Optical fiber fault detection system and method based on optical time domain reflection technology

Abstract

The invention discloses an optical fiber fault detection system and method based on an optical time domain reflectometry (OTDR) technology. According to the method, reflected signals are collected through an OTDR detection module, the signals are decomposed, denoised and reconstructed through wavelet transform, high-frequency noise is removed, key features are reserved, then the processed signals are input into a convolutional neural network (CNN) model for feature extraction and classification, and automatic fault type identification and accurate position positioning are achieved. Wherein in the positioning of the fault point, the threshold value adopts a dynamically adjustable threshold value. The system comprises a light source module, a circulator, an optical fiber to be detected, an OTDR detection module and a data processing module, the light source module emits dual-wavelength light pulses to solve the problem of short-distance blind areas of a traditional OTDR, and the complex environment adaptability is enhanced in combination with a dynamic threshold algorithm. Through intelligent fusion of wavelet transform and CNN, the signal denoising capability, the fault detection precision and the automation level are remarkably improved, the manual analysis cost is reduced, and the high-precision and real-time detection requirements of a modern optical fiber network are met.

Description

Technical Field

[0001] The present invention relates to the technical field of optical fiber communication, and particularly to an optical fiber fault detection system and method based on optical time domain reflectometry (OTDR). Background Art

[0002] An optical time domain reflectometer (OTDR) is a detection tool widely used in the field of optical fiber communication. It emits high-energy optical pulses into the optical fiber. When the optical pulses encounter positions such as joints, breakpoints, and loss change points during the transmission in the optical fiber, reflected light and backscattered light will be generated. The OTDR receives and analyzes these returned optical signals, and precisely measures key parameters such as the length of the optical fiber, loss distribution, and the position of the fault point by measuring the round-trip time of the optical signal and the change in optical power. However, traditional OTDRs have blind spots in short-distance optical fiber measurements, and there are also limitations in the detection accuracy of weak signals, making it difficult to meet the requirements of modern optical fiber networks for high-precision and high-resolution detection. Summary of the Invention

[0003] In view of the problems existing in the prior art, an optical fiber fault detection method based on optical time domain reflection of the present invention is characterized in that the method includes the following steps:

[0004] S1: Collect reflection signals by using an OTDR detection module;

[0005] S2: Perform wavelet transform denoising processing on the signals;

[0006] S3: Input the denoised signals into a CNN model for feature extraction;

[0007] S4: The CNN model outputs the fault type and position;

[0008] The specific steps of step S2 include the following steps:

[0009] S21: Perform wavelet transform on the reflection signals and decompose the signals by using the wavelet;

[0010] S22: After decomposing the signals, remove the high-frequency noise components in the wavelet coefficients that are lower than a certain threshold and retain the main signal features;

[0011] S23: Reconstruct the signals through inverse wavelet transform to obtain clear denoised signals as the input of the CNN model.

[0012] Preferably, the specific steps of step S21 include the following steps:

[0013] S211: Select a wavelet basis function;

[0014] S212: Set the decomposition level N;

[0015] S213: Decompose the signal into a low-frequency part and a high-frequency part using discrete wavelet transform;

[0016] S214: Continue to decompose the low-frequency part until the set number of layers is reached.

[0017] Preferably, the detection method for the fault point E is as follows:

[0018]

[0019] where CD i is the high-frequency component of the i-th layer; when E is greater than the set threshold, it is considered that there is a fault at this point.

[0020] Preferably, the detection method for the fault point E is as follows: Find the mutation point through the wavelet modulus maximum detection method; calculate the first derivative of the wavelet coefficient to find the local maximum, and the point where the maximum is located is the fault point.

[0021] Preferably, the step of determining that there is a fault at this point when E is greater than the set threshold further includes a step of calculating this threshold; this threshold is a dynamically adjustable threshold.

[0022] The present invention also provides a detection system for implementing the above detection method, and the system includes a light source module, a circulator, a fiber optic cable to be measured, an OTDR detection module, and a data processing module.

