Optical fiber communication equipment monitoring method and system based on big data, and medium

Abstract

The invention provides an optical fiber communication equipment monitoring method and system based on big data and a medium, and belongs to the technical field of optical fiber communication and big data. The method comprises the following steps: synchronously acquiring operating parameters and environmental parameters of optical fiber equipment through a multi-source sensor network, and generating a standard parameter information matrix through space-time alignment and noise suppression; further extracting characteristics such as an optical signal stability index and vibration energy anomaly through time-frequency domain conjoint analysis, and constructing an equipment health state matrix; and predicting a degradation process based on a bidirectional LSTM model of an attention mechanism, and outputting a visual diagnosis result in combination with a dynamic threshold grading alarm mechanism. According to the scheme, the problems of high false alarm rate and insufficient prediction capability of a traditional monitoring method are solved, the false alarm rate is reduced, and the fault prediction accuracy is improved.

Description

Technical Field

[0001] This application relates to the fields of optical fiber communication and big data technology. Specifically, it relates to a method, system, and medium for monitoring optical fiber communication devices based on big data. Background Art

[0002] Traditional optical fiber communication device monitoring methods rely on single sensors such as optical power meters and do not integrate multi-dimensional data such as temperature and vibration, resulting in a high false alarm rate for faults. The fixed threshold alarm mechanism cannot adapt to environmental changes such as day-night temperature differences, and the false trigger rate is high. 80% of faults rely on manual experience judgment, lacking the ability to analyze the degradation trend based on historical data. In the prior art, although a monitoring scheme based on optical signal analysis has been proposed, it does not solve the problems of multi-source data fusion and dynamic threshold setting, and there are significant limitations especially in the real-time anomaly detection scenario of long-distance optical fiber networks.

[0003] Therefore, there is an urgent need for a method for monitoring optical fiber communication devices based on big data to reduce the false alarm rate and improve the accuracy of fault prediction. Summary of the Invention

[0004] The purpose of this application is to provide a method, system, and medium for monitoring optical fiber communication devices based on big data. The core lies in synchronously collecting the operating parameters and environmental parameters of optical fiber devices through a multi-source sensor network, generating a standard parameter information matrix through spatio-temporal alignment and noise suppression; further extracting features such as optical signal stability index and vibration energy anomaly through joint time-frequency domain analysis, and constructing a device health status matrix; predicting the degradation process based on a bidirectional LSTM model with an attention mechanism, and outputting a visual diagnosis result in combination with a dynamic threshold grading alarm mechanism. This solution solves the problems of high false alarm rate and insufficient prediction ability of traditional monitoring methods, realizes a reduction in the false alarm rate, and improves the accuracy of fault prediction.

[0005] This application provides a method for monitoring optical fiber communication devices based on big data, including the following steps: Synchronously collect the operating parameter set and environmental parameter set of the optical fiber device within a preset time window through a multi-source sensor network and perform preprocessing to generate a standard parameter information matrix of the optical fiber device; Extract time-frequency domain feature data according to the standard parameter information matrix of the optical fiber device and construct a health status matrix of the optical fiber device; Predict the device degradation process based on a preset deep learning model and output a diagnosis result through a visual interface.

[0006] Among them, in a method for monitoring optical fiber communication devices based on big data according to this application, the multi-source sensors include: Optical time domain reflectometer, three-axis MEMS accelerometer, and distributed optical fiber temperature sensor.

[0007] Among them, in a method for monitoring optical fiber communication equipment based on big data described in this application, synchronously collecting the operation parameter set and environmental parameter set of the optical fiber equipment within a preset time window and performing preprocessing to generate a standard parameter information matrix of the optical fiber equipment specifically includes: Synchronously collecting the operation parameter set and environmental parameter set of the optical fiber equipment within a preset time window; The operation parameter set of the optical fiber equipment includes optical signal intensity, wavelength offset, bit error rate, and polarization mode dispersion; The environmental parameter set includes the device surface temperature gradient, environmental relative humidity, and three-dimensional vibration acceleration; Generating a standard parameter information matrix of the optical fiber equipment through spatio-temporal alignment and noise suppression processing based on the operation parameter set and environmental parameter set of the optical fiber equipment.

[0008] Among them, in a method for monitoring optical fiber communication equipment based on big data described in this application, extracting time-frequency domain feature data according to the standard parameter information matrix of the optical fiber equipment and constructing a health state matrix of the optical fiber equipment specifically includes: Extracting time-frequency domain feature data according to the standard parameter information matrix of the optical fiber equipment, including the optical signal time domain, optical signal frequency domain, vibration characteristics, and environmental compensation parameters; Constructing a health state matrix of the optical fiber equipment based on the optical signal time domain, optical signal frequency domain, vibration characteristics, and environmental compensation parameters.

[0009] Among them, in a method for monitoring optical fiber communication equipment based on big data described in this application, predicting the equipment degradation process based on a preset deep learning model and outputting a diagnostic result through a visualization interface specifically includes: Predicting the equipment degradation process based on a preset deep learning model; The preset deep learning model is a bidirectional LSTM prediction model based on an attention mechanism, including an input layer, a bidirectional LSTM layer, a time attention layer, and an output layer; Outputting a diagnostic result through a visualization interface, including a fault probability heat map, an AR augmented reality view, and a maintenance recommendation generation rule.

