Long-distance communication optical cable fault automatic detection and positioning system and method

Abstract

The invention discloses a long-distance communication optical cable fault automatic detection and positioning system and method, and the method comprises the steps: collecting optical signal data in an optical fiber link in real time through the deployment of a multi-optical fiber sensor or a distributed optical fiber sensing technology; denoising processing is carried out on the collected data, abnormal fluctuation is analyzed by applying a self-adaptive algorithm, and the type of a fault which is about to occur is identified; classifying the faults by using a machine learning classification algorithm; time domain reflection and wavelength multiplexing technologies are combined, a composite algorithm is adopted to accurately position a fault point, and path optimization and error correction are carried out through sensing node data comprehensive analysis; an environment compensation algorithm is introduced, influences of environment factors on the optical signals are analyzed in real time, and errors are eliminated; and based on a fault positioning and analysis result, an AI algorithm is adopted to evaluate a fault influence range, a repair scheme is automatically generated, and priorities are scheduled for repair.

Description

Technical Field

[0001] The present invention belongs to the field of automatic detection and location of long-distance communication optical cable faults, and particularly relates to an automatic detection and location system and method for long-distance communication optical cable faults. Background Art

[0002] An automatic detection and location system for long-distance communication optical cable faults is a system used to monitor and locate optical cable faults in a long-distance optical fiber communication network. Its main functions are to monitor the status of the optical fiber link in real time. Once a fault occurs, the system can automatically detect the specific location where the fault occurs and provide accurate location information for maintenance personnel to repair; utilize the characteristics of the optical fiber itself to judge whether there is a problem with the optical fiber link by detecting changes such as attenuation, reflection, and scattering of optical signals; based on the characteristics of the optical fiber link and combined with the signal changes after the fault occurs, accurately locate the position of the fault through algorithms; can collect and analyze the health status of the optical fiber link in real time, send out fault alarms in a timely manner, and transmit the detailed position of the fault to maintenance personnel; optimize the fault detection and location capabilities of the system through continuous data collection and analysis to improve accuracy and response speed.

[0003] However, existing long-distance communication optical cable fault detection and location systems, although able to identify and locate faults in the optical fiber link, still have problems such as insufficient location accuracy, limited fault type identification, and poor system adaptability in some complex network environments. In addition, existing technologies still face certain challenges in dealing with different types of faults, adapting to network topology changes, and real-time fault response. Summary of the Invention

[0004] Aiming at the deficiencies in the prior art, the purpose of the present invention is to provide an automatic detection and location system and method for long-distance communication optical cable faults, aiming to solve the defects in the prior art through innovative technical means and achieve the goals of accurate fault location, diverse fault type identification, strong environmental adaptability, and fast real-time response.

[0005] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0006] An automatic detection and location system for long-distance communication optical cable faults, including:

[0007] An optical signal monitoring module: uses multi-fiber sensors and distributed sensing technology to monitor the optical cable in real time, and detects signal attenuation, time delay, and temperature and humidity changes in the optical fiber;

[0008] A signal processing and fault analysis module: analyzes optical signal data based on an adaptive signal processing algorithm and a machine learning model, automatically identifies the fault type of the optical fiber, and estimates the location where the fault occurs;

[0009] Fault Location Module: Combining time-domain reflectometry technology and wavelength division multiplexing technology, it uses a composite algorithm to optimize the accuracy of fault location, calculates the location of the optical fiber fault point, and in a long-distance and complex topology network environment, improves the location accuracy through a multi-path correction algorithm;

[0010] Intelligent Decision-making and Repair Suggestion Module: According to the fault detection and location results, it combines artificial intelligence algorithms to automatically evaluate the severity of the fault, generates optimized repair plans and priorities, and guides the maintenance personnel to carry out repairs;

[0011] Network Topology Adaptive Module: Through dynamic network topology analysis, this module senses the changes in the network topology in real time to ensure that the fault location system accurately adapts to different network environments.

[0012] Automatic Detection and Location Method for Long-distance Communication Optical Cable Faults, including the following steps:

[0013] By deploying multiple optical fiber sensors or using distributed optical fiber sensing technology, it collects the optical signal data in the optical fiber link in real time, including the attenuation, delay, temperature and humidity change parameters of the optical signal;

[0014] Denoise the collected optical signal data, and use an adaptive algorithm to analyze the abnormal fluctuations in the signal, identify the upcoming fault types, and identify and judge whether to further locate the fault;

[0015] Use a classification algorithm based on machine learning to classify optical fiber faults and identify fault types, including optical fiber breakage, joint faults, connector looseness, and bending damage;

[0016] Combining time-domain reflectometry technology and wavelength division multiplexing technology, it uses a composite algorithm for precise location of the optical fiber fault point. For a long-distance optical fiber link, through comprehensive analysis of the data of multiple sensing nodes, it conducts path optimization and error correction to accurately determine the fault location;

[0017] Introduce an environmental compensation algorithm in the optical fiber network to analyze the impact of environmental changes on optical signals in real time, and identify and eliminate errors caused by environmental factors;

[0018] Based on the fault location and analysis results, use artificial intelligence algorithms to evaluate the scope of influence of the fault, automatically generate repair plans and present them to the maintenance personnel, and schedule according to the priorities of the repair plans.

[0019] Preferably, the method of collecting the optical signal data in the optical fiber link in real time, including the attenuation, delay, temperature and humidity change parameters of the optical signal, by deploying multiple optical fiber sensors or using distributed optical fiber sensing technology is:

[0020] Signal attenuation in an optical fiber refers to the gradual weakening of the optical signal during fiber transmission due to absorption and scattering. Its attenuation rate is expressed in decibels, and the formula is:

[0021]

[0022] α is the fiber attenuation coefficient;

[0023] P in is the input optical power;

[0024] P out is the output optical power;

[0025] L is the fiber length;

[0026] The time delay in an optical fiber refers to the time required for an optical signal to propagate from the light source to the receiver. The time delay is related to the fiber length, the speed of light, and the refractive index in the fiber. The formula is:

[0027]

[0028] t dclay is the signal propagation time delay;

[0029] L is the fiber length;

[0030] v is the propagation speed of the optical signal in the fiber;

[0031] The propagation speed v of the fiber is related to the refractive index n of the fiber. The propagation speed of the fiber is:

[0032]

[0033] c is the speed of light in a vacuum;

[0034] n is the refractive index of the fiber;

[0035] Fiber Bragg grating technology detects temperature and stress changes in an optical fiber by the wavelength change of the reflected optical signal. The temperature and wavelength change formula is:

[0036] Δλ B =λ B ·α T ·ΔT

[0037] Δλ B is the change in the Bragg wavelength;

[0038] λ B is the original Bragg wavelength;

[0039] α T is the temperature sensitivity coefficient;

[0040] ΔT is the temperature change;

[0041] Distributed temperature sensing technology uses optical fiber as a sensing element to measure temperature along the entire length of the optical fiber. The Rayleigh scattering signal formula is:

[0042]

[0043] P R (z) is the Rayleigh scattering signal at position z;

[0044] P in is the power of the input optical signal;

[0045] L0 is the starting length of the optical fiber;

[0046] The relationship between temperature and the intensity of the scattering signal is established through the correlation formula of Brillouin scattering and temperature. The formula is:

[0047] ΔP Brillouin (T) = γ·(T - T0)

[0048] ΔP Brillouin is the change in the Brillouin scattering signal;

[0049] γ is the temperature sensitivity coefficient;

[0050] T is the current temperature;

[0051] T0 is the reference temperature.

