Multi-dimensional state monitoring and fault prediction method and system for optical cable

Abstract

An optical cable multi-dimensional state monitoring and fault prediction method and system are provided, and the method comprises the steps: obtaining optical cable environment characteristic data, historical fault distribution probability and operation risk indexes, selecting key nodes, deploying state sensors, and collecting multi-dimensional state data to generate an initial state data set; establishing a fault judgment model based on the initial data set, and calculating multi-dimensional state feature data and determining an initial distribution range in combination with historical fault probability and environment features; arranging the physical state value and the communication state value according to a time sequence, and generating a dynamic threshold range of the key node by combining dynamic correction of environment characteristic data; and when the feature data exceed the threshold range, marking abnormity, calculating a judgment parameter through a fault model, and comparing the judgment parameter with a preset value to confirm a fault.

Description

Technical Field

[0001] The present invention belongs to the technical field of optical cable monitoring, and more specifically, relates to a multi-dimensional state monitoring and fault prediction method and system for optical cables. Background Art

[0002] In today's information society, as an important infrastructure for carrying massive data transmission, the security and stability of optical cable communication networks are of vital importance to ensure the continuity and reliability of information flow. The reliable operation of optical cable systems is directly related to whether telecommunications services can efficiently and stably serve social and economic activities. However, with the continuous expansion of optical cable networks and the increasing complexity of application scenarios, optical cable systems face various potential threats and challenges, including natural environmental factors (such as earthquakes, lightning, floods, etc.), human factors (such as construction excavation, theft, etc.), and failures caused by aging equipment.

[0003] At present, the optical cable monitoring system mainly relies on traditional fault detection methods, which can only detect faults when the optical cable is damaged or the performance is degraded, but it is difficult to effectively monitor the potential and hidden optical cable faults. In addition, the existing monitoring systems often lack the ability to continuously, real-time and comprehensively monitor the status of the optical cable, and cannot conduct in-depth analysis of the health status of the optical cable, which results in limited response speed of repair work when a fault occurs, affecting the quality of communication services. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] To solve the problem that the existing optical cable monitoring system is difficult to effectively monitor potential and hidden optical cable faults on the one hand, and lacks the ability to continuously, real-time and comprehensively monitor the optical cable status, and is unable to conduct in-depth analysis of the health status of the optical cable on the other hand.

[0006] The present invention adopts the following technical solution.

[0007] A first aspect of the present invention provides a multi-dimensional state monitoring and fault prediction method for an optical cable, wherein the optical cable comprises a plurality of sections, and comprises the following steps:

[0008] Step 1, obtain and use the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable to select key nodes from each section of the optical cable; deploy different types of state sensors at the key nodes to collect multi-dimensional state data of the optical cable, including physical state value δ and communication state value η;

[0009] Step 2, using the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable to correct the historical data of the multi-dimensional state data, and using the mean and standard deviation of the corrected historical data of the multi-dimensional state data to establish a dynamic threshold range of the multi-dimensional state data;

[0010] Step 3: Establish a fitting model for multi-dimensional state data based on the working principles of different types of state sensors, and establish a fault state judgment model based on the fitting model for multi-dimensional state data; use the corrected historical data of the multi-dimensional state data to obtain a fault state judgment value based on the fault state judgment model; and use the mean and standard deviation of the fault state judgment value to establish a distribution range of the fault state judgment value;

[0011] Step 4, when the physical state value of the key node collected in real time is not within the dynamic threshold range of the physical state value and / or the communication state value is not within the dynamic threshold range of the communication state value, the key node is marked as an abnormal section, and the physical state value and communication state value of the key node collected in real time are used to obtain the real-time value of the fault state judgment of the key node based on the fault state judgment model; when the real-time value of the fault state judgment of the key node is not within the distribution range of the fault state judgment value, it is determined that a fault has occurred at the key node.

[0012] Preferably, in step 1, the environmental characteristic data of the optical cable include the temperature change rate and relative humidity; the historical fault frequency and historical fault location are extracted according to the historical fault distribution probability; the operation risk index includes the external electromagnetic interference intensity;

[0013] The temperature change rate, relative humidity, historical fault frequency, external electromagnetic interference intensity and historical fault location are used to form multiple criteria. When any section of the optical cable meets at least one criterion, it is determined to be a critical node.

