization-maintaining fiber is of a reasonable length, results in output light composed of two orthogonal polarization components with equal intensity and zero correlation coefficient. This makes the backscattering curve of the interferometric fiber optic sensing system smoother, reduces random fluctuations and facilitates differential processing, effectively reducing the polarization fading noise of the interferometric fiber optic sensing system.

[0035] The signal receiving and processing module 9 of this invention integrates linear and nonlinear scattering for simultaneous dual-signal detection, and demodulates the output signal from both intensity and frequency perspectives, including time domain, frequency domain and spectral dimensions, enabling real-time and uninterrupted multi-parameter sensing and detection of vibration, temperature and stress.

[0036] The integrated signal processing module 10 of this invention performs a series of processing steps on the differential backscattered signal, including sampling, filtering, periodic heterodyne, periodic intradyne, and data compression. The processed data is then uploaded to the host computer software. Combined with deep learning pattern recognition algorithms, a multimodal optical signal feature decoupling model of neural network and deep convolutional neural network is constructed to achieve demodulation and detection of signals such as vibration, temperature, and stress.

[0037] The integrated signal processing module 10 of this invention is used to receive two backscattered signals from the sensing optical path and perform differential amplification, DC bias adjustment, multi-stage signal switching, and backscattered trace differential functions. The main processing flow is as follows: 1. The two backscattered signals are converted into photoelectric signals by a photodetector and then enter the preamplifier. They are then filtered by a low-pass filter to remove interference noise signals and are subjected to voltage-controlled operational amplifier and DC bias.

[0038] 2. The bias signal is introduced into the subtractor for differential processing, and then amplified and filtered again.

[0039] Figure 2The graph shows the relationship between the total emitted optical power and the injected current (PI curve), the spectrum, and the modulated pulsed light output of the broadband light source under test. From the PI characteristic fitting curve, it can be seen that the threshold current is 17.4 mA, the slope efficiency is 0.192 mA / mW, and it approaches its output optical power limit at 700 mA. Figure 2 (a)). The spectrum shows a center wavelength of 1550 nm and a 3 dB bandwidth of 5.195 nm. Figure 2 (b)). Figure 2 (c) shows a detailed view of a single pulse output from the light source, but waveform distortion exists. After introducing a high-speed operational amplifier, the pulse height is stable, the duty cycle is 1%, and the output waveform is better. Figure 2 (d)).

[0040] Figure 3 This diagram illustrates the structure and physical model of a polarization-maintaining fiber depolarizer. It comprises three panda-shaped polarization-maintaining fibers fused at 45° angles along their main axes. Utilizing the high birefringence of polarization-maintaining fibers, it effectively reduces polarization-induced signal fading. When this structure is applied to an interferometric fiber optic sensing system and connected to two couplers, the polarization state of the backscattered light during interference is effectively controlled, preventing random changes in the intensity distribution during interference.

[0041] Figure 4 The backscattering curve of the interference when the depolarization structure is not connected to the fiber optic sensing system shows fluctuations. It can be seen that at 5.1 km–5.2 km, the interference light intensity peaks due to polarization-induced signal fading. At 5.35 km–5.55 km, random fluctuations in the interference light intensity occur due to polarization-induced signal fading. It can be seen that throughout the backscattering curve from 5.1–5.6 km, the polarization-induced signal fading is effectively controlled, and the random fluctuations in the interference light power decrease from 0.137 dB to 0.012 dB.

[0042] Figure 5 This is a multimodal optical signal feature decoupling model for deep convolutional neural networks. Leveraging the powerful feature extraction capabilities of deep convolutional neural networks, a unique multimodal input structure is constructed for different signal representations in spectra and light spots. By designing a combination of convolutional layers, pooling layers, and fully connected layers, key features such as wavelength, intensity distribution, and absorption peak positions in the input spectrum, as well as important information such as shape, intensity distribution, and centroid position in the light spot image, are accurately captured. A multi-channel data acquisition card is used to acquire multidimensional spectral or light spot data and train the model features. The trained network model is then used to classify, identify, and demodulate sensor data in multi-physics interference environments, enabling multi-parameter intelligent sensing applications.