[0023] Preferably, the light source module emits optical pulses of two different wavelengths.

[0024] Preferably, the circulator is used to introduce the optical pulses emitted by the light source module into the fiber optic cable to be measured, and guide the returned reflected optical signal to the OTDR detection module.

[0025] Preferably, the OTDR detection module includes a photodetector and a data acquisition system; the photodetector converts the received optical signal into an electrical signal, and measures the time delay and the change in optical power through the data acquisition system.

[0026] Preferably, the data processing module uses the wavelet transform algorithm to denoise the reflected signal and improve the signal quality; then, uses the CNN model to extract features and classify the processed signal, automatically identify the fault type and calculate the fault location.

[0027] Compared with the prior art solutions, the present invention has at least the following beneficial effects:

[0028] 1) The present invention emits optical pulse signals of two different wavelengths, and uses the reflected signals received by the OTDR, combined with an intelligent signal processing algorithm, to accurately calculate the position of the optical fiber fault point. This technology effectively solves the blind area problem of traditional OTDR in short-distance optical fiber detection and improves the fault location accuracy;

[0029] 2) The wavelet transform denoising and convolutional neural network (CNN) fault recognition algorithms are introduced to denoise and enhance the reflected signals received by the OTDR, and automatically identify fault types, such as optical fiber breakage, excessive joint loss, optical fiber bending, etc. This technology reduces the manual analysis workload and improves the detection efficiency;

[0030] 3) Using wavelet denoising + fault point identification can more accurately detect optical fiber faults and improve the maintenance efficiency, which includes the step of setting a threshold T for the fault point; this method combines wavelet analysis and noise estimation to calculate an adaptive threshold (T), which can accurately detect the optical fiber fault point and can be combined with CNN for intelligent classification. Compared with the traditional fixed threshold method, this method is more suitable for complex environments, especially when the signal-to-noise ratio changes greatly. The advantages of applying wavelet transform (db4) to OTDR signal analysis are: it can provide both time-domain and frequency-domain information and is suitable for non-stationary signals; it can effectively separate noise and mutation signals and improve the fault recognition ability; compared with Fourier transform, it is more suitable for mutation detection (such as optical fiber breakage, bending, etc.). BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a schematic structural diagram of the optical fiber fault detection system of the present invention.

[0032] Reference numerals: 100, optical fiber fault detection system; 10, light source module; 20, circulator; 30, optical fiber to be measured; 40, OTDR detection module; 50, data processing module.

[0033] The present invention will be further described in detail below. However, the following examples are only simple examples of the present invention and do not represent or limit the scope of the claimed protection of the present invention. The scope of protection of the present invention shall be subject to the claims. SPECIFIC EMBODIMENTS

[0034] The technical solution of the present invention will be further described below in conjunction with the drawings and specific embodiments.

[0035] As Figure 1 shown, the optical fiber fault detection system 100 provided by the present invention includes a light source module 10, a circulator 20, an optical fiber to be measured 30, an OTDR detection module 40, and a data processing module 50.

[0036] (1) Light Source Module

[0037] The light source module is one of the key components of the system and is responsible for emitting light pulses of two different wavelengths. The light source module uses a tunable pulsed laser, which can adjust the wavelength and pulse width of the emitted light pulses as needed. For different detection requirements, the wavelength selection of the light source module should be determined according to the characteristics of the optical fiber and the receiving range of the OTDR detection module.

[0038] (2) Circulator

[0039] The circulator is used to guide the light pulses emitted by the light source module into the optical fiber under test and direct the returned reflected light signal to the OTDR detection module. The circulator should have low insertion loss and high isolation to ensure the signal quality and transmission efficiency.

[0040] (3) OTDR Detection Module

[0041] The OTDR detection module includes a highly sensitive photodetector and a high-speed data acquisition system. The photodetector converts the received optical signal into an electrical signal, and the data acquisition system measures the time delay and optical power change. This module needs to have high time resolution and high dynamic range to ensure high-precision fault location.