[0010] Among them, in a method for monitoring optical fiber communication equipment based on big data described in this application, the maintenance recommendation generation rule specifically includes: Obtaining the fault probability of the optical fiber equipment and comparing it with a preset first fault probability threshold and a preset second fault probability threshold respectively; The preset first fault probability threshold is greater than the preset second fault probability threshold; If the fault probability is greater than or equal to the preset first fault probability threshold, then send a shutdown for maintenance message; If the failure probability is less than the preset first failure probability threshold and greater than or equal to the preset second failure probability threshold, then send priority troubleshooting information; If the failure probability is less than the preset second failure probability threshold, then send continuous monitoring information.

[0011] In a second aspect, the present application provides a fiber optic communication device monitoring system based on big data. The system includes: a memory and a processor. The memory includes a program for a method for monitoring a fiber optic communication device based on big data. When the program for the method for monitoring a fiber optic communication device based on big data is executed by the processor, the following steps are implemented: Synchronously collect the fiber optic device operation parameter set and the environmental parameter set within a preset time window through a multi-source sensor network and perform preprocessing to generate a fiber optic device standard parameter information matrix; Extract time-frequency domain feature data according to the fiber optic device standard parameter information matrix and construct a fiber optic device health status matrix; Predict the device degradation process based on a preset deep learning model and output the diagnosis result through a visualization interface.

[0012] Among them, in the fiber optic communication device monitoring system based on big data described in the present application, the multi-source sensor includes: An optical time domain reflectometer, a three-axis MEMS accelerometer, and a distributed fiber optic temperature sensor.

[0013] Among them, in the fiber optic communication device monitoring system based on big data described in the present application, the step of synchronously collecting the fiber optic device operation parameter set and the environmental parameter set within a preset time window and performing preprocessing to generate a fiber optic device standard parameter information matrix is specifically: Synchronously collect the fiber optic device operation parameter set and the environmental parameter set within a preset time window; The fiber optic device operation parameter set includes optical signal intensity, wavelength offset, bit error rate, and polarization mode dispersion; The environmental parameter set includes the device surface temperature gradient, environmental relative humidity, and three-dimensional vibration acceleration; Generate a fiber optic device standard parameter information matrix according to the fiber optic device operation parameter set and the environmental parameter set through spatio-temporal alignment and noise suppression processing.

[0014] In a third aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium includes a program for a method for monitoring a fiber optic communication device based on big data. When the program for the method for monitoring a fiber optic communication device based on big data is executed by a processor, the steps of the method for monitoring a fiber optic communication device based on big data as described in any one of the above are implemented.

[0015] As can be seen from the above, a monitoring method, system, and medium for optical fiber communication equipment based on big data provided by the embodiments of the present application synchronously collect the operating parameters and environmental parameters of optical fiber equipment through a multi-source sensor network, and generate a standard parameter information matrix through spatio-temporal alignment and noise suppression; further extract features such as the optical signal stability index and vibration energy abnormality through joint time-frequency domain analysis, and construct a device health status matrix; predict the deterioration process based on a bidirectional LSTM model with an attention mechanism, and output a visual diagnosis result in combination with a dynamic threshold classification warning mechanism. This solution solves the problems of high false alarm rates and insufficient prediction capabilities of traditional monitoring methods, reduces the false alarm rate, and improves the accuracy of fault prediction.

[0016] Other features and advantages of the present application will be described in the subsequent description, and part of them will become obvious from the description, or be understood by implementing the embodiments of the present application. The objectives and other advantages of the present application can be achieved and obtained through the structures specifically pointed out in the written description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is a flowchart of a monitoring method for optical fiber communication equipment based on big data provided by the embodiments of the present application; Figure 2 It is a flowchart of generating a standard parameter information matrix of optical fiber equipment for a monitoring method for optical fiber communication equipment based on big data provided by the embodiments of the present application; Figure 3 It is a flowchart of constructing a health status matrix of optical fiber equipment for a monitoring method for optical fiber communication equipment based on big data provided by the embodiments of the present application; Figure 4 It is a flowchart of predicting the deterioration process of equipment and outputting a diagnosis result through a visual interface for a monitoring method for optical fiber communication equipment based on big data provided by the embodiments of the present application; DETAILED DESCRIPTION OF THE EMBODIMENTS The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and illustrated herein can be arranged and designed in a variety of different configurations. Therefore, the detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the claimed present application, but merely represents selected embodiments of the present application. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative efforts fall within the scope of protection of the present application.

[0019] It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present application, the terms "first", "second", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance. It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0020] Please refer to Figure 1 , Figure 1 which is a flowchart of a method for monitoring optical fiber communication devices based on big data in some embodiments of the present application. This method for monitoring optical fiber communication devices based on big data is used in terminal devices, such as computer, mobile phone terminals, etc. This method for monitoring optical fiber communication devices based on big data includes the following steps: S101. Synchronously collect the operation parameter set and environmental parameter set of the optical fiber device within a preset time window through a multi-source sensor network and perform preprocessing to generate a standard parameter information matrix of the optical fiber device; S102. Extract time-frequency domain feature data according to the standard parameter information matrix of the optical fiber device and construct a health status matrix of the optical fiber device; S103. Predict the device degradation process based on a preset deep learning model and output the diagnosis result through a visualization interface.