[0014] According to the requirements of optical cable status monitoring, determine the type of status sensors arranged at each key node, including physical status sensors and communication status sensors; physical status sensors include temperature sensors, pressure sensors, humidity sensors and strain gauges, which respectively collect the temperature, pressure, humidity and bending radius of the optical cable; communication status sensors include optical power sensors, optical attenuators and high-speed data acquisition cards, which respectively measure optical power, attenuation rate and transmission rate;

[0015] Physical status sensors and communication status sensors are integrated and installed on each key node. The data collected by the physical status sensors and the data collected by the communication status sensors constitute multi-dimensional status data. The multi-dimensional status data is transmitted to the central server through the data transmission module, and the initial status data set of the key nodes is established in the central server.

[0016] Preferably, the multiple criteria include:

[0017] 3) The temperature change rate is greater than ±20℃ / hour;

[0018] 4) Relative humidity is greater than 90%;

[0019] 3) The historical failure frequency is greater than once a month;

[0020] 4) The external electromagnetic interference intensity is greater than 50μT;

[0021] 5) The location of historical faults occurs repeatedly in the same section.

[0022] Preferably, based on the working principle of the physical state sensor, the fitting model of the physical state value satisfies the following relationship:

[0023]

[0024] In the formula, δ is the physical state value; T is the temperature; P is the pressure; S is the humidity; W is the bending radius; dx is the constant integral operation;

[0025] Based on the working principle of the communication status sensor, the fitting model of the communication status value satisfies the following relationship:

[0026]

[0027] In the formula, η is the communication state value; G is the optical power; J is the attenuation rate; C is the transmission rate;

[0028] Based on the fitting model of the physical state value and the fitting model of the communication state value, the fault state judgment model satisfies the following relationship:

[0029]

[0030] Where μ is the fault state judgment parameter.

[0031] Preferably, step 3 comprises:

[0032] Calculate the mean of the physical state values and standard deviation σ δ , to determine the distribution range of the physical state value Calculate the mean of the communication status values and standard deviation ση , to determine the distribution range of the communication status value

[0033] The physical state correction value is calculated using the following relationship:

[0034] δ adj =δ+k 1 ΔT+k 2 ·S+k 3 ΔP

[0035] In the formula, δ adj is the physical state correction value, ΔT is the temperature change, S is the humidity value, ΔP is the pressure change, k 1 , k 2 , k 3 All are weight coefficients;

[0036] The communication status correction value is calculated using the following relationship:

[0037] η adj =η+k 4 ·E+k 1 ΔT+k 2 ·S

[0038] Where η adj is the communication status correction value, E is the electromagnetic interference intensity, k 4 is the weight coefficient;

[0039] Among them, each weight coefficient satisfies the following relationship:

[0040]

[0041] In the formula, k i Represents the weight coefficient of the i-th characteristic variable, i = 1, 2, 3, 4 and satisfies y is the correlation coefficient between the sum of all physical state correction values ​​and communication state correction values ​​of key nodes and the failure risk y; It represents the sum of the absolute values ​​of the correlation coefficients of all variables;

[0042] The dynamic threshold range of the physical state value of the key node is generated as follows:

[0043]

[0044] The dynamic threshold range of the key node communication status value is generated as follows:

[0045]

[0046] In the formula, is the mean of the physical state correction values, is the standard deviation of the physical state correction value, is the mean value of the communication status correction value, is the standard deviation of the communication status correction value, α and β are dynamic adjustment coefficients respectively. If the abnormal proportion of key nodes is >10%, α=1.5, β=1.5; if the abnormal proportion of key nodes is <2%, α=2.5, β=2.5; the abnormal proportion of key nodes includes the proportion of data points whose physical state values ​​and / or communication state values ​​of key nodes exceed the dynamic threshold range to the total monitored data points.

[0047] Preferably, the physical state correction value and the communication state correction value are used to calculate the fault state judgment value based on the fault state judgment model; and the mean of the fault state judgment value is calculated. and standard deviation σ μ , to determine the distribution range of the fault status judgment value

[0048] Extract the fault judgment parameter μ calculated by the fault judgment model from the historical fault data, and calculate the distribution characteristics of normal and abnormal states by segment; in the normal state, calculate the mean value of μ and standard deviation σ μ , build a distribution model for normal operation; for abnormal conditions, extract μ value samples that exceed the normal range, calculate the mean and standard deviation respectively, and build a distribution model for abnormal conditions;

[0049] Based on the distribution characteristics of μ values ​​under normal conditions, the upper and lower limits of the global preset judgment value are set, covering more than 95% of the normal operating data, and setting the global preset judgment value.

[0050] Preferably, step 4 comprises:

[0051] When the physical state value of the key node collected in real time is not within the dynamic threshold range of the physical state value and / or the communication state value is not within the dynamic threshold range of the communication state value, the key node is marked as an abnormal section, and the physical state value and communication state value of the key node collected in real time are used to obtain the real-time value of the fault state judgment of the key node based on the fault state judgment model;

[0052] When the fault state judgment real-time value of the key node is not within the distribution range of the fault state judgment value, it is determined that the key node has a fault.