[0042] (4) Data Processing Unit

[0043] The data processing unit uses the wavelet transform algorithm to denoise the reflected signal and improve the signal quality. Then, it uses a convolutional neural network (CNN) to extract features and classify the processed signal, automatically identifying the fault type and calculating the fault location. Through this intelligent algorithm, the system can greatly reduce manual intervention and improve the efficiency and accuracy of fault diagnosis.

[0044] (5) Optical Fiber Under Test

[0045] The optical fiber under test is the object to be detected. Different types of faults may exist in this optical fiber network, such as breaks, excessive splice losses, and fiber bends.

[0046] Example 1: Dual-Wavelength Pulse Fault Location

[0047] Assume that the length of the optical fiber under test is 1000 meters, and the light source module emits light pulses with wavelengths of 1310 nm and 1550 nm. The OTDR detection module records the return times of the light pulses of the two wavelengths respectively. Assume that the return time of the 1310-nm light pulse is 10 microseconds, and the return time of the 1550-nm light pulse is 10.5 microseconds. By calculation, the fault point is 500 meters away from the light source module. This technology effectively solves the blind area problem of traditional OTDR in short-distance optical fiber detection.

[0048] Specific calculation steps: Assume that the light source module emits optical pulses. The optical pulses propagate through the optical fiber. When encountering a fault point, part of the optical pulses are reflected back to the OTDR detection module. The OTDR calculates the distance to the fault point based on the return time of the optical pulses.

[0049] The calculation formula is as follows:

[0050] Distance S = (c * Δt) / 2;

[0051] Where c is the propagation speed of light in the optical fiber (the propagation speed of light in the optical fiber is usually about 2×10 8 m / s, depending on the refractive index of the optical fiber); Δt is the time difference from the emission of the pulse to the return of the reflected pulse (unit: second).

[0052] The division by 2 in the formula is because the optical signal travels from the light source to the fault point and then back, so the time needs to be divided by 2 to obtain the one-way distance.

[0053] Assume that the time difference of the return signal received by the OTDR is 10 microseconds. Then, the distance to the fault point can be calculated as:

[0054] Distance S = ((2×10 8 m / s) × (10×10 -6 s)) / 2 = 500 meters

[0055] This result indicates that the fault point is 500 meters away from the light source module.

[0056] In the above embodiments, 1) Solve the short - distance blind - zone problem: Traditional OTDRs encounter a "blind - zone" problem when measuring short - distance optical fibers. The blind - zone is due to the fact that the propagation time of the optical pulse is too short, and the time difference between the reflected signal and the emitted signal is very small, resulting in an inability to accurately measure faults close to the light - source end. By using dual - wavelength pulses (such as 1310 nm and 1550 nm), optical signals of different wavelengths can propagate separately, and the time difference between them can be used to improve the measurement accuracy at short distances. Because the propagation speeds of light of different wavelengths in the optical fiber are slightly different, this can be used to improve fault location at short distances.

[0057] 2) Improve the positioning accuracy: With two optical pulses of different wavelengths, more information about the characteristics of the optical fiber can be obtained, which makes the calculation of the fault location more accurate. Specifically, the propagation loss and scattering characteristics of light of different wavelengths in the optical fiber are different, which provides more references for data analysis, thus more accurately calculating the fault point.

[0058] 3) Enhanced signal analysis ability: After using the dual-wavelength technology, the system can receive more data. After being processed by intelligent algorithms (such as wavelet transform denoising and CNN fault identification), these data can not only accurately locate the fault point, but also automatically determine the fault type according to the characteristics of the reflected light, such as fiber breakage, connector loss, or excessive bending. Compared with the traditional single-wavelength system, the increased signal processing ability greatly improves the intelligence level of detection and reduces the complexity of manual analysis.