[0021] Among them, first, through a heterogeneous sensor network composed of an optical time domain reflectometer, a three-axis MEMS accelerometer, and a distributed fiber optic temperature sensor, the operating parameters (optical signal intensity, polarization mode dispersion, bit error rate) and environmental parameters (temperature gradient, three-dimensional vibration acceleration, environmental humidity) of the fiber optic device are synchronously collected. The dynamic time warping algorithm (DTW) is used to align the multi-source time series data, and the wavelet soft threshold denoising is combined to generate a standard parameter information matrix. Subsequently, wavelet packet decomposition is performed on the optical signal to extract the fixed-frequency band energy entropy ratio feature. At the same time, the vibration signal is subjected to modal decomposition, and the sample entropy of the first 3 IMF components is calculated. The multi-dimensional features are fused through a sparse autoencoder to construct a 6-dimensional health status matrix including the optical stability index, vibration anomaly degree, temperature coupling factor, etc. On this basis, a bidirectional LSTM prediction model is used to analyze the 72-hour historical data, and the time attention mechanism is combined to output the failure probability in the next 24 hours. According to the dynamic threshold grading alarm strategy, a multi-level response is triggered - sending a text message to notify the maintenance personnel or automatically switching to the standby link. Finally, the fault point is marked through an AR enhanced view and a priority maintenance work order is generated (the inspection path is optimized by combining the Dijkstra algorithm). This solution achieves remarkable effects of reducing the false alarm rate, improving the accuracy of fault prediction, and shortening the response delay, overcoming the technical bottlenecks of the traditional monitoring method with a single data dimension and poor environmental adaptability, and providing an efficient and reliable intelligent solution for the operation and maintenance of the fiber optic network.

[0022] According to an embodiment of the present invention, the multi-source sensor includes: An optical time domain reflectometer, a three-axis MEMS accelerometer, and a distributed fiber optic temperature sensor.

[0023] Among them, the optical time domain reflectometer (OTDR) adopts a 1260 - 1650nm wide spectrum design, and can detect fiber optic link loss events in real time (accuracy ±0.02dB / km), and accurately locate breakpoints and bending faults with a spatial resolution of 0.1m level; the three-axis MEMS accelerometer has a range of ±16g and a wide frequency response of 0.5Hz - 5kHz (noise density 100μg / √Hz), and eliminates electromagnetic interference through Butterworth band-pass filtering, and accurately captures abnormal mechanical vibrations of the device; the distributed fiber optic temperature sensor (DTS) realizes the monitoring of the temperature field of the entire link. The multi-source cooperation technology has remarkable effects, with the false alarm rate decreasing and the fault location accuracy improving. The core of the innovation lies in the spectrum cooperation coverage and data fusion mechanism (vibration-temperature joint alarm threshold, co-fiber deployment of OTDR and DTS). Engineering-wise, an IP68 protection level MEMS sensor is adapted to the harsh environment, providing a highly reliable solution for the intelligent operation and maintenance of the fiber optic network.

[0024] Please refer to Figure 2 , Figure 2It is a flowchart of generating a standard parameter information matrix of optical fiber devices for a method of monitoring optical fiber communication devices based on big data in some embodiments of the present application. According to an embodiment of the present invention, synchronously collect the operating parameter set and environmental parameter set of the optical fiber device within a preset time window and perform preprocessing to generate a standard parameter information matrix of the optical fiber device. Specifically: S201. Synchronously collect the operating parameter set and environmental parameter set of the optical fiber device within a preset time window; S202. The operating parameter set of the optical fiber device includes optical signal intensity, wavelength offset, bit error rate, and polarization mode dispersion; S203. The environmental parameter set includes the device surface temperature gradient, environmental relative humidity, and three-dimensional vibration acceleration; S204. Generate a standard parameter information matrix of the optical fiber device through spatio-temporal alignment and noise suppression processing based on the operating parameter set and environmental parameter set of the optical fiber device.

[0025] Among them, in step S201, based on the timestamp synchronization trigger mechanism, synchronously collect the operating parameter set and environmental parameter set of the optical fiber device within a preset time window (a typical value is 10 minutes, which can be dynamically adjusted according to network load). The operating parameter set defined in step S202 includes optical signal intensity, wavelength offset, bit error rate, and polarization mode dispersion; the environmental parameter set covered by step S203 consists of the device surface temperature gradient, environmental relative humidity, and three-dimensional vibration acceleration. Step S204 performs spatio-temporal alignment and noise suppression processing on the original data: uses the dynamic time warping algorithm (DTW) to eliminate the transmission delay of multi-sensors and achieve data point-level alignment; suppresses the random noise in the optical signal through wavelet soft threshold denoising, and combines adaptive Kalman filtering to remove the electromagnetic interference component in the vibration signal. The finally generated standard parameter information matrix is stored in a structured form, with the false alarm rate reduced compared to the traditional method and the data processing efficiency improved.

[0026] Please refer to Figure 3 , Figure 3 It is a flowchart of constructing a health status matrix of optical fiber devices for a method of monitoring optical fiber communication devices based on big data in some embodiments of the present application. According to an embodiment of the present invention, extract time-frequency domain feature data based on the standard parameter information matrix of the optical fiber device and construct a health status matrix of the optical fiber device. Specifically: S301. Extract time-frequency domain feature data based on the standard parameter information matrix of the optical fiber device, including the optical signal time domain, optical signal frequency domain, vibration characteristics, and environmental compensation parameters; S302. Construct a health status matrix of the optical fiber device based on the optical signal time domain, optical signal frequency domain, vibration characteristics, and environmental compensation parameters.