[0053] A second aspect of the present invention provides a multi-dimensional state monitoring and fault prediction system for optical cables, comprising: a data acquisition module, a data correction module, a model building module and a fault determination module;

[0054] The data acquisition module is used to obtain the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable, select key nodes from each section of the optical cable, and deploy different types of status sensors at the key nodes to collect multi-dimensional status data of the optical cable, including physical status value δ and communication status value η;

[0055] A data correction module is used to correct the historical data of the multi-dimensional state data by using the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable, and to establish a dynamic threshold range of the multi-dimensional state data by using the mean and standard deviation of the corrected multi-dimensional state data;

[0056] A model building module is used to build a fitting model of multi-dimensional state data based on the working principles of different types of state sensors, and to build a fault state judgment model based on the fitting model of the multi-dimensional state data; at the same time, using the corrected multi-dimensional state data, the fault state judgment value is calculated based on the fault state judgment model, and the distribution range of the fault state judgment value is established based on the mean and standard deviation of the fault state judgment value;

[0057] A fault judgment module is used to collect the physical state value and communication state value of the key node in real time. When the physical state value is not within the dynamic threshold range of the physical state value and / or the communication state value is not within the dynamic threshold range of the communication state value, the key node is marked as an abnormal section, and the real-time value of the fault state judgment of the key node is obtained based on the fault state judgment model using the real-time collected physical state value and communication state value of the key node; when the real-time value of the fault state judgment is not within the distribution range of the fault state judgment value, it is determined that a fault occurs at the key node.

[0058] A third aspect of the present invention provides an electronic device, comprising: a processor; a memory for storing instructions executable by the processor; and the processor for executing a multi-dimensional state monitoring and fault prediction method for an optical cable.

[0059] The fourth aspect of the present invention provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is executed by a processor, the device where the storage medium is located is controlled to execute a multi-dimensional state monitoring and fault prediction method for optical cables.

[0060] Compared with the prior art, the beneficial effects of the present invention include at least:

[0061] 1. The present invention arranges integrated sensor modules for physical status and communication status at key nodes of optical cables, which can collect physical parameters such as temperature, humidity, pressure, bending radius and communication parameters such as optical power and attenuation rate of optical cables in multiple dimensions, thereby realizing comprehensive monitoring of optical cables. Compared with the traditional single data monitoring method, the present invention can more comprehensively and accurately reflect the operating status of optical cables, reducing the risk of underreporting and misjudgment;

[0062] 2. For different sections of the optical cable line, the present invention sets an initial fault judgment threshold for each section based on environmental characteristics and historical fault conditions, combined with the risk weight of the section, and dynamically optimizes these thresholds in the subsequent process, so that the monitoring system can be personalized according to the actual needs of each section, avoiding the monitoring error problem of the prior art that cannot distinguish different environmental conditions;

[0063] 3. The physical state and communication state of each section are monitored in real time through the fault judgment model, and the optimized dynamic threshold is used for state judgment. When an abnormal state is detected, a secondary screening is performed by comparing the fault state judgment parameter with the preset judgment value, so as to timely identify potential faults and trigger early warnings, thus avoiding the problem of fault expansion caused by delayed monitoring or untimely response in the existing technology;

[0064] 4. The design of the present invention combines a variety of environmental and operating data. Through intelligent and partitioned dynamic threshold optimization methods, it can reduce unnecessary false alarms and maintenance frequency, significantly reduce operation and maintenance costs, and maintain monitoring accuracy and high efficiency of fault prediction when the optical cable operating environment changes, overcoming the problems of threshold fixation and poor adaptability in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a multi-dimensional state monitoring and fault prediction flow chart of an optical cable provided in accordance with an embodiment of the present invention;

[0066] Figure 2 It is an overall system module diagram of an intelligent monitoring system for optical cables provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0068] At present, the optical cable monitoring system mainly relies on traditional fault detection methods, which can only detect faults when the optical cable is damaged or the performance is degraded, but it is difficult to effectively monitor the potential and hidden optical cable faults. In addition, the existing monitoring systems often lack the ability to continuously, real-time and comprehensively monitor the status of the optical cable, and cannot conduct in-depth analysis of the health status of the optical cable, which results in limited response speed of repair work when a fault occurs, affecting the quality of communication services.