[0059] 4) Enhanced ability to process weak signals: When detecting weak signals or long-distance transmission fibers, the use of dual wavelengths can provide more signal redundancy. Even in weak reflected signals, more information can be extracted through the signal differences of different wavelengths, thereby improving the detection of weak faults. The greatest advantage of the dual-wavelength pulse method lies in improving the accuracy of short-distance fault location, solving the blind zone problem of traditional OTDR in short distances, and being able to perform more accurate fault identification and location through richer signal information. This enables the system to provide higher accuracy and more intelligent analysis when dealing with complex fiber optic networks, especially in the processing of weak signals and fault identification, which is superior to traditional single-wavelength OTDRs. However, the introduction of this technology will also increase the complexity and cost of the system, and pose higher requirements for hardware and algorithms.

[0060] Compared with traditional conventional calculation methods, the dual-wavelength pulse scheme can provide significant improvements in some specific application scenarios (such as short-distance and weak signal detection). Therefore, this calculation method is not completely the same as the conventional calculation method, but has more accurate and intelligent advantages, especially suitable for modern fiber optic network detection capabilities that require high precision and high resolution.

[0061] Embodiment 2

[0062] The OTDR device will receive complex reflected signals during fiber optic fault detection. These signals usually include noise, signal attenuation, scattering effects, and weak fault signals (such as fiber breakage, connector loss, etc.). Traditional OTDR detection methods rely on simple signal analysis (such as peak detection, time difference calculation, etc.). These methods have low accuracy in a strong noise environment, especially limited ability to identify weak signals and complex fault types.

[0063] Therefore, the goal of this embodiment is to use wavelet transform denoising and convolutional neural network (CNN) fault identification algorithms to process the reflected signals received by the OTDR, improve the accuracy of fault location, and automatically identify the fault type.

[0064] Before the convolutional neural network (CNN), it is necessary to process the noise of the OTDR reflection signal first. Traditional OTDR reflection signals contain noise components, which may come from the fluctuations of the light source, the scattering effect of the optical fiber itself, or other external interferences. To improve the accuracy and stability of the CNN model, we use wavelet transform to denoise the signal.

[0065] Steps of wavelet transform denoising:

[0066] S21: Perform wavelet transform on the reflection signal received by OTDR, and use wavelets such as Daubechies wavelet (db4) for signal decomposition.

[0067] S22: After decomposing the signal, remove the high-frequency noise components below a certain threshold in the wavelet coefficients, and retain the main signal features;

[0068] S23: Reconstruct the signal through inverse wavelet transform to obtain a denoised clear signal as the input of the CNN.

[0069] Performing wavelet transform (such as Daubechies wavelet (db4)) on the reflection signal received by OTDR, the main purpose is to perform time-frequency analysis of the signal to extract effective features, remove noise, and improve the accuracy of fault detection. This processing method can effectively reduce the interference of noise on the signal, retain useful fault features, and ensure that the CNN model can identify faults more accurately.

[0070] Wavelet Transform (WT) is a time-frequency analysis method that can provide information about the signal in both the time domain and the frequency domain, and is particularly suitable for processing non-stationary signals (such as OTDR reflection signals). Compared with the traditional Fourier Transform (FFT), wavelet transform can provide different resolutions in different frequency ranges, which helps to identify mutation points in the optical fiber (such as faults like joint loss, fracture, and microbend).

[0071] Daubechies wavelet (db4) is a wavelet basis with good compact support and smoothness, which is suitable for the analysis of optical fiber reflection signals.

[0072] The measurement result of OTDR is a one-dimensional time-series signal, where the abscissa is the propagation time or the optical fiber distance (unit: meter), and the ordinate is the echo signal intensity (unit: dB). Assume the signal collected by OTDR is as follows:

[0073] Distance (m) Reflection intensity (dB) 0 -30 10 -25 20 -40 30 -35 40 -45 50 -60 60 -45 70 -35 80 -30 90 -20 100 -60

[0074] …….

[0075] As can be seen from the above table, a fracture fault (a large drop in the signal) occurred at 100 meters.

[0076] Specific steps of signal decomposition S21:

[0077] S211: Select a wavelet basis function (such as db4);

[0078] S212: Set the decomposition level N (generally select 2 - 4 levels);

[0079] S213: Use discrete wavelet transform (DWT) to decompose the signal into a low - frequency part (approximation coefficients, CA) and a high - frequency part (detail coefficients, CD);

[0080] S214: Continue to decompose the low - frequency part until the set level is reached.