[0027] Among them, in step S301, four types of core features are extracted based on the standard parameter information matrix. Time-domain features of the optical signal: Calculate the mean optical power, kurtosis (to detect pulse anomalies), and zero-crossing rate (to monitor the signal oscillation frequency); Frequency-domain features of the optical signal: Use wavelet packet decomposition to extract the energy entropy ratio of a fixed frequency band; Vibration features: Decompose the vibration signal through the CEEMDAN algorithm and extract the sample entropy and instantaneous frequency variance of the first three IMF components; Environmental compensation parameters: Establish a temperature-optical attenuation dynamic compensation model to reduce the false alarm rate. In step S302, the 14-dimensional initial features are dimensionally reduced through a sparse autoencoder. Among them, the time domain of the optical signal contains 3 features, the frequency domain of the optical signal contains 4 features, the vibration features contain 2 features, and the environmental compensation parameters contain 5 features. A 6-dimensional health status matrix including the optical stability index, vibration anomaly degree, temperature coupling factor, etc. is constructed. Its row vectors correspond to the time series, and after the column vectors are normalized, they are input into the bidirectional LSTM model to improve the model training efficiency. The technical breakthrough lies in cross-domain feature fusion (optics, mechanics, thermodynamics) and the dynamic weight mechanism. Engineeringly, the time-consuming for feature extraction meets the requirements of telecom-level real-time performance, providing core support for the intelligent operation and maintenance of fiber optic networks.

[0028] Please refer to Figure 4 , Figure 4 is a flowchart of a prediction device for predicting the deterioration process of an optical fiber communication device based on big data and outputting a diagnosis result through a visualization interface in some embodiments of the present application. According to an embodiment of the present invention, the prediction of the device deterioration process based on a preset deep learning model and the output of the diagnosis result through a visualization interface are specifically as follows: S401. Predict the device deterioration process based on a preset deep learning model; S402. The preset deep learning model is a bidirectional LSTM prediction model based on an attention mechanism, including an input layer, a bidirectional LSTM layer, a time attention layer, and an output layer; S403. Output the diagnosis result through a visualization interface, including a fault probability heat map, an AR augmented reality view, and a maintenance advice generation rule.

[0029] Among them, in step S401, the historical health status matrix (time span ≥ 72 hours) is analyzed based on a preset deep learning model to predict the development trend of equipment failures in the next 24 hours; the prediction model defined in step S402 adopts a bidirectional LSTM architecture based on the attention mechanism. Input layer: Receives multi-dimensional feature vectors. Bidirectional LSTM layer: Time step 72 (corresponding to 3 days of historical data); Time attention layer, calculates the weights of each moment through the Softmax function. Output layer: The Sigmoid function generates a failure probability value between 0 and 1. The training uses the Focal Loss loss function; the visualization output system in step S403 includes three core modules. Fault probability heat map: Can display the historical failure rates of similar devices; AR augmented reality view: After scanning the device QR code on the mobile phone, superimpose and display the optical fiber stress distribution cloud map and three-dimensional navigation arrows to guide the nearest maintenance entrance. Through the deep integration of deep learning prediction and augmented reality visualization, pain points such as decision-making lag and information silos in traditional operation and maintenance are overcome, and a new benchmark is set for the intelligent operation and maintenance of optical fiber networks.

[0030] According to an embodiment of the present invention, the maintenance recommendation generation rule is specifically: Obtain the failure probability of the optical fiber device and compare it with a preset first failure probability threshold and a preset second failure probability threshold respectively; The preset first failure probability threshold is greater than the preset second failure probability threshold; If the failure probability is greater than or equal to the preset first failure probability threshold, send a shutdown and maintenance message; If the failure probability is less than the preset first failure probability threshold and greater than or equal to the preset second failure probability threshold, send a priority troubleshooting message; If the failure probability is less than the preset second failure probability threshold, send a continuous monitoring message.

[0031] Among them, first obtain the failure probability of the optical fiber device calculated by the deep learning model, and then compare this probability value with two preset thresholds (where the first failure probability threshold > the second failure probability threshold). When the failure probability ≥ the first threshold, trigger the highest-level alarm, and the system automatically sends an instruction to require immediate shutdown and maintenance; when the failure probability is between the first threshold and the second threshold, generate a medium-level alarm and send a prompt message suggesting priority troubleshooting; if the failure probability < the second threshold, it is determined as a low-risk state, and the system only sends a regular prompt to maintain continuous monitoring. This three-level threshold judgment mechanism realizes a gradient response strategy from emergency handling, preventive maintenance to regular monitoring.

[0032] According to an embodiment of the present invention, it further includes: Extract the optical signal stability index according to the optical fiber device health status matrix and query it in the preset optical fiber device health list to obtain the health status deviation value; Compare the health status deviation value with a preset health status deviation threshold to obtain multiple deviation rate of deviation values; Compare the deviation rate of the deviation value with a preset deviation rate threshold; If the deviation rate of the deviation value is greater than or equal to the preset deviation rate threshold, send an unhealthy status message; If the deviation rate of the deviation value is less than the preset deviation rate threshold, send a healthy status message.

[0033] Among them, by collecting a time series dataset composed of optical signal parameters such as optical power and wavelength offset in real time, analyzing the fluctuation characteristics of optical signals within a preset time period (such as 10 minutes), identifying abnormal pulses and trend attenuation, and then performing normalization processing to map the calculation results to a 0-100 standard scoring interval, the higher the value, the better the stability. The preset fiber optic device health list database stores the optical signal stability reference values of different models of devices under standard operating conditions, records the linear attenuation coefficient of device aging on the stability index (such as a 1.2% decrease per year), and automatically adjusts the reference value according to the real-time ambient temperature (such as a 2% decrease in the reference value for every 5°C increase in temperature). The query process is realized through two-factor matching. Determine the device model according to the MAC address / serial number, and calculate the current effective reference value in combination with the device operation time and ambient temperature. The deviation value represents the degree of difference between the real-time stability index and the database reference value.

[0034] According to an embodiment of the present invention, it further includes: Obtain the historical health status matrix of the fiber optic device; Perform unsupervised training on the historical health status matrix based on the isolation forest algorithm to generate a device normal state reference model; Dynamically adjust the anomaly determination threshold according to the environmental temperature gradient to perform multi-level anomaly determination on the real-time collected data, including primary alarm, intermediate alarm and emergency alarm.