[0052] Preferably, the collected optical signal data is denoised, and an adaptive algorithm is used to analyze the abnormal fluctuations in the signal, identify the types of upcoming faults, and identify and determine whether to further locate the faults. The method is:

[0053] Kalman filtering is an optimal filter for estimating and predicting signals with noise. It is based on recursive least squares estimation, and the recurrence formula is:

[0054] Prediction step:

[0055]

[0056] P k = AP k1 A T + Q

[0057] Update step:

[0058] K k = P k H T (HP k H T + R) 1

[0059]

[0060] P k =(I - K k H)P k

[0061] is the estimated state;

[0062] A is the system dynamic matrix;

[0063] B is the control matrix;

[0064] u k is the control input;

[0065] y k is the observed value;

[0066] P k is the covariance matrix;

[0067] Q is the process noise covariance matrix;

[0068] R is the measurement noise covariance matrix;

[0069] H is the measurement matrix;

[0070] After denoising the optical signal, the abnormal fluctuations in the signal are analyzed through an adaptive algorithm. The LMS algorithm is a commonly used adaptive filtering algorithm, and the LMS algorithm formula is:

[0071]

[0072]

[0073] w(n + 1)=w(n)+μx(n)e(n)

[0074] is the estimated signal value

[0075] w(n) is the filter coefficient;

[0076] x(n) is the input signal vector;

[0077] e(n) is the error;

[0078] μ is the step size factor;

[0079] d(n) is the target signal;

[0080] The support vector machine is a supervised learning algorithm. In the fault identification of optical fiber signals, the SVM extracts and classifies the features of the denoised signal to identify different types of faults. The SVM optimization formula:

[0081]

[0082] w is the weight vector of the decision hyperplane;

[0083] b is the bias;

[0084] x i is the input sample;

[0085] y i is the label;

[0086] C is the penalty factor;

[0087] For the abnormal fluctuations of signals, statistical methods are used to detect abnormal points. The Z-Score formula is:

[0088]

[0089] x is the current signal value;

[0090] μ is the mean value of the signal;

[0091] σ is the standard deviation of the signal;

[0092] Once the fault type is identified, the location where the fault occurs can then be determined through location methods. Common fault location methods include optical time domain reflectometer technology and time-delay-based location methods;

[0093] The optical time domain reflectometer locates faults by analyzing the time when the optical signal is reflected in the optical fiber. The OTDR locates faults by sending pulsed light, recording the reflected optical signal, and calculating the time when the optical signal propagates in the optical fiber. The OTDR location formula is:

[0094]

[0095] d is the fault location;

[0096] c is the speed of light;

[0097] t is the time delay when the optical signal returns.

[0098] Preferably, a classification algorithm based on machine learning is used to classify optical fiber faults and identify the fault types. The methods for identifying fiber breakage, joint faults, connector looseness, and bending damage are:

[0099] SVM is a supervised learning algorithm. By constructing a hyperplane, samples of different categories are separated. For the classification of optical fiber faults, different fault types are distinguished through SVM. SVM classifies by finding the hyperplane with the maximum margin. The formula is as follows:

[0100]

[0101] w is the weight vector of the decision hyperplane;

[0102] b is the bias term;

[0103] x i is the feature vector of the i-th sample;

[0104] y i is the class label, indicating which fault type the sample belongs to;

[0105] In practical applications, fiber optic signal data is non-linearly separable. Therefore, a kernel function is used to map the data to a high-dimensional space for linear separability. Commonly used kernel functions include:

[0106] Linear kernel function:

[0107] K(x, x′) = x · x′

[0108] Gaussian kernel function:

[0109]

[0110] Polynomial kernel function:

[0111] K(x, x′) = (x · x′ + 1) d

[0112] where d is the degree of the polynomial and σ is the parameter of the Gaussian kernel;

[0113] The training process includes the following steps:

[0114] Prepare data: Extract the features of various fault signals according to the signal data collected by the OTDR device;

[0115] Data preprocessing: Standardize or normalize the feature data;

[0116] Train the SVM model: Use the training set data to learn the optimal parameters of the SVM;

[0117] Model validation: Verify the generalization ability of the model through the cross-validation method and adjust the parameters to improve the classification accuracy;

[0118] Classification is based on distance metric and is applicable to cases where the similarity of the sample feature space is relatively high. Its basic idea is that for a sample to be classified, find the K nearest neighbors in the feature space and select the most frequent class as the predicted class for this sample. The KNN classification formula is:

[0119]

[0120] where, is the predicted class, y iThe label for the neighbor.

[0121] Preferably, by combining time-domain reflectometry technology with wavelength division multiplexing technology and adopting a composite algorithm for precise positioning of optical fiber fault points. For long-distance optical fiber links, through comprehensive analysis of data from multiple sensing nodes, the method for path optimization and error correction to accurately determine the fault location is as follows:

[0122] Time-domain reflectometry technology is used to detect the fault point of the optical fiber link. It locates the fault by sending optical pulses and measuring the time of the reflected signal. The basic formula for OTDR measurement is:

[0123]

[0124] d is the distance from the fault point to the sensor;

[0125] c is the speed of light in the optical fiber;

[0126] t is the time delay of the signal round trip;

[0127] When performing fault location, multiple sensing nodes will work at different wavelengths. Therefore, by comprehensively analyzing the reflection data of multiple wavelength signals, more information can be obtained;

[0128] Assume there are N reflected signals of different wavelengths, and define the time delay at each wavelength as t n (n = 1, 2,..., N), and the fault distance for each wavelength is d n , then there is a similar formula for the reflected signal of each wavelength:

[0129]

[0130] The fault distance data of multiple sensing nodes are synthesized by the weighted average method. Assume the fault distances obtained from M sensing nodes are d1, d2,..., d M , and the weighting coefficients are w1, w2,..., u M , then the weighted average fault location d avg is calculated by the following formula:

[0131]

[0132] In the case of multiple sensing nodes, the true position of the fault point is determined by minimizing the error between the measured data and the theoretical value. Assume there are M measured values d1, d2,..., d M , and the theoretical fault point is d true , then the goal of the least squares method is to minimize the following error function:

[0133]

[0134] Minimizing this error function gives:

[0135]

[0136] In the process of optical fiber fault location, data from multiple sensing nodes are fused through Kalman filtering, and the system noise is estimated. The state equation of Kalman filtering is:

[0137] x k+1 = Ax k + Bu k + w k

[0138] z k = Hx k + v k

[0139] x k is the state of the system;

[0140] A is the state transition matrix;

[0141] B is the control input matrix;

[0142] u k is the control vector;

[0143] w k is the process noise;

[0144] z k is the measurement value;

[0145] H is the measurement matrix;

[0146] v k is the measurement noise;

[0147] To optimize the path, the goal of path optimization is to select the shortest path and minimize error propagation. A graph-based algorithm is used to calculate the optimal path. The optimized path distance d opt is obtained by the following method:

[0148]

[0149] where d i is the actual distance between nodes, and ∈ is the error correction term.