[0043] Figure 6To obtain multi-parameter experimental results of vibration, temperature, and stress using intensity and frequency, real-time demodulation was implemented from three dimensions: time-domain spectrum, frequency-domain spectrum, and spectral spectrum. Four signals (impact, trampling, temperature, and stress) were selected for identification and classification. To improve the system's response sensitivity, optical fibers were bent and laid on a wire mesh. For each interference signal, 200 sets of interference data were collected, with 150 sets used for model training and 50 sets used to test the model's recognition results. The time-domain waveforms of the two vibration signals obtained using intensity demodulation are shown below. Figure 6 As shown in (a) and (b), the striking and stomping signals exhibit periodic variations. The amplitude of the stomping signal gradually decreases within one cycle, while the amplitude of the striking signal varies slightly within one cycle, with a shorter duration. The frequency domain spectrum and wavelength time domain spectrum of the stress and temperature signals obtained using frequency demodulation are shown in (a) and (b). Figure 6 As shown in (c) and 6(d), different stress signals are applied, corresponding to different frequencies in the frequency domain. Based on the relationship between stress and frequency, real-time demodulation of the stress signal is achieved. By demodulating the change of the optical frequency signal at different temperatures, as the temperature gradually increases, the wavelength of the interference spectrum gradually shifts. A temperature-wavelength fitting curve is constructed, enabling real-time demodulation of the temperature signal. Simultaneously, to achieve simultaneous demodulation of three parameters, the output signals under conditions of simultaneous application of vibration, stress, and temperature signals are collected. Combined with a pattern recognition algorithm, different signals are identified from both intensity and frequency perspectives. Based on the single-signal recognition model, a dual-signal recognition model is introduced to effectively enrich the database. The light intensity and frequency, simultaneously affected by the two parameters, are modulated to achieve demodulation of different signals.

[0044] This invention discloses a low-polarization fading noise fusion-type multi-parameter distributed optical fiber sensing system, comprising a broadband light source module, an optical fiber circulator, a 2×2 optical fiber coupler, a depolarization module, a 2×1 optical fiber coupler, a photodetector, a signal receiving and processing module, and a comprehensive signal processing module. The broadband light source module generates a modulated optical signal that is transmitted to the optical fiber circulator. A 2×2 optical fiber coupler cascaded with a 2×1 optical fiber coupler and a depolarization module constitutes an optical fiber interferometer. The two output ports of the optical fiber circulator are connected to the incident end of the optical fiber interferometer, and the output end of the 2×1 optical fiber coupler is connected to a long-distance sensing optical fiber. The downlink output light of the 2×2 optical fiber coupler and the output light from port 3 of the optical fiber circulator enter the 2×2 optical fiber coupler, and are connected to the signal receiving and processing module via a photoelectric converter for simultaneous dual-signal detection. The optical information is demodulated from both intensity and frequency components to achieve real-time detection of physical parameters such as vibration, temperature, and stress, ensuring uninterrupted real-time operation. A convolutional neural network multimodal optical signal feature decoupling model is constructed to capture feature information of different parameter data in real time and online, intelligently predicting different parameters. The integrated signal processing module controls the broadband light source module, signal receiving and processing module, and preprocesses the transmitted data. This invention integrates both linear and nonlinear scattering photosensitive mechanisms, combined with a polarization-maintaining depolarization structure to effectively reduce fading noise, while simultaneously achieving real-time collaborative detection of multiple parameters such as vibration, temperature, and stress along long-distance optical fibers. It overcomes the shortcomings of traditional distributed optical fiber sensing systems, such as single demodulation methods, high noise, and complex structures, effectively reducing polarization fading noise in interferometric optical fiber sensing systems while enabling multi-parameter sensing applications.

[0045] All devices in this invention adopt an all-fiber structure, which is compact, stable, and has strong anti-electromagnetic interference capability. It can effectively reduce polarization-induced signal fading and has high application value in interferometric fiber optic sensing systems.

[0046] This invention has multiple embodiments, and all technical solutions formed by equivalent transformations or equivalent modifications fall within the protection scope of this invention.