[0035] Among them, continuously collect device operation parameters from sensor nodes deployed in the fiber optic network, including key indicators such as optical power, wavelength offset, and vibration spectrum, and organize historical data of at least 6 months into a multi-dimensional matrix according to a time window (usually at 10-minute intervals). Each time, randomly select several features to construct an isolation tree, divide the data until a preset depth is reached, establish a temperature binning model, classify and store historical data according to working temperature ranges (such as 20-25°C, 25-30°C, etc.), and independently train sub-models for each temperature range to form a temperature-adaptive model cluster. The dynamic threshold multi-level alarm mechanism is divided into primary alarm, intermediate alarm and emergency alarm. The response measures for the primary alarm are log recording and yellow warning on the interface. The response measures for the intermediate alarm are to send notifications to the operation and maintenance personnel. The response measures for the emergency alarm are to automatically start the protection device and make a phone alarm.

[0036] The present invention also discloses a monitoring system for optical fiber communication devices based on big data, including a memory and a processor. A monitoring method program for optical fiber communication devices based on big data is included in the memory. When the monitoring method program for optical fiber communication devices based on big data is executed by the processor, the following steps are implemented: Synchronously collect the operation parameter set and environmental parameter set of the optical fiber device within a preset time window through a multi-source sensor network and perform preprocessing to generate a standard parameter information matrix of the optical fiber device; Extract time-frequency domain feature data according to the standard parameter information matrix of the optical fiber device and construct a health state matrix of the optical fiber device; Predict the device degradation process based on a preset deep learning model and output the diagnosis result through a visualization interface.

[0037] Among them, first, through a heterogeneous sensor network composed of an optical time domain reflectometer, a three-axis MEMS accelerometer, and a distributed optical fiber temperature sensor, synchronously collect the operation parameters (optical signal intensity, polarization mode dispersion, bit error rate) and environmental parameters (temperature gradient, three-dimensional vibration acceleration, environmental humidity) of the optical fiber device, and use the dynamic time warping algorithm (DTW) to align the multi-source time series data, and combine wavelet soft threshold denoising to generate a standard parameter information matrix; Subsequently, perform wavelet packet decomposition on the optical signal, extract the fixed-band energy entropy ratio feature, and at the same time perform modal decomposition on the vibration signal, calculate the sample entropy of the first 3 IMF components, and fuse multi-dimensional features through a sparse autoencoder to construct a 6-dimensional health state matrix including optical stability index, vibration anomaly degree, temperature coupling factor, etc.; On this basis, use a bidirectional LSTM prediction model to analyze 72-hour historical data, combine the time attention mechanism to output the failure probability in the next 24 hours, and trigger a multi-level response according to the dynamic threshold grading warning strategy - send a text message to notify the maintenance personnel or automatically switch to the standby link, and finally mark the fault point through an AR enhanced view and generate a priority maintenance work order (optimize the maintenance path in combination with the Dijkstra algorithm). This solution achieves remarkable effects of reducing the false alarm rate, improving the accuracy of fault prediction, and shortening the response delay, overcomes the technical bottlenecks of the traditional monitoring method with a single data dimension and poor environmental adaptability, and provides an efficient and reliable intelligent solution for the operation and maintenance of optical fiber networks.

[0038] According to an embodiment of the present invention, the multi-source sensor includes: An optical time domain reflectometer, a three-axis MEMS accelerometer, and a distributed optical fiber temperature sensor.

[0039] Among them, the optical time domain reflectometer (OTDR) adopts a wide spectral design of 1260 - 1650 nm, real-time detects fiber optic link loss events (accuracy ±0.02 dB / km), and accurately locates breakpoints and bending faults with a spatial resolution at the 0.1 m level; the three-axis MEMS accelerometer has a range of ±16 g and a wide frequency response of 0.5 Hz - 5 kHz (noise density 100 μg / √Hz), eliminates electromagnetic interference through Butterworth band-pass filtering, and accurately captures abnormal mechanical vibrations of the device; the distributed fiber optic temperature sensor (DTS) realizes full-link temperature field monitoring. The multi-source collaboration technology has remarkable effects, with the false alarm rate decreasing and the fault location accuracy improving. The core of the innovation lies in the spectrum collaborative coverage and data fusion mechanism (vibration-temperature joint alarm threshold, co-fiber deployment of OTDR and DTS). Engineering-wise, an IP68 protection level MEMS sensor is adapted to the harsh environment, providing a highly reliable solution for the intelligent operation and maintenance of fiber optic networks.

[0040] According to the embodiment of the present invention, synchronously collect the operation parameter set and environment parameter set of the fiber optic device within a preset time window and perform preprocessing to generate a standard parameter information matrix of the fiber optic device, specifically as follows: Synchronously collect the operation parameter set and environment parameter set of the fiber optic device within a preset time window; The operation parameter set of the fiber optic device includes optical signal intensity, wavelength offset, bit error rate, and polarization mode dispersion; The environment parameter set includes the surface temperature gradient of the device, environmental relative humidity, and three-dimensional vibration acceleration; Generate a standard parameter information matrix of the fiber optic device through space-time alignment and noise suppression processing according to the operation parameter set and environment parameter set of the fiber optic device.