[0150] Preferably, an environmental compensation algorithm is introduced into the optical fiber network to analyze the influence of environmental changes on optical signals in real time. The method for identifying and eliminating errors caused by environmental factors is:

[0151] The attenuation of optical fiber is mainly affected by environmental factors such as temperature change, humidity change, and mechanical stress. Assuming that the optical fiber attenuation α is affected by temperature T, humidity H, and mechanical stress σ, it is described by the following formula:

[0152] α(T, H, σ) = α0 + Δα T (T) + Δα H (H) + Δα σ (σ)

[0153] Where α0 is the attenuation under standard environmental conditions;

[0154] Δα T (T) is the attenuation change caused by temperature change, usually the influence coefficient of temperature on the optical fiber multiplied by the temperature change amount;

[0155] Δα H (H) is the attenuation change caused by humidity change, usually the influence coefficient of humidity on the optical fiber multiplied by the humidity change amount;

[0156] Δα σ (σ) is the attenuation change caused by mechanical stress, usually the influence coefficient of stress on the optical fiber multiplied by the stress change amount;

[0157] In order to analyze the influence of environmental changes on optical signals in real time and eliminate the errors caused by environmental factors, environmental compensation is achieved through the following steps:

[0158] Monitor environmental factors in real time. Usually, by installing temperature and humidity sensors and stress sensors, the changes of environmental parameters T, H, and σ are obtained in real time;

[0159] When the environmental parameters change, the signal attenuation value of the optical fiber is corrected through the following compensation formula:

[0160] α comp = α measurcd - (Δα T (T) + Δα H (H) + Δα σ (σ))

[0161] Where α mcasured is the optical fiber attenuation value measured in real time by OTDR means;

[0162] α comp is the compensated attenuation value;

[0163] Δα T (T, Δα H (H), Δα σ (σ) is the correction value of the optical fiber attenuation based on the real-time monitored temperature, humidity, and stress environmental parameters;

[0164] Environmental changes are dynamic. Therefore, real-time correction is carried out according to the changing environmental factors. By introducing a dynamic update mechanism, the attenuation model is corrected in real time. Assume that at time t, the temperature, humidity, and stress change, and the following dynamic correction formula is adopted:

[0165] α comp (t) = α mcasurcd (t) - [Δα T (T(t)) + Δα H (H(t)) + Δα σ (σ(t))]

[0166] To eliminate the errors caused by environmental factors, filtering technology is used to identify and correct environmental errors. The Kalman filter formula is:

[0167]

[0168] where, is the estimated attenuation value after environmental correction;

[0169] K k is the Kalman gain;

[0170] z k is the attenuation value measured by OTDR;

[0171] H k is the environmental correction matrix;

[0172] After environmental compensation, the fault location of the optical fiber is corrected by the following formula:

[0173]

[0174] where, d comp is the fault location of the optical fiber after environmental compensation;

[0175] t measurcd is the propagation time of the fault signal initially measured by OTDR;

[0176] Δt cnvironmcnt is the propagation time error caused by environmental changes.

[0177] Preferably, based on the fault location and analysis results, an artificial intelligence algorithm is used to evaluate the influence range of the fault, automatically generate a repair plan and present it to the operation and maintenance personnel. The method for scheduling according to the priority of the repair plan is:

[0178] Assume that the specific location d of the fault is obtained through the optical fiber fault location technology fault, and calculate the affected range based on historical data and real-time monitoring data. Assume that the affected range of the fault can be evaluated by the following formula:

[0179] R impact = f(d fault , α fibcr , ΔT, ΔH, σ strcss )

[0180] where, R impact is the affected range of the fault

[0181] d fault is the specific location where the fault occurs;

[0182] α fibcr is the attenuation coefficient of the optical fiber;

[0183] ΔT, ΔH, σ strcss are the correction factors for the effects of temperature change, humidity change, and mechanical stress on the optical fiber;

[0184] The generation of the repair plan depends on the fault location d fault and the affected range R impact . Different repair plans are generated according to the type of the fault. Assume that the generation of the repair plan is automated through an artificial intelligence algorithm, and the algorithm takes into account the type and location of the fault, the affected range of the fault, and the resource input required for repair;

[0185] Through the AI algorithm, the generated repair plan is represented as a set of plan sets:

[0186] S = {s1, s2, s3,..., s n}

[0187] where, S is all possible repair plans, and s i is the i-th repair plan among them, representing specific repair measures and steps;

[0188] To evaluate the priority of the repair plan, a comprehensive priority scoring model is introduced, considering multiple factors. Assume that each repair plan s i has a priority score P i , and this score can be calculated by the following formula:

[0189] P i = w1·S severity (s i ) + w2·S busincss_impact (s i ) + w3·S rcpair_timc (s i ) - w4·S cost (s i )

[0190] Among them, S scvcrity (s i ) is the severity score of the repair plan s i .

[0191] S busincss_inpact (s i ) is the score of the impact on the business of the repair plan s i .

[0192] S rcpair_timc (s i ) is the time required for the repair plan s i .

[0193] S cost (s i ) is the cost score of the repair plan s i .

[0194] w1, w2, w3, w4 are weight coefficients used to adjust the influence of each factor on the final priority score;

[0195] After generating multiple repair plans and calculating the priority scores, according to the priorities, the repair plans are divided into three degrees. Suppose there are n repair plans, and the operation and maintenance personnel sort the repair tasks according to the priorities. According to the priority sorting algorithm, the scheduling order of the repair plans is:

[0196] Scheduled_order = sort(P1, P2,..., P n )

[0197] Among them, the sort() function sorts the repair plans according to the priority score P i from high to low to obtain the scheduling order;

[0198] According to the scheduling order, the operation and maintenance personnel will start to execute the repair tasks according to the priorities of the repair plans. Suppose the execution progress of the repair tasks can be represented by a status variable Status i :

[0199] Status i = f(P i , T elapscd , resources_used)

[0200] Among them, Status i is the current execution status of the repair plan s i ;

[0201] T clapscd is the time used for the repair task;

[0202] resources_used are the resources consumed for the repair task.

[0203] The beneficial effects of the present invention are as follows:

[0204] Deploying multiple fiber optic sensors or distributed fiber optic sensing technology can collect key signal data in the fiber optic link in real time, including parameters such as optical attenuation, time delay, temperature and humidity, etc. After data collection, denoising technology is used to remove noise interference, and at the same time, the abnormal fluctuations in the signal are analyzed through an adaptive algorithm. Using a classification algorithm based on machine learning, the types of faults can be accurately identified, such as fiber breakage, joint failure, connector looseness, bending damage, etc. Combining time-domain reflectometry (OTDR) and wavelength division multiplexing technology can achieve precise positioning of fiber optic fault points. Especially for long-distance fiber optic links, path optimization and error correction can be carried out by integrating data from multiple sensing nodes. Environmental factors (such as changes in temperature and humidity) will affect the optical signal. By introducing an environmental compensation algorithm, the interference of environmental changes on the optical signal can be analyzed in real time, and the errors caused by environmental factors can be corrected. Based on the fault location and analysis results, the system can automatically generate a repair plan using an artificial intelligence algorithm and present the plan to the operation and maintenance personnel. Through automated fault detection, location, repair plan generation and scheduling, the workload of operation and maintenance personnel is greatly reduced, and they can concentrate on handling more complex tasks. Through multi-level technical support and intelligent fault diagnosis means, the network can maintain efficient and stable operation for a long time. This solution has strong scalability and can support the management of large-scale fiber optic networks. Whether it is a city-level or cross-provincial fiber optic link, unified monitoring, data analysis and fault handling of a large number of sensing nodes can be achieved through this technical architecture, greatly improving the scale and intelligence level of network management. Brief Description of the Drawings

[0205] Figure 1 It is a schematic flow diagram of the automatic detection and location system for long-distance communication optical cable faults of the present invention. Detailed Embodiments

[0206] The principles and features of the present invention are described below. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention. The present invention will be described more specifically by way of example in the following paragraphs. The advantages and features of the present invention will be clearer according to the following description and claims.