[0069] In order to meet the above challenges, it is necessary to develop an intelligent optical cable monitoring system that can perform all-round monitoring and real-time analysis of the optical cable status. The system should be able to continuously, in real time and comprehensively monitor the physical integrity, environmental adaptability, electrical performance, mechanical performance, etc. of the optical cable, and use advanced analysis algorithms to process the monitoring data to achieve accurate evaluation and prediction of the health status of the optical cable. Through this intelligent monitoring system, the sensitivity and accuracy of fault detection can be greatly improved, and the fault repair time can be shortened, thereby ensuring the efficient and stable operation of the optical cable communication network and improving the quality of communication services.

[0070] like Figure 1 As shown, embodiment 1 of the present invention provides a multi-dimensional state monitoring and fault prediction method for an optical cable, comprising the following steps:

[0071] Step 1, obtain and use the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable to select key nodes from each section of the optical cable; deploy different types of state sensors at the key nodes to collect multi-dimensional state data of the optical cable, including physical state value δ and communication state value η;

[0072] Preferably, step 1 comprises:

[0073] The temperature change rate, relative humidity, historical fault frequency, external electromagnetic interference intensity and historical fault location constitute multiple criteria; when any section of the optical cable meets at least one criterion, it is determined to be a critical node;

[0074] According to the geographical location, environmental characteristics and historical fault distribution of the optical cable line, the operation risks of different sections of the optical cable are determined, and sections with a temperature and humidity change rate greater than the set threshold, a fault frequency greater than once a month, and external electromagnetic interference are selected as key nodes;

[0075] According to the requirements of optical cable status monitoring, determine the type of status sensors deployed at each key node, including physical status sensors and communication status sensors;

[0076] Physical state sensors and communication state sensors are integrated and installed at each key node to synchronously collect multi-dimensional state parameters; the physical state sensors include temperature sensors, pressure sensors, humidity sensors and strain gauges, which respectively collect the temperature, pressure, humidity and bending radius of the optical cable; the communication state sensors include optical power sensors, optical attenuators and high-speed data acquisition cards, which respectively measure optical power, attenuation rate and transmission rate;

[0077] The data collected by the physical state sensor and the data collected by the communication state sensor are transmitted to the central server through the data transmission module.

[0078] Preferably, the multiple criteria include:

[0079] 1) The temperature change rate is greater than ±20℃ / hour;

[0080] 2) Relative humidity is greater than 90%;

[0081] 3) The historical failure frequency is greater than once a month;

[0082] 4) The external electromagnetic interference intensity is greater than 50μT;

[0083] 5) The location of historical faults occurs repeatedly in the same section.

[0084] Preferably, according to the geographical location, environmental characteristics and historical fault distribution of the optical cable line, the operation risks of different sections are analyzed, and sections where the temperature and humidity change rate exceeds the set threshold, the fault frequency exceeds once a month, and the sections significantly affected by external electromagnetic interference are preferentially selected as key monitoring nodes, including:

[0085] In areas where temperature, humidity and air pressure fluctuations exceed the set thresholds, the physical properties of the optical cable are deformed, including: temperature and humidity changes exceeding ±20°C and relative humidity reaching more than 90%, and air pressure changes exceeding 20hPa; key nodes are set up first, and sensors are deployed to collect and monitor the physical state parameters of the optical cable in real time;

[0086] Monitor the distribution of historical faults and prioritize key nodes in areas where the same fault recurs.

[0087] Specifically, the impact of environmental characteristics:

[0088] Sections with large changes in temperature, humidity and air pressure: In areas with large fluctuations in ambient temperature, humidity and air pressure, the physical properties of optical cables (such as material strength and bending radius) are more likely to change. These environmental factors will cause thermal expansion and contraction or changes in the physical structure of the optical cable, increasing the risk of fatigue damage and performance degradation of the optical cable. For example, in sections with high humidity or large temperature differences, optical cables are more likely to age or bend due to environmental influences. Therefore, these areas should be selected as monitoring nodes in priority, and sensors should be deployed to monitor changes in physical conditions in real time.

[0089] Areas prone to natural disasters: For areas that are exposed to natural disasters such as earthquakes and floods, optical cables are more likely to be damaged or have performance fluctuations. Deploying monitoring equipment in these high-risk areas can obtain optical cable status data in a timely manner after a disaster occurs, providing support for emergency repairs.

[0090] Historical fault distribution:

[0091] Fault-prone sections: By analyzing the historical fault distribution data of the optical cable line, areas with a high frequency of faults can be identified. These sections may have a higher risk of failure due to unstable geological conditions, frequent ground construction, or aging equipment. The present invention sets monitoring nodes for these fault-prone