[0041] Among them, in step S201, based on the timestamp synchronization trigger mechanism, synchronously collect the operation parameter set and environment parameter set of the fiber optic device within a preset time window (the typical value is 10 minutes, which can be dynamically adjusted according to the network load). The operation parameter set defined in step S202 includes optical signal intensity, wavelength offset, bit error rate, and polarization mode dispersion; the environment parameter set covered by step S203 consists of the surface temperature gradient of the device, environmental relative humidity, and three-dimensional vibration acceleration. Step S204 performs space-time alignment and noise suppression processing on the original data: uses the dynamic time warping algorithm (DTW) to eliminate the transmission delay of multiple sensors and achieve data point-level alignment; suppresses the random noise in the optical signal through wavelet soft threshold denoising, and combines adaptive Kalman filtering to remove the electromagnetic interference component in the vibration signal. The finally generated standard parameter information matrix is stored in a structured form, with the false alarm rate reduced compared to the traditional method and the data processing efficiency improved.

[0042] According to an embodiment of the present invention, when extracting time-frequency domain feature data based on the standard parameter information matrix of the optical fiber device and constructing a health status matrix of the optical fiber device, specifically: Extract time-frequency domain feature data based on the standard parameter information matrix of the optical fiber device, including the time domain of the optical signal, the frequency domain of the optical signal, vibration characteristics, and environmental compensation parameters; Construct a health status matrix of the optical fiber device based on the time domain of the optical signal, the frequency domain of the optical signal, vibration characteristics, and environmental compensation parameters.

[0043] Among them, in step S301, four types of core features are extracted based on the standard parameter information matrix. For the time domain feature of the optical signal: calculate the mean optical power, kurtosis (detect pulse anomalies), and zero-crossing rate (monitor the signal oscillation frequency); for the frequency domain feature of the optical signal: use wavelet packet decomposition to extract the energy entropy ratio of a fixed frequency band; for vibration characteristics: decompose the vibration signal through the CEEMDAN algorithm and extract the sample entropy and instantaneous frequency variance of the first 3 IMF components; for environmental compensation parameters: establish a temperature-optical attenuation dynamic compensation model to reduce the false alarm rate. In step S302, the 14-dimensional initial features are dimensionally reduced through a sparse autoencoder. Among them, the time domain of the optical signal contains 3 features, the frequency domain of the optical signal contains 4 features, the vibration characteristics contain 2 features, and the environmental compensation parameters contain 5 features. A 6-dimensional health status matrix including the optical stability index, vibration anomaly degree, temperature coupling factor, etc. is constructed. Its row vectors correspond to the time series, and after the column vectors are normalized, they are input into the bidirectional LSTM model to improve the model training efficiency. The technical breakthrough lies in cross-domain feature fusion (light, mechanical, thermal) and the dynamic weight mechanism, and the time-consuming for feature extraction in engineering meets the real-time requirements of the telecommunications level, providing core support for the intelligent operation and maintenance of the optical fiber network.

[0044] According to an embodiment of the present invention, when predicting the deterioration process of the device based on a preset deep learning model and outputting the diagnosis result through a visualization interface, specifically: Predict the deterioration process of the device based on a preset deep learning model; The preset deep learning model is a bidirectional LSTM prediction model based on the attention mechanism, including an input layer, a bidirectional LSTM layer, a time attention layer, and an output layer; Output the diagnosis result through a visualization interface, including a fault probability heat map, an AR augmented reality view, and a maintenance advice generation rule.

[0045] Among them, in step S401, the historical health status matrix (time span ≥ 72 hours) is analyzed based on a preset deep learning model to predict the development trend of equipment failures in the next 24 hours; the prediction model defined in step S402 adopts a bidirectional LSTM architecture based on the attention mechanism. Input layer: receives a multi-dimensional feature vector. Bidirectional LSTM layer: time step 72 (corresponding to 3 days of historical data); Time attention layer, calculates the weights at each moment through the Softmax function. Output layer: the Sigmoid function generates a failure probability value between 0 and 1. The training uses the Focal Loss function; the visualization output system in step S403 includes three core modules. Fault probability heat map: can display the historical failure rates of similar equipment; AR augmented reality view: after scanning the equipment QR code on the mobile phone, superimpose and display the optical fiber stress distribution cloud map and three-dimensional navigation arrows to guide the nearest maintenance entrance. Through the deep integration of deep learning prediction and augmented reality visualization, it has overcome the pain points such as decision-making lag and information silos in traditional operation and maintenance, and set a new benchmark for the intelligent operation and maintenance of optical fiber networks.

[0046] According to an embodiment of the present invention, the maintenance suggestion generation rule is specifically: Obtain the failure probability of the optical fiber equipment and compare it with a preset first failure probability threshold and a preset second failure probability threshold respectively; The preset first failure probability threshold is greater than the preset second failure probability threshold; If the failure probability is greater than or equal to the preset first failure probability threshold, then send a shutdown and maintenance message; If the failure probability is less than the preset first failure probability threshold and greater than or equal to the preset second failure probability threshold, then send a priority troubleshooting message; If the failure probability is less than the preset second failure probability threshold, then send a continuous monitoring message.

[0047] Among them, first obtain the failure probability of the optical fiber equipment calculated by the deep learning model, and then compare this probability value with two preset thresholds (where the first failure probability threshold > the second failure probability threshold). When the failure probability ≥ the first threshold, trigger the highest-level alarm, and the system automatically sends an instruction to require immediate shutdown and maintenance; when the failure probability is between the first threshold and the second threshold, generate a medium-level alarm and send a prompt message suggesting priority troubleshooting; if the failure probability < the second threshold, it is determined to be in a low-risk state, and the system only sends a regular prompt for continuous monitoring. This three-level threshold judgment mechanism realizes a gradient response strategy from emergency disposal, preventive maintenance to regular monitoring.