[0207] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0208] Embodiment

[0209] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0210] An automatic detection and location system for long-distance communication optical cable faults, comprising:

[0211] Optical signal monitoring module: Using multi-fiber sensors and distributed sensing technology to monitor the optical cable in real time, detecting signal attenuation, time delay, temperature and humidity changes in the optical fiber;

[0212] Signal processing and fault analysis module: Based on adaptive signal processing algorithms and machine learning models, analyze the optical signal data, automatically identify the fault type of the optical fiber, and estimate the location where the fault occurs;

[0213] Fault location module: Combining time-domain reflectometry technology and wavelength division multiplexing technology, using a composite algorithm to optimize the accuracy of fault location, calculating the location of the optical fiber fault point, and in a long-distance and complex topology network environment, improving the location accuracy through a multi-path correction algorithm;

[0214] Intelligent decision-making and repair suggestion module: According to the fault detection and location results, automatically evaluate the severity of the fault in combination with artificial intelligence algorithms, and generate an optimized repair plan and priority to guide the maintenance personnel to carry out repairs;

[0215] Network topology adaptive module: Through dynamic network topology analysis, this module perceives the changes in the network topology in real time to ensure that the fault location system accurately adapts to different network environments.

[0216] An automatic detection and location method for long-distance communication optical cable faults, comprising the following steps:

[0217] By deploying multi-fiber sensors or using distributed optical fiber sensing technology, real-time collection of optical signal data in the optical fiber link, including optical signal attenuation, time delay, temperature and humidity change parameters;

[0218] Denoise the collected optical signal data, and use an adaptive algorithm to analyze the abnormal fluctuations in the signal, identify the types of faults that are about to occur, and identify and judge whether to further locate the fault;

[0219] Use a classification algorithm based on machine learning to classify optical fiber faults, identify fault types, including optical fiber breakage, joint faults, connector looseness, and bending damage;

[0220] Combine time-domain reflectometry technology and wavelength division multiplexing technology, and use a composite algorithm for precise location of the optical fiber fault point. For long-distance optical fiber links, through comprehensive analysis of data from multiple sensing nodes, path optimization and error correction are carried out to accurately determine the fault location;

[0221] Introduce an environmental compensation algorithm in the optical fiber network to analyze the impact of environmental changes on optical signals in real time, identify and eliminate errors caused by environmental factors;

[0222] Based on the fault location and analysis results, use an artificial intelligence algorithm to evaluate the scope of the fault impact, automatically generate a repair plan and present it to the operation and maintenance personnel, and schedule according to the priority of the repair plan.

[0223] The method of collecting optical signal data in the optical fiber link in real time, including the attenuation, time delay, temperature and humidity change parameters of the optical signal, by deploying multiple optical fiber sensors or using distributed optical fiber sensing technology is as follows:

[0224] The signal attenuation in the optical fiber refers to the gradual weakening of the optical signal intensity during the optical fiber transmission process due to absorption and scattering. Its attenuation rate is expressed in decibels, and the formula is:

[0225]

[0226] α is the optical fiber attenuation coefficient; P in is the input optical power; P out is the output optical power; L is the optical fiber length;

[0227] The time delay in the optical fiber refers to the time required for the optical signal to propagate from the light source to the receiver. The time delay is related to the length of the optical fiber, the speed of light, and the refractive index in the optical fiber. The formula is:

[0228]

[0229] t dclay is the signal propagation time delay; L is the optical fiber length; v is the propagation speed of the optical signal in the optical fiber;

[0230] The propagation speed v of the optical fiber is related to the refractive index n of the optical fiber. The propagation speed of the optical fiber is:

[0231]

[0232] c is the speed of light in vacuum; n is the refractive index of the optical fiber;

[0233] The fiber Bragg grating technology detects the temperature and stress changes in the optical fiber by the wavelength change of the reflected optical signal. The temperature and wavelength change formula is:

[0234] Δλ B =λ B ·α T ·ΔT

[0235] Δλ B is the change in the Bragg wavelength; λ B is the original Bragg wavelength; α Tis the temperature sensitivity coefficient; ΔT is the temperature change;

[0236] Distributed temperature sensing technology uses optical fiber as a sensing element to measure temperature along the entire length of the optical fiber. The Rayleigh scattering signal formula is:

[0237]

[0238] P R (z) is the Rayleigh scattering signal at position z; P in is the power of the input optical signal; L0 is the starting length of the optical fiber;

[0239] The relationship between temperature and the scattering signal intensity is established through the correlation formula of Brillouin scattering and temperature. The formula is:

[0240] ΔP Brillouin (T) = γ·(T - T0)

[0241] ΔP Brillouin is the change in the Brillouin scattering signal; γ is the temperature sensitivity coefficient; T is the current temperature; T0 is the reference temperature.

[0242] The method of denoising the collected optical signal data, analyzing the abnormal fluctuations in the signal using an adaptive algorithm, identifying the type of impending failure, and determining whether to further locate the failure is:

[0243] Kalman filtering is an optimal filter for estimating and predicting signals with noise. It is based on recursive least squares estimation, and the recursive formula is:

[0244] Prediction step:

[0245]

[0246] P k = AP k1 A T + Q

[0247] Update step:

[0248] K k = P k H T (HP k H T + R) 1

[0249]

[0250] P k = (I - K k H)P k

[0251] is the estimated state; A is the system dynamic matrix; B is the control matrix; u k is the control input; y k is the observed value; P k is the covariance matrix; Q is the process noise covariance matrix; R is the observation noise covariance matrix; H is the observation matrix;

[0252] After denoising the optical signal, analyze the abnormal fluctuations in the signal through an adaptive algorithm. The LMS algorithm is a commonly used adaptive filtering algorithm, and the LMS algorithm formula is:

[0253]

[0254]

[0255] w(n + 1) = w(n) + μx(n)e(n)

[0256] is the estimated signal value; w(n) is the filter coefficient; x(n) is the input signal vector; e(n) is the error; μ is the step size factor; d(n) is the target signal;

[0257] Support vector machine is a supervised learning algorithm. In the fault identification of optical fiber signals, SVM extracts features and classifies the denoised signals to identify different types of faults. The SVM optimization formula:

[0258]

[0259] w is the weight vector of the decision hyperplane; b is the bias; x i is the input sample; y i is the label; C is the penalty factor;

[0260] For the abnormal fluctuations of the signal, use statistical methods to detect abnormal points. The Z - Score formula is:

[0261]

[0262] x is the current signal value; μ is the mean of the signal; σ is the standard deviation of the signal;

[0263] Once the fault type is identified, the location where the fault occurs can be determined through a location method next. Common fault location methods include optical time domain reflectometer technology and time - delay - based location methods;

[0264] The optical time domain reflectometer locates faults by analyzing the time when the optical signal is reflected in the optical fiber. The OTDR locates faults by sending pulsed light, recording the reflected optical signal, and calculating the time when the optical signal propagates in the optical fiber. The OTDR location formula is:

[0265]

[0266] d is the fault location; c is the speed of light; t is the time delay for the optical signal to return.

[0267] Through the Kalman filter and LMS algorithm, the noise in the optical signal can be removed and the signal quality can be improved; through Z-Score for anomaly fluctuation detection, possible faults can be detected in a timely manner; the Support Vector Machine (SVM) algorithm can accurately classify different types of faults, reducing misdiagnosis and missed diagnosis; through the OTDR technology, the fault location in the optical fiber link can be quickly and accurately located, improving the fault response speed and maintenance efficiency; this solution can process optical fiber signals in real time and give fault warnings, discovering potential problems in advance and preventing the expansion of faults; this solution realizes the full-automatic processing from signal acquisition to fault diagnosis and location, improving the efficiency and intelligence level of optical fiber maintenance.