[0081] Among them, CA 2 : The low - frequency approximation component of the second layer, representing the overall trend of the signal; CD 2 : The high - frequency detail component of the second layer, representing medium - frequency components, such as optical fiber joint loss, etc.; CD 1 : The high - frequency detail component of the first layer, representing high - frequency noise, such as environmental interference.

[0082] Select the Daubechies 4 - th order wavelet (db4) as the mother wavelet. Its characteristics are: having good time - frequency localization characteristics; being able to effectively decompose complex signals, suitable for signal denoising and fault point detection.

[0083] Analyzing the results of wavelet decomposition, we can find that:

[0084] The low - frequency part (CA 2 ): Represents the overall trend of the signal, relatively smooth. The high - frequency part (CD 2 , CD 1 ): The position of the fracture fault (100 meters) will show obvious mutation points in CD 2 or CD 1 ; Noise can be removed through threshold processing, and fault characteristics can be highlighted.

[0085] Fault detection method:

[0086] Method 1: Calculate the energy of the detail coefficients. If the energy at a certain point is higher than the set threshold, it may be a fault point:

[0087] Fault point

[0088] If E > T (set threshold), it is considered that there is a fault at this point; where T is the threshold.

[0089] Among them, the process of setting the threshold is as follows:

[0090] Adopt a method that combines the energy analysis of wavelet detail coefficients specially designed for detecting the position of optical fiber faults with noise estimation to calculate the threshold:

[0091] T = μ + β·σ

[0092] Where: μ is the mean value of the detail coefficients (CD 2 ); σ is the standard deviation of the detail coefficients; β is an adjustment coefficient (generally taken as 2 - 4, adjusted according to experiments).

[0093] The steps of the method for calculating the threshold T include: 1. Calculate the mean value and standard deviation of the wavelet detail coefficients (CD 2 ); 2. Set the threshold T, and the points exceeding T are considered as fault points; 3. Applicable to the detection of mutation points and can effectively filter noise.

[0094] The following is a practical example of the calculation of the threshold T in the above optical fiber fault point detection: The threshold is not a fixed value but a dynamically adjustable threshold.

[0095] (1) OTDR reflection signal

[0096] Suppose the optical fiber reflection signal obtained from a certain OTDR test is as follows (unit: dB):

[0097]

[0098] There is an optical fiber break at 100 meters, and the signal suddenly drops to -9.5 dB. The signal is stable at other positions, and the normal signal is about -5 dB.

[0099] (2) Wavelet transform (Daubechies db4)

[0100] Perform Daubechies (db4) wavelet decomposition on the signal to obtain the detail coefficients (CD 2 )

[0101]

[0102] At 100 meters, CD 2 = -3.5, far exceeding other positions, indicating that there may be a fault here; while at other positions, CD 2 changes little and belongs to the normal noise range.

[0103] (3) Calculate the threshold T

[0104] Calculate the mean value and standard deviation of CD 2 :

[0105] μ = mean({0.1, 0.2, 0.1, -0.1, 0.0, -3.5, 0.2, 0.1, 0.0, -0.1, 0.1}) = -0.27

[0106] σ = std({0.1, 0.2, 0.1, -0.1, 0.0, -3.5, 0.2, 0.1, 0.0, -0.1, 0.1}) = 1.06

[0107] Set β = 3 and calculate the threshold:

[0108] T = μ + β·σ = -0.27 + 3 × 1.06 = 2.91

[0109] (4) Determine the fault point

[0110] CD at 100 meters 2 = -3.5, significantly lower than -2.91 (exceeding the threshold), determined as the fault point. The CD of other points 2 are all within the threshold range and belong to normal signals.

[0111] This method combines wavelet analysis and noise estimation to calculate the adaptive threshold (T), which can accurately detect the fiber optic fault point and can be combined with CNN for intelligent classification. Compared with the traditional fixed threshold method, this method is more suitable for complex environments, especially when the signal-to-noise ratio changes greatly.