Claims

1. A fusion type multi-parameter distributed optical fiber sensing system with low polarization fading noise, characterized in that, The application relates to a fiber-optic interferometer system for real-time detection of physical parameters, which comprises a broadband light source module (1), a fiber-optic circulator (2), a first 2*2 fiber-optic coupler (3), a fiber-optic depolarization structure (4), a 2*1 fiber-optic coupler (5), a second 2*2 fiber-optic coupler (6), a first photoelectric detector (7), a second photoelectric detector (8), a signal receiving and processing module (9), a comprehensive signal processing module (10), a linear signal processing module (11) and a nonlinear signal processing module (12). The broadband light source module (1) generates a modulated light signal which is transmitted to the input end of the fiber-optic circulator (2), the first 2*2 fiber-optic coupler (3), the second 2*2 fiber-optic coupler (6) and the 2*1 fiber-optic coupler (5) are cascaded to form a fiber-optic interferometer, the two output ports of the fiber-optic circulator (2) are connected with the incident ends of the fiber-optic interferometer, and the output end of the 2*1 fiber-optic coupler (5) is connected with a sensing optical fiber; the fiber-optic depolarization structure (4) is arranged between the first 2*2 fiber-optic coupler (3) and the 2*1 fiber-optic coupler (5); the downlink output light of one group of the second 2*2 fiber-optic couplers (6) and the output light of the fiber-optic circulator (2) enter the 2*2 fiber-optic coupler together; the output light of the other two groups of the second 2*2 fiber-optic couplers (6) is transmitted to the signal receiving and processing module (9) through the photoelectric converter formed by the first photoelectric detector (7) and the second photoelectric detector (8); the signal of the signal receiving and processing module (9) is demodulated to optical information through the linear signal processing module (11) and the nonlinear signal processing module (12), so that the real-time detection of physical parameters is realized; the differential backscattering signal of the signal receiving and processing module (9) is transmitted to the comprehensive signal processing module (10) and is fed back to the broadband light source module (1), so as to control the broadband light source module, the signal receiving and processing module and the pre-processing of transmission data.

2. The low polarization-fading-noise fusion-type multi-parameter distributed optical fiber sensing system according to claim 1, characterized in that, The broadband light source module (1) is a broadband light source, the central wavelength of which is 1550 nm, and the bandwidth is 5 nm.

3. The low polarization-fading-noise fusion-type multi-parameter distributed optical fiber sensing system according to claim 1, characterized in that, The broadband light source module (1) can output a pulse modulated light signal, the pulse peak power of which is 130 mW, and can effectively suppress scattering noise while generating linear and nonlinear scattering in the sensing optical fiber.

4. The low polarization-fading-noise fusion-type multi-parameter distributed optical fiber sensing system according to claim 1, characterized in that, The fiber-optic depolarization structure (4) adopts three sections of panda-type polarization-maintaining optical fibers with specific lengths, which are fused at the fast and slow axes at an angle of 45 degrees, so that the polarization fading noise can be reduced and the signal-to-noise ratio of the system can be improved.

5. The low polarization-fading-noise fusion-type multi-parameter distributed optical fiber sensing system according to claim 4, characterized in that, Under the driving of the broadband light source module, the fiber-optic depolarization structure (4) separates two beams of orthogonal polarized light in time, outputs two light signals with equal intensity values, makes the backscattering curve of the interferometric fiber-optic sensing system smoother, reduces the random fluctuation, and is used for signal differentiation and amplification processing.

6. The low polarization-fading-noise fusion-type multi-parameter distributed optical fiber sensing system according to claim 1, characterized in that, The signal receiving and processing module (9) performs double-signal synchronous detection by fusing linear and nonlinear scattering, demodulates the output signal from the aspects of intensity and frequency, includes time domain, frequency domain and spectral dimension, can realize real-time and uninterrupted vibration, temperature and stress multi-parameter sensing detection.

7. The low polarization-fading-noise fusion-type multi-parameter distributed optical fiber sensing system according to claim 1, characterized in that, The integrated signal processing module (10) performs sampling, filtering, periodic heterodyne, periodic intradyne, and data compression on the differential backscattered signal. Combined with the deep learning pattern recognition algorithm on the host computer, it constructs a deep convolutional neural network multimodal optical signal feature decoupling model, and intelligently realizes the demodulation, recognition, and prediction of multi-parameter signals of vibration, temperature, and stress.

Images