[0268] The method for classifying optical fiber faults and identifying fault types, including fiber breakage, joint faults, connector looseness, and bending damage, by using a classification algorithm based on machine learning is as follows:

[0269] SVM is a supervised learning algorithm. By constructing a hyperplane to separate samples of different classes, for the classification of optical fiber faults, different fault types are distinguished by SVM. SVM classifies by finding the hyperplane with the maximum margin, and the formula is as follows:

[0270]

[0271] w is the weight vector of the decision hyperplane; b is the bias term; x i is the feature vector of the i-th sample; y i is the class label, indicating which fault type the sample belongs to;

[0272] In practical applications, the optical fiber signal data is non-linearly separable. Therefore, the kernel function is used to map the data to a high-dimensional space for linear separability. Commonly used kernel functions include:

[0273] Linear kernel function:

[0274] K(x, x′) = x · x′

[0275] Gaussian kernel function:

[0276]

[0277] Polynomial kernel function:

[0278] K(x, x′) = (x · x′ + 1) d

[0279] where d is the degree of the polynomial and σ is the parameter of the Gaussian kernel;

[0280] The training process includes the following steps:

[0281] Prepare data: Extract the features of various fault signals from the signal data collected by the OTDR device;

[0282] Data preprocessing: Standardize or normalize the feature data;

[0283] Train the SVM model: Use the training set data to learn the optimal parameters of the SVM;

[0284] Model validation: Verify the generalization ability of the model through the cross-validation method and adjust the parameters to improve the classification accuracy;

[0285] Classification based on distance metric is applicable to the case where the similarity of the sample feature space is relatively high. Its basic idea is that for a sample to be classified, find the K nearest neighbors in the feature space and select the most frequent class as the predicted class for this sample. The KNN classification formula is:

[0286]

[0287] where is the predicted class and y i is the label of the neighbor.

[0288] Combining time-domain reflectometry technology with wavelength division multiplexing technology, a composite algorithm is used for precise positioning of fiber optic fault points. For long-distance fiber optic links, through comprehensive analysis of data from multiple sensing nodes, path optimization and error correction are carried out. The method for accurately determining the fault location is:

[0289] Time-domain reflectometry technology is used to detect the fault point of the fiber optic link. It locates the fault by sending optical pulses and measuring the time of the reflected signal. The basic formula for OTDR measurement is:

[0290]

[0291] d is the distance from the fault point to the sensor; c is the speed of light in the fiber; t is the time delay of the signal round trip;

[0292] When performing fault location, multiple sensing nodes will work at different wavelengths. Therefore, by comprehensively analyzing the reflection data of multiple wavelength signals, more information can be obtained;

[0293] Assume there are N reflection signals at different wavelengths, and define the time delay at each wavelength as t n (n = 1, 2,..., N), and the fault distance at each wavelength is d n, there is a similar formula for the reflected signal of each wavelength:

[0294]

[0295] The fault distance data of multiple sensing nodes are synthesized by the weighted average method. Assuming that the fault distances obtained from M sensing nodes are d1, d2,..., d M , and the weighting coefficients are w1, w2,..., w M , then the weighted average fault position d avg is calculated by the following formula:

[0296]

[0297] In the case of multiple sensing nodes, the true position of the fault point is determined by minimizing the error between the measured data and the theoretical value. Assuming there are M measured values d1, d2,..., d M , and the theoretical fault point is d truc , then the goal of the least squares method is to minimize the following error function:

[0298]

[0299] Minimizing this error function gives:

[0300]

[0301] In the process of optical fiber fault location, the data of multiple sensing nodes are fused by Kalman filtering, and the system noise is estimated. The state equation of Kalman filtering is:

[0302] x k+1 = Ax k + Bu k + w k

[0303] z k = Hx k + v k

[0304] x k is the state of the system; A is the state transition matrix; B is the control input matrix; u k is the control vector; w k is the process noise; z k is the measured value; H is the measurement matrix; v k is the measurement noise;

[0305] To optimize the path, the goal of path optimization is to select the shortest path and minimize the error propagation. A graph-based algorithm is used to calculate the optimal path. The optimized path distance d optobtained by the following means:

[0306]

[0307] where d i is the actual distance between nodes, and ∈ is the error correction term.

[0308] By combining time-domain reflectometry technology with multi-wavelength signals and comprehensively analyzing the data of multiple sensing nodes, more accurate fault location results can be obtained. The combination of the weighted average method, the least squares method, and Kalman filtering makes the positioning process more accurate and robust; the reflection data of multiple wavelength signals can provide more comprehensive information, helping to overcome the errors that may occur with a single wavelength and improving the stability of fault location; by minimizing the error between the measured data and the theoretical value and combining with a path optimization algorithm, error propagation is reduced, thereby improving the accuracy of fault location. Path optimization not only selects the shortest path but also considers error correction, ensuring optimal performance during data transmission; Kalman filtering improves the overall performance of the system by fusing the data of multiple sensing nodes and estimating and correcting noise, ensuring the reliability of fault location; this method is particularly suitable for long-distance optical fiber links and can gradually correct the positioning error through the collaborative work of multiple nodes, ensuring accurate fault location even over long distances; due to the combination of multi-wavelength and composite algorithms, the system can flexibly adapt to different network environments and actual application requirements and has strong versatility.

[0309] The method of introducing an environmental compensation algorithm into the optical fiber network to analyze the impact of environmental changes on optical signals in real time and identify and eliminate errors caused by environmental factors is as follows:

[0310] Optical fiber attenuation is mainly affected by environmental factors such as temperature change, humidity change, and mechanical stress. Assuming that the optical fiber attenuation α is affected by temperature T, humidity H, and mechanical stress σ, it is described by the following formula:

[0311] α(T, H, σ) = α0 + Δα T (T) + Δα H (H) + Δα σ (σ)

[0312] where α0 is the attenuation under standard environmental conditions;

[0313] Δα T (T) is the attenuation change caused by temperature change, usually the influence coefficient of temperature on the optical fiber multiplied by the temperature change amount;

[0314] Δα H (H) is the attenuation change caused by humidity change, usually the influence coefficient of humidity on the optical fiber multiplied by the humidity change amount;

[0315] Δα σ (σ) is the attenuation change caused by mechanical stress, usually the influence coefficient of stress on the optical fiber multiplied by the stress change amount;

[0316] To analyze the influence of environmental changes on optical signals in real time and eliminate the errors caused by environmental factors, environmental compensation is achieved through the following steps:

[0317] Monitor environmental factors in real time. Usually, by installing temperature and humidity sensors and stress sensors, the changes of environmental parameters T, H, and σ are obtained in real time;

[0318] When the environmental parameters change, the signal attenuation value of the optical fiber is corrected through the following compensation formula:

[0319] α comp = α measurcd -(Δα T (T)+Δα H (H)+Δα σ (σ))

[0320] Among them, α mcasurcd is the optical fiber attenuation value measured in real time by OTDR means;

[0321] α comp is the compensated attenuation value;

[0322] Δα T (T), Δα H (H), Δα σ (σ) are the correction values of the optical fiber attenuation based on the real-time monitored temperature, humidity, and stress environmental parameters;

[0323] Since environmental changes are dynamic, real-time correction is made according to the changing environmental factors. By introducing a dynamic update mechanism, the attenuation model is corrected in real time. Assuming that at time t, the temperature, humidity, and stress change, the following dynamic correction formula is adopted:

[0324] α comp (t) = α measurcd (t)-[Δα T (T(t))+Δα H (H(t))+Δα σ (σ(t))]

[0325] To eliminate the errors caused by environmental factors, filtering technology is used to identify and correct environmental errors. The Kalman filter formula is:

[0326]

[0327] Among them, is the estimated attenuation value after environmental correction;

[0328] K k is the Kalman gain;

[0329] z k is the attenuation value obtained by OTDR measurement;

[0330] H k is the environmental correction matrix;

[0331] After environmental compensation, the fault location of the optical fiber is corrected by the following formula:

[0332]

[0333] where d comp is the fault location of the optical fiber after environmental compensation;

[0334] t mcasured is the propagation time of the fault signal obtained preliminarily by OTDR measurement;

[0335] Δt cnvironmcnt is the propagation time error caused by environmental changes.