[0112] Method 2: Use modulus maximum detection: Find the mutation point through the wavelet modulus maximum detection method:

[0113] Calculate the first derivative of the wavelet coefficient; find the local maximum (fault point).

[0114] After wavelet decomposition and analysis, we can observe that:

[0115] 1) Wavelet decomposition successfully extracts the high-frequency anomaly of the fault point (at 100 meters);

[0116] 2) The detail coefficient (CD 2 ) shows the mutation in the signal and can help identify fiber optic loss and breakage;

[0117] 3) Through threshold energy analysis or modulus maximum detection, the fault point can be automatically identified.

[0118] Convolutional neural network (CNN) structure

[0119] Once the signal is processed by wavelet denoising, it can be input into the CNN for feature extraction and fault classification. CNN is a powerful deep learning model, especially suitable for processing signal data with spatial local dependencies, such as time series signals (OTDR reflection signals).

[0120] CNN model architecture:

[0121] 1. Input layer: The input layer receives the OTDR reflection signals denoised by wavelet transform and is usually input into the CNN network in the form of a one-dimensional time series. The signal length can be adjusted according to specific applications, and a common length is several thousand data points.

[0122] 2. Convolutional layer: The convolutional layer contains multiple convolutional kernels for extracting local features in the signal. Each convolutional kernel performs a convolution operation on the time series through a sliding window to identify feature patterns in the signal, such as peaks, reflection points, signal mutations, etc.

[0123] The size of the convolutional kernel can be set to 3×3 or 5×5, etc., the stride is 1, and the number of convolutional kernels is 32 or 64, which can be adjusted as needed.

[0124] The purpose of the convolution operation is to extract important local features from the original signal, such as the peak position and reflection amplitude of the reflected light.

[0125] 3. Activation layer: The ReLU (Rectified Linear Unit) activation function is used to map the output of the convolutional layer to a non-linear space, improving the network's fitting ability for complex signals.

[0126] 4. Pooling layer: The max pooling operation is adopted to reduce the size of the feature map, thereby reducing the computational complexity and helping to prevent overfitting. The commonly used pooling operation is 2×2 max pooling.

[0127] 5. Fully connected layer: The feature maps extracted by the convolutional layer are further processed through the fully connected layer. The fully connected layer flattens the features output by the convolutional layer into a one-dimensional vector and performs a weighted sum through a weight matrix, finally outputting the prediction results of the fault types.

[0128] 6. Output layer: The Softmax activation function is used in the output layer to convert the output of the CNN model into the probabilities of different fault types. Different fault types, such as fiber breakage, excessive joint loss, fiber bending, etc., will correspond to an output probability value indicating the likelihood of that fault type.

[0129] Training and optimization

[0130] To train the CNN model, a large number of OTDR reflection signal samples need to be collected, including normal signals and different types of fault signals. These signal samples are preprocessed (including wavelet transform denoising) and then divided into a training set and a test set.

[0131] During the training process: (1) The Cross-Entropy Loss Function is used to measure the difference between the prediction results and the true labels. (2) The Adam optimizer (Adaptive Moment Estimation) is used to optimize the network parameters and adjust hyperparameters such as the learning rate and batch size. (3) Data Augmentation techniques are used to generate more training samples, for example, enhancing the robustness of the model by translating, scaling, and adding noise along the time axis.

[0132] Fault Identification and Location

[0133] After training is completed, the CNN model can perform real-time analysis on new OTDR reflection signals. After wavelet denoising and feature extraction, the model can automatically identify the fault type in the signal and give the location of the fault.

[0134] Fault Identification Process:

[0135] S1: The OTDR detection module collects the reflection signal.

[0136] S2: Perform wavelet transform denoising on the signal.

[0137] S3: Input the denoised signal into the CNN model for feature extraction.

[0138] S4: The CNN model outputs the fault type and location. For example, if the model identifies an obvious spike in the signal, it may be a fiber break; if a large and flat reflection signal is detected, it may be excessive joint loss.