[0048] According to an embodiment of the present invention, it further includes: Extract the optical signal stability index according to the optical fiber equipment health status matrix and input it into a preset optical fiber equipment health list for query to obtain the health status deviation value; Compare the health status deviation value with a preset health status deviation threshold to obtain a plurality of deviation value deviation rates; Compare the deviation value deviation rate with a preset deviation rate threshold; If the deviation value deviation rate is greater than or equal to the preset deviation rate threshold, send an unhealthy status message; If the deviation value deviation rate is less than the preset deviation rate threshold, send a healthy status message.

[0049] Among them, by collecting in real time a time series data set composed of optical signal parameters such as optical power and wavelength offset, analyzing the fluctuation characteristics of the optical signal within a preset time period (such as 10 minutes), identifying abnormal pulses and trend attenuation, and then performing normalization processing to map the calculation results to the 0-100 standard scoring interval. The higher the value, the better the stability. The preset optical fiber device health list database stores the optical signal stability reference values of different models of devices under standard working conditions, records the linear attenuation coefficient of device aging on the stability index (such as a 1.2% decrease per year), and automatically adjusts the reference value according to the real-time ambient temperature (such as a 2% decrease in the reference value for every 5°C increase in temperature). The query process is achieved through two-factor matching. Determine the device model according to the MAC address / serial number, and calculate the current effective reference value in combination with the device commissioning time and ambient temperature. The deviation value represents the degree of difference between the real-time stability index and the database reference value.

[0050] According to an embodiment of the present invention, it further includes: Obtain the historical health status matrix of the optical fiber device; Perform unsupervised training on the historical health status matrix based on the isolation forest algorithm to generate a device normal state reference model; Dynamically adjust the anomaly determination threshold according to the environmental temperature gradient to perform multi-level anomaly determination on the real-time collected data, including primary alarm, intermediate alarm, and emergency alarm.

[0051] Among them, continuously collect device operation parameters from sensor nodes deployed in the optical fiber network, including key indicators such as optical power, wavelength offset, and vibration spectrum. Organize historical data of at least 6 months into a multi-dimensional matrix according to a time window (usually at 10-minute intervals). Each time, randomly select several features to construct an isolation tree, divide the data until reaching a preset depth, establish a temperature binning model, classify and store the historical data according to working temperature intervals (such as 20-25°C, 25-30°C, etc.), and independently train sub-models for each temperature interval to form a temperature-adaptive model cluster. The dynamic threshold multi-level alarm mechanism is divided into primary alarm, intermediate alarm, and emergency alarm. The response measures for the primary alarm are log recording and yellow warning on the interface. The response measures for the intermediate alarm are sending notifications to the operation and maintenance personnel. The response measures for the emergency alarm are automatically starting the protection device and giving a phone alarm.

[0052] In the third aspect of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium includes a program for a method of monitoring optical fiber communication devices based on big data. When the program for the method of monitoring optical fiber communication devices based on big data is executed by a processor, the steps of a method of monitoring optical fiber communication devices based on big data as described in any one of the above are implemented.

[0053] A method, system, and medium for monitoring optical fiber communication devices based on big data disclosed by the present invention. The core technology lies in achieving high-precision fault prediction and hierarchical response through multi-dimensional data fusion and a dynamic deep learning model. First, a heterogeneous sensor network composed of an optical time domain reflectometer (OTDR, operating wavelength 1260 - 1650 nm), a three-axis MEMS accelerometer (±16 g range), and a distributed fiber optic temperature sensor (DTS, ±0.5 °C accuracy) synchronously collects the operating parameters of the fiber optic device (optical signal intensity, wavelength offset, bit error rate, polarization mode dispersion) and environmental parameters (temperature gradient, three-dimensional vibration acceleration, relative humidity), and aligns the time series data through the dynamic time warping algorithm (DTW). Combining wavelet soft threshold denoising and Kalman filtering to eliminate noise, a structured standard parameter information matrix (600×6 dimensions) is generated. Subsequently, an optical signal stability index, vibration characteristics (CEEMDAN decomposition sample entropy), and temperature compensation parameters are extracted from the time-frequency domain, and fused into a 6-dimensional health state matrix using a sparse autoencoder. On this basis, a bidirectional LSTM model based on the attention mechanism is used to analyze historical data, dynamically predict the fault probability for the next 24 hours in combination with time attention weights, and trigger a response through a dynamic threshold hierarchical alarm mechanism: when the fault probability ≥ the first threshold (such as 0.8), a shutdown for maintenance instruction is sent; when it is between the first and second thresholds (such as 0.6), it is prompted to prioritize troubleshooting; when it is lower than the second threshold, continuous monitoring is carried out. The system innovatively integrates an AR augmented reality view to intuitively display the stress distribution of the fault point, optimizes the maintenance path in combination with the Dijkstra algorithm, and establishes an anomaly detection model with temperature adaptability through the isolation forest algorithm, achieving a reduction in the false alarm rate and an improvement in the accuracy of fault prediction, overcoming the technical bottlenecks of traditional methods relying on a single data source and poor environmental adaptability, and providing a real-time and accurate intelligent operation and maintenance solution for the optical fiber network.

[0054] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces. The indirect coupling or communication connection of devices or units can be electrical, mechanical, or other forms.

[0055] The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0056] In addition, each functional unit in the embodiments of the present invention can be all integrated in a processing unit, or each unit can be separately used as a unit, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware, or in the form of a combination of hardware and software functional units.

[0057] Those of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a readable storage medium. When the program is executed, it executes the steps including the above method embodiments. The foregoing storage medium includes various media that can store program codes, such as removable storage devices, read-only memories, random access memories, magnetic disks, or optical discs.