[0336] This solution can ensure accurate compensation of optical fiber attenuation under changing environmental conditions and provide real-time and accurate optical fiber fault location by monitoring changes in environmental factors such as temperature, humidity, and mechanical stress in real time and dynamically correcting the attenuation model; environmental factors have the characteristic of dynamic change, and the solution can adapt to different environmental changes in real time through a dynamic update mechanism and the Kalman filtering algorithm; through filtering techniques such as the Kalman filter, errors caused by environmental factors can be effectively identified and excluded, reducing the interference of noise and ensuring more accurate fault location. This is especially important in complex environments, avoiding the influence of environmental factors on the positioning results; by compensating for the influence of environmental factors, the fault location of the optical fiber can be corrected more precisely; this solution makes optical fiber maintenance more precise, reduces misjudgments caused by environmental changes, and improves the overall operation and maintenance efficiency of the optical fiber network.

[0337] Based on the fault location and analysis results, the method for using an artificial intelligence algorithm to evaluate the influence range of the fault, automatically generate a repair plan and present it to the operation and maintenance personnel, and schedule according to the priority of the repair plan is as follows:

[0338] Assume that the specific location d of the fault is obtained through the optical fiber fault location technology fault , and the influence range is calculated based on historical data and real-time monitoring data. Assume that the influence range of the fault can be evaluated by the following formula:

[0339] R impact = f(d fault, α fibcr , ΔT, ΔH, σ strcss )

[0340] Among them, R impact is the influence range of the fault

[0341] d fault is the specific location where the fault occurs;

[0342] α fiber is the attenuation coefficient of the optical fiber;

[0343] ΔT, ΔH, σ strcss are the correction amounts for the influence of temperature change, humidity change, and mechanical stress on the optical fiber;

[0344] The generation of the repair plan depends on the fault location d fault and the influence range R impact . Different repair plans are generated according to the type of the fault. It is assumed that the generation of the repair plan is automated through an artificial intelligence algorithm, and the algorithm considers the type and location of the fault, the influence range of the fault, and the resource input required for repair;

[0345] Through the AI algorithm, the generated repair plan is represented as a set of plan sets:

[0346] S = {s1, s2, s3,..., s n}

[0347] Among them, S is all possible repair plans, and s i is the i-th repair plan, representing specific repair measures and steps;

[0348] In order to evaluate the priority of the repair plan, a comprehensive priority scoring model is introduced, considering multiple factors. It is assumed that each repair plan s i has a priority score P i , and this score can be calculated by the following formula:

[0349] P i = w1·S severity (s i ) + w2·S busincss_impact (s i ) + w3·S rcpair_timc (s i ) - w4·S cost (s i )

[0350] Among them, S scvcrity (s i ) is the severity score of the repair plan s i ;

[0351] Sbusincss_impact (s i ) is the repair plan s i The score of the impact on the business;

[0352] S rcpair_timc (s i ) is the repair plan s i The time required;

[0353] S cost (s i ) is the repair plan s i The cost score of;

[0354] w1, w2, w3, w4 are weight coefficients used to adjust the impact of each factor on the final priority score;

[0355] After generating multiple repair plans and calculating the priority scores, the repair plans are scheduled according to the priority. Assuming there are n repair plans, the operation and maintenance personnel sort the repair tasks according to the priority. According to the priority sorting algorithm, the scheduling order of the repair plans is:

[0356] Scheduled_order = sort(P1, P2,..., P n )

[0357] Among them, the sort() function sorts the repair plans according to the priority score P i from high to low to obtain the scheduling order;

[0358] According to the scheduling order, the operation and maintenance personnel will start to execute the repair tasks according to the priority of the repair plans. Assuming that the execution progress of the repair tasks can be represented by a status variable Status i :

[0359] Status i = f(P i , T clapscl , resources_used)

[0360] Among them, Status i is the current execution status of the repair plan s i ;

[0361] T clapscd is the time used for the repair task;

[0362] resources_used is the resources consumed by the repair task.

[0363] This solution automatically evaluates the impact of faults, generates repair plans through artificial intelligence algorithms, and schedules them according to priority scores; by introducing a comprehensive priority scoring model and considering multiple factors (severity, business impact, time cost, and cost), it ensures a more scientific and reasonable prioritization of repair plans; through automated repair plan scheduling and execution progress management, it can ensure a quick response to faults and optimize resource allocation; the evaluation of resource consumption and repair time in the repair plan enables more accurate resource management during the execution of repair tasks, avoids resource waste, and ensures that each task can be completed on time; through priority scheduling and efficient repair, it can reduce the operation interruption time caused by faults and ensure the stable operation of the optical fiber network; the solution uses historical data and real-time monitoring data to continuously optimize the generation process of fault location and repair plans, and the AI algorithm can improve decision-making quality through continuous learning and further optimize the fault repair process.

[0364] This embodiment also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the long-distance communication optical cable fault automatic detection and location system and method as described above.

[0365] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the long-distance communication optical cable fault automatic detection and location system and method as described above.

[0366] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0367] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

[0368] The above embodiments of the present invention do not limit the protection scope of the present invention. The implementation manners of the present invention are not limited thereto. All kinds of modifications, substitutions, or changes made to the above structure of the present invention according to the above content of the present invention, in accordance with the common general knowledge and conventional means in the art, without departing from the above basic technical idea of the present invention, shall fall within the protection scope of the present invention.

Claims

1. An automatic detection and positioning system for long-distance communication optical cable faults, characterized in that, It includes: Optical signal monitoring module: Using multi-fiber sensors and distributed sensing technology to monitor the optical cable in real time, detecting signal attenuation, time delay, temperature and humidity changes in the optical fiber; Signal processing and fault analysis module: Based on adaptive signal processing algorithms and machine learning models, analyze optical signal data, automatically identify the fault types of optical fibers, and estimate the location where the fault occurs; Fault location module: Combining time-domain reflectometry technology and wavelength multiplexing technology, using a composite algorithm to optimize the accuracy of fault location, calculate the location of the optical fiber fault point, and in a long-distance and complex topology network environment, improve the location accuracy through a multi-path correction algorithm; Intelligent decision-making and repair suggestion module: According to the fault detection and location results, automatically evaluate the severity of the fault by combining artificial intelligence algorithms, and generate optimized repair plans and priorities to guide the maintenance personnel for repair; Network topology adaptive module: Through dynamic network topology analysis, this module perceives the changes in the network topology in real time to ensure that the fault location system accurately adapts to different network environments.