[0139] By combining CNN and wavelet transform, the fiber optic fault detection system of the present invention has been significantly improved in the following aspects: (1) Accuracy improvement: Through the deep learning ability of CNN, it can automatically learn to extract complex features from the original signal, identify various types of faults, and avoid the limitations of relying on manual analysis and empirical judgment in traditional methods. (2) Noise suppression: Wavelet transform denoising can effectively remove the noise in the reflection signal, especially the common interference signals in the fiber optic system, enhance the signal quality, and provide a cleaner input for CNN. (3) Automatic fault type identification: The CNN model can identify different types of faults (such as fiber breaks, joint losses, etc.) and give the accurate fault location, reducing the workload of manual analysis and improving work efficiency. (4) Processing speed: The parallel computing ability of CNN accelerates the fault diagnosis process, significantly improving the detection speed and meeting the requirements of modern fiber optic networks for real-time detection.

[0140] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various simple modifications can be made to the technical solutions of the present invention, and these simple modifications all fall within the protection scope of the present invention.

[0141] In addition, it should be noted that, in the various specific technical features described in the above specific embodiments, they can be combined in any appropriate manner without conflict. To avoid unnecessary repetition, the present invention will not separately describe various possible combination manners.

[0142] Furthermore, any combination can be made between various different embodiments of the present invention, as long as it does not violate the idea of the present invention, and it should also be regarded as the content disclosed by the present invention.

Claims

1. A method for optical fiber fault detection based on optical time domain reflection, characterized in that: The method comprises the following steps: S1: Collect reflection signals using OTDR detection module; S2: Perform wavelet transform denoising on the signal; S3: Input the denoised signal into the CNN model for feature extraction; S4: CNN model outputs fault type and location; The step S2 specifically includes the following steps: S21: performing wavelet transform on the reflection signal, and using the wavelet to decompose the signal; S22: After decomposing the signal, remove the high-frequency noise components below a certain threshold in the wavelet coefficients and retain the main signal features; S23: Reconstruct the signal through inverse wavelet transform to obtain a clear signal after denoising as the input of the CNN model.

2. The detection method according to claim 1, characterized in that: The step S21 specifically includes the following steps: S211: Select wavelet basis function; S212: Setting the number of decomposition levels N; S213: Decomposing the signal into a low-frequency part and a high-frequency part by using discrete wavelet transform; S214: Continue to decompose the low-frequency part until a set number of layers is reached.

3. The detection method according to claim 2, characterized in that: The detection method of fault point E is as follows: Among them, CD i is the high-frequency component of the i-th layer; when E is greater than the set threshold, it is considered that there is a fault at this point.

4. The detection method according to claim 2, characterized in that: The detection method of the fault point E is as follows: find the mutation point through the wavelet modulus maximum detection method; calculate the first-order derivative of the wavelet coefficient and find the local maximum. The point where the maximum is located is the fault point.

5. The detection method according to claim 2, wherein the step of determining that a fault exists at the point when E is greater than a set threshold value further comprises a step of calculating the threshold value; the threshold value is a dynamically adjustable threshold value.

6. A detection system for implementing any detection method according to claims 1 to 4, characterized in that: The system comprises a light source module, a circulator, an optical fiber to be tested, an OTDR detection module and a data processing module.

7. The detection system according to claim 1, characterized in that: The light source module emits light pulses of two different wavelengths.

8. The detection system according to claim 5, characterized in that: The circulator is used to guide the light pulse emitted by the light source module into the optical fiber to be tested, and guide the returned reflected light signal to the OTDR detection module.

9. The detection system according to claim 5, characterized in that: The OTDR detection module includes a photoelectric detector and a data acquisition system; the photoelectric detector converts the received optical signal into an electrical signal, and measures the time delay and optical power change through the data acquisition system.

10. The detection system according to claim 5, characterized in that: The data processing module uses a wavelet transform algorithm to denoise the reflected signal and improve the signal quality; then, the CNN model is used to extract and classify the processed signal to automatically identify the fault type and calculate the fault location.

Images