[0058] Alternatively, if the above integrated units of the present invention are implemented in the form of software functional modules and sold or used as independent products, they can also be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiments of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The foregoing storage medium includes various media that can store program codes, such as removable storage devices, ROM, RAM, magnetic disks, or optical discs.

Claims

1. A method for monitoring optical fiber communication equipment based on big data, characterized in that: The following steps are involved: The optical fiber equipment operation parameter set and environmental parameter set within a preset time window are synchronously collected and preprocessed through a multi-source sensor network to generate an optical fiber equipment standard parameter information matrix; Extracting time-frequency domain feature data according to the optical fiber equipment standard parameter information matrix and constructing an optical fiber equipment health status matrix; Predict equipment degradation process based on preset deep learning models and output diagnostic results through a visual interface.

2. The optical fiber communication equipment monitoring method based on big data according to claim 1, characterized in that: The multi-source sensor comprises: Optical time domain reflectometer, triaxial MEMS accelerometer and distributed fiber optic temperature sensor.

3. The optical fiber communication equipment monitoring method based on big data according to claim 1, characterized in that: The synchronous collection of the optical fiber equipment operation parameter set and the environmental parameter set within the preset time window and preprocessing to generate the optical fiber equipment standard parameter information matrix is ​​specifically: Synchronously collect the optical fiber equipment operating parameter set and environmental parameter set within the preset time window; The optical fiber equipment operation parameter set includes optical signal strength, wavelength offset, bit error rate and polarization mode dispersion; The environmental parameter set includes equipment surface temperature gradient, environmental relative humidity and three-dimensional vibration acceleration; According to the optical fiber equipment operation parameter set and the environmental parameter set, a standard parameter information matrix of the optical fiber equipment is generated through time-space alignment and noise suppression processing.

4. The optical fiber communication equipment monitoring method based on big data according to claim 1, characterized in that: The extracting of time-frequency domain feature data and constructing a fiber optic equipment health status matrix according to the fiber optic equipment standard parameter information matrix is ​​specifically as follows: Extracting time-frequency domain characteristic data according to the optical fiber equipment standard parameter information matrix, including optical signal time domain, optical signal frequency domain, vibration characteristics and environmental compensation parameters; An optical fiber equipment health status matrix is ​​constructed according to the optical signal time domain, optical signal frequency domain, vibration characteristics and environmental compensation parameters.

5. The optical fiber communication equipment monitoring method based on big data according to claim 1, characterized in that: The prediction of equipment degradation process based on the preset deep learning model and output of diagnosis results through a visual interface are specifically as follows: Predict equipment degradation process based on preset deep learning models; The preset deep learning model is a bidirectional LSTM prediction model based on the attention mechanism, including an input layer, a bidirectional LSTM layer, a temporal attention layer and an output layer; The diagnostic results are output through a visual interface, including fault probability heat map, AR augmented reality view and maintenance suggestion generation rules.

6. The optical fiber communication equipment monitoring method based on big data according to claim 5 is characterized in that: The maintenance suggestion generation rules are specifically as follows: Obtaining the failure probability of the optical fiber device and comparing it with a preset first failure probability threshold and a preset second failure probability threshold respectively; The preset first fault probability threshold is greater than the preset second fault probability threshold; If the fault probability is greater than or equal to the preset first fault probability threshold, then send shutdown maintenance information; If the fault probability is less than the preset first fault probability threshold and greater than or equal to the preset second fault probability threshold, a priority troubleshooting message is sent; If the fault probability is less than the preset second fault probability threshold, continuous monitoring information is sent.

7. A fiber optic communication equipment monitoring system based on big data, characterized in that: The invention comprises a memory and a processor, wherein the memory comprises a method program for monitoring optical fiber communication equipment based on big data, and when the method program for monitoring optical fiber communication equipment based on big data is executed by the processor, the following steps are implemented, specifically: The optical fiber equipment operation parameter set and environmental parameter set within a preset time window are synchronously collected and preprocessed through a multi-source sensor network to generate an optical fiber equipment standard parameter information matrix; Extracting time-frequency domain feature data according to the optical fiber equipment standard parameter information matrix and constructing an optical fiber equipment health status matrix; Predict equipment degradation process based on preset deep learning models and output diagnostic results through a visual interface.

8. The optical fiber communication equipment monitoring system based on big data according to claim 7, characterized in that: The multi-source sensor comprises: Optical time domain reflectometer, triaxial MEMS accelerometer and distributed fiber optic temperature sensor.

9. The optical fiber communication equipment monitoring system based on big data according to claim 7, characterized in that: The synchronous collection of the optical fiber equipment operation parameter set and the environmental parameter set within the preset time window and preprocessing to generate the optical fiber equipment standard parameter information matrix is ​​specifically: Synchronously collect the optical fiber equipment operating parameter set and environmental parameter set within the preset time window; The optical fiber equipment operation parameter set includes optical signal strength, wavelength offset, bit error rate and polarization mode dispersion; The environmental parameter set includes equipment surface temperature gradient, environmental relative humidity and three-dimensional vibration acceleration; According to the optical fiber equipment operation parameter set and the environmental parameter set, a standard parameter information matrix of the optical fiber equipment is generated through time-space alignment and noise suppression processing.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a fiber optic communication equipment monitoring method based on big data, a system and a medium program. When the fiber optic communication equipment monitoring method based on big data, the system and the medium program are executed by a processor, the steps of the fiber optic communication equipment monitoring method based on big data as described in any one of claims 1 to 6 are implemented.

Images