2. Automatic detection and location method for long-distance communication optical cable faults, characterized in that It includes the following steps: By deploying multi-fiber sensors or using distributed fiber sensing technology, real-time collect the optical signal data in the optical fiber link, including the attenuation, time delay, temperature and humidity change parameters of the optical signal; Perform denoising processing on the collected optical signal data, and use an adaptive algorithm to analyze the abnormal fluctuations in the signal, identify the upcoming fault types, and identify and judge whether to further locate the fault; Use a classification algorithm based on machine learning to classify optical fiber faults, identify fault types, including fiber breakage, joint faults, connector looseness, and bending damage; Combine time-domain reflectometry technology and wavelength multiplexing technology, and use a composite algorithm to accurately locate the optical fiber fault point. For long-distance optical fiber links, through comprehensive analysis of the data of multiple sensing nodes, perform path optimization and error correction to accurately determine the fault location; Introduce an environmental compensation algorithm in the optical fiber network, analyze the impact of environmental changes on optical signals in real time, and identify and eliminate errors caused by environmental factors; Based on the fault location and analysis results, use artificial intelligence algorithms to evaluate the scope of influence of the fault, automatically generate a repair plan and present it to the maintenance personnel, and schedule according to the priority of the repair plan.

3. The automatic fault detection and location method for long-distance communication optical cables according to claim 2, characterized in that, The method of real-time collecting the optical signal data in the optical fiber link by deploying multi-fiber sensors or using distributed fiber sensing technology is: The signal attenuation in the optical fiber refers to the gradual weakening of the optical signal intensity during the optical fiber transmission process due to absorption and scattering. Its attenuation rate is expressed in decibels, and the formula is: α is the optical fiber attenuation coefficient; P in is the input optical power; P out is the output optical power; L is the optical fiber length; The time delay in the optical fiber refers to the time required for the optical signal to propagate from the light source to the receiver. The time delay is related to the length of the optical fiber, the speed of light, and the refractive index in the optical fiber. The formula is: t dclay is the signal propagation delay; L is the optical fiber length; v is the propagation speed of the optical signal in the optical fiber; The propagation speed v of the optical fiber is related to the refractive index n of the optical fiber. The propagation speed of the optical fiber is: c is the speed of light in a vacuum; n is the refractive index of the optical fiber; Fiber Bragg grating technology detects temperature and stress changes in optical fibers by the wavelength change of reflected optical signals. The formula for temperature and wavelength change is as follows: △λ B = λ B · α T · △T Δλ B is the change in the Bragg wavelength; λ B is the original Bragg wavelength; α T is the temperature sensitivity coefficient; ΔT is the temperature change; Distributed temperature sensing technology uses optical fiber as a sensing element to measure temperature along the entire length of the optical fiber. The Rayleigh scattering signal formula is as follows: P R (z) is the Rayleigh scattering signal at position z; P in is the power of the input optical signal; L0 is the starting length of the optical fiber; The relationship between temperature and scattering signal intensity is established through the correlation formula of Brillouin scattering and temperature. The formula is as follows: △P Brillouin (T) = γ·(T - T0) ΔP Brillouin is the change in the Brillouin scattering signal; γ is the temperature sensitivity coefficient; T is the current temperature; T0 is the reference temperature.

4. The automatic fault detection and location method for long-distance communication optical cables according to claim 3, characterized in that, The method for denoising the collected optical signal data, analyzing abnormal fluctuations in the signal using an adaptive algorithm, identifying the type of impending fault, and determining whether to further locate the fault is as follows: Kalman filter is an optimal filter for estimating and predicting signals with noise. It is based on recursive least squares estimation, and the recursive formula is as follows: Prediction step: P k = AP k-1 A T + Q Update step: K k = P k H T (HP k H T + R) T P k = (I - K k H)P k For an estimated state; A is the system dynamic matrix; B is the control matrix; u k For controlling the input; y k is the observed value; P k is the covariance matrix; Q is the process noise covariance matrix; R is the observation noise covariance matrix; H is the observation matrix; After denoising the optical signal, analyze the abnormal fluctuations in the signal through an adaptive algorithm. The LMS algorithm is a commonly used adaptive filtering algorithm. The LMS algorithm formula is as follows: w(n + 1) = w(n) + μx(n)e(n) To estimate the signal value w(n) is the filter coefficient; x(n) is the input signal vector; e(n) is the error; μ is the step size factor; d(n) is the target signal; Support vector machine is a supervised learning algorithm. In the fault identification of optical fiber signals, SVM extracts features and classifies the denoised signals to identify different types of faults. The SVM optimization formula: w is the weight vector of the decision hyperplane; b is the bias; x i is the input sample; y i is a label; C is the penalty factor; For the abnormal fluctuations of the signal, statistical methods are used to detect abnormal points. The Z-Score formula is as follows: x is the current signal value; μ is the mean of the signal; σ is the standard deviation of the signal; Once the fault type is identified, the location where the fault occurs can be determined through a location method next. Common fault location methods include optical time domain reflectometer technology and time-delay-based location methods; The optical time domain reflectometer locates faults by analyzing the time when the optical signal is reflected in the optical fiber. OTDR locates faults by sending pulsed light, recording the reflected optical signal, and calculating the time when the optical signal propagates in the optical fiber. The OTDR location formula is as follows: d is the fault location; c is the speed of light; t is the time delay when the optical signal returns.

5. The automatic detection and location method for long-distance communication optical cable faults according to claim 4, characterized in that The method for classifying optical fiber faults using a machine learning-based classification algorithm to identify fault types, including fiber breakage, joint faults, connector looseness, and bending damage, is as follows: SVM is a supervised learning algorithm. By constructing a hyperplane to separate samples of different classes, for the classification of optical fiber faults, SVM distinguishes different fault types. SVM classifies by finding the hyperplane with the maximum margin. The formula is as follows: w is the weight vector of the decision hyperplane; b is the bias term; x i is the feature vector of the i-th sample; y i It is a class label indicating which fault type the sample belongs to; In practical applications, fiber optic signal data is non-linearly separable. Therefore, a kernel function is used to map the data into a high-dimensional space for linear separability. Commonly used kernel functions include: Linear kernel function: K(x,x') = x·x′ Gaussian kernel function: Polynomial kernel function: K(x,x′) = (x·x′ + 1) d where d is the degree of the polynomial and σ is the parameter of the Gaussian kernel; The training process includes the following steps: Prepare data: Extract the characteristics of various fault signals based on the signal data collected by the OTDR device; Data preprocessing: Standardize or normalize the feature data; Train the SVM model: Use the training set data to learn the optimal parameters of the SVM; Model verification: Verify the generalization ability of the model through the cross-validation method and adjust the parameters to improve the classification accuracy; Classification based on distance metric is applicable to the case where the similarity of the sample feature space is relatively high. Its basic idea is that for a sample to be classified, find the K nearest neighbors of it in the feature space and select the most frequent class as the predicted class of this sample. The KNN classification formula is: Among them, is the predicted category, y i is the label of the neighbor.

6. The automatic detection and location method for long-distance communication optical cable faults according to claim 5, characterized in that, Combining time-domain reflectometry technology and wavelength division multiplexing technology, a composite algorithm is used for precise positioning of fiber optic fault points. For long-distance fiber optic links, through comprehensive analysis of data from multiple sensing nodes, path optimization and error correction are carried out. The method for accurately determining the fault location is: Time-domain reflectometry technology is used to detect the fault points of the fiber optic link. It locates the fault by sending optical pulses and measuring the time of the reflected signal. The basic formula for OTDR measurement is: d is the distance from the fault point to the sensor; c is the speed of light in the fiber; t is the time delay of the signal round trip; When performing fault location, multiple sensing nodes will work at different wavelengths. Therefore, by comprehensively analyzing the reflection data of multiple wavelength signals, more information can be obtained; Assume there are N reflected signals with different wavelengths, and define the time delay at each wavelength as t n (n = 1, 2,..., N), and the fault distance for each wavelength is d n , then there is a similar formula for the reflected signal of each wavelength: The fault distance data of multiple sensing nodes are synthesized by the weighted average method. Assume that the fault distances obtained from M sensing nodes are d1, d2,..., d M , and the weighting coefficients are w1, w2,..., w M , then the fault location d avg is calculated by the following formula: In the case of multiple sensing nodes, the true location of the fault point is determined by minimizing the error between the measured data and the theoretical value. Suppose there are M measured values d1, d2,..., d M , and the theoretical fault point is d truc . Then the goal of the least squares method is to minimize the following error function: Minimizing this error function gives: During the process of fiber optic fault location, data from multiple sensing nodes is fused through Kalman filtering, and the system noise is estimated. The state equation of Kalman filtering is: x k+1 = Ax k + Bu k + w k z k = Hx k + v k x k is the status of the system; A is the state transition matrix; B is the control input matrix; u k is the control vector; w k is the process noise; z k is the measured value; H is the measurement matrix; v k for measuring noise; To optimize the path, the goal of path optimization is to select the shortest path and minimize error propagation. A graph-based algorithm is used to calculate the optimal path, and the distance d of the optimized path opt is obtained by the following method: where d i is the actual distance between nodes, and ∈ is the error correction term.

7. The automatic detection and location method for long-distance communication optical cable faults according to claim 6, characterized in that The method for introducing an environmental compensation algorithm in the fiber optic network to analyze the impact of environmental changes on optical signals in real time and identify and eliminate errors caused by environmental factors is: Fiber optic attenuation is mainly affected by environmental factors such as temperature change, humidity change, and mechanical stress. Assuming that the fiber optic attenuation α is affected by temperature T, humidity H, and mechanical stress σ, it is described by the following formula: α(T, H, σ) = α0 + △α T (T + △α H (H) + △α σ (σ) where α0 is the attenuation under standard environmental conditions; Δα T (T) is the attenuation change caused by temperature change, usually the influence coefficient of temperature on the optical fiber multiplied by the temperature change amount; Δα H (H) is the attenuation change caused by humidity change, usually the influence coefficient of humidity on the optical fiber multiplied by the humidity change amount; Δα σ (σ) is the attenuation change caused by mechanical stress, usually the influence coefficient of stress on the optical fiber multiplied by the stress change amount; In order to analyze the impact of environmental changes on optical signals in real time and eliminate errors caused by environmental factors, environmental compensation is achieved through the following steps: Monitor environmental factors in real time. Usually, by installing temperature and humidity sensors and stress sensors, the changes of environmental parameters T, H, σ are obtained in real time; When the environmental parameters change, the signal attenuation value of the fiber optic is corrected through the following compensation formula: α comp = α measured - (Δα T (T) + Δα H (H) + Δα σ (σ)) Among them, α mcasured is the optical fiber attenuation value measured in real time by OTDR; σ comp is the compensated attenuation value; Δα T (T), Δα H (H), Δα σ (σ) is the correction value of the optical fiber attenuation based on the temperature, humidity, and stress environment parameters monitored in real time; Environmental changes are dynamic. Therefore, according to the changing environmental factors, real-time correction is carried out. By introducing a dynamic update mechanism, the attenuation model is corrected in real time. Assuming that at time t, the temperature, humidity, and stress change, the following dynamic correction formula is adopted: α comp (t) = measured (t) - [Δα T (T(t)) + △α H (H(t)) + △α σ (σ(t))] To exclude the errors caused by environmental factors, filtering technology is used to identify and correct environmental errors. The Kalman filter formula is as follows: Among them, is the estimated attenuation value after environmental correction; K k is the Kalman gain; z k is the attenuation value obtained by OTDR measurement; H k is an environmental correction matrix; After environmental compensation, the fault location of the optical fiber is corrected by the following formula: where d comp is the optical fiber fault location after environmental compensation; t measured is the propagation time of the fault signal initially obtained by OTDR measurement; Δt cnvironment is the propagation time error caused by environmental changes.

8. The automatic detection and location method for long-distance communication optical cable faults according to claim 7, characterized in that Based on the fault location and analysis results, an artificial intelligence algorithm is used to evaluate the influence range of the fault, automatically generate a repair plan and present it to the operation and maintenance personnel. The method for scheduling according to the priority of the repair plan is as follows: Suppose the specific location d of the fault is obtained by the fault location technology through the optical fiber fault , and the affected range is calculated based on historical data and real-time monitoring data. Suppose the affected range of the fault can be evaluated by the following formula: R impact = f(d fault , α fiber , ΔT, △H, σ stress ) Among them, R impact is the influence range of the fault d fault is the specific location where the fault occurs; α fiber is the attenuation coefficient of the optical fiber; ΔT, ΔH, σ stress are the correction amounts for the effects of temperature change, humidity change, and mechanical stress on the optical fiber; The generation of the repair plan depends on the fault location d fault and the influence range R impact , and different repair plans are generated according to the type of the fault. Assume that the generation of the repair plan is an automated decision through an artificial intelligence algorithm, and the algorithm considers the type and location of the fault, the range affected by the fault, and the resource input required for the repair; Through the AI algorithm, the generated repair plan is represented as a set of plan sets: S = {s1, s2, s3, ..., s n} Among them, S is all possible repair solutions, and s i is the i-th repair solution among them, representing specific repair measures and steps; To evaluate the priority of repair solutions, a comprehensive priority scoring model is introduced, considering multiple factors. Assume that each repair solution s i has a priority score P i , and this score can be calculated by the following formula: P i = w1·Ss everity (s i ) + w2·S business_impact (s i ) + w3·S repair_time (s i ) - w4·S cost (s i ) Among them, S scvcrity (s i ) is the severity score of the repair solution s i ; S busincss_impact (s i ) is the repair plan s i The score of the impact on the business; S rcpair_time (s i ) is the time required for the repair solution s i ; S cost (s i ) is the cost score of the repair plan s i ; w1, w2, w3, w4 are weight coefficients used to adjust the influence of each factor on the final priority score; After generating multiple repair plans and calculating the priority scores, the repair plans are scheduled according to the priority. Assuming there are n repair plans, the operation and maintenance personnel sort the repair tasks according to the priority. According to the priority sorting algorithm, the scheduling order of the repair plans is as follows: Scheduled order = sort(P1, P2, ..., P n ) Among them, the sort() function sorts the repair solutions according to the priority score P i from high to low to obtain the scheduling order; According to the scheduling order, the operation and maintenance personnel will start to execute the repair tasks according to the priority of the repair plan. Assume that the execution progress of the repair tasks can be represented by a status variable Status i : Status i = f(P, T clapsed , resources_used) Among them, Status i is the current execution status of the repair plan s i ; T clapscd Elapsed time for the repair task; resources_used is the resources consumed by the repair task.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the automatic detection and location method for long-distance communication optical cable faults as described in any one of claims 2-8.

10. A computer-readable storage medium, on which a computer program is stored, characterized in that, When the program is executed by the processor, it implements the automatic detection and location method for long-distance communication optical cable faults as described in any one of claims 2-8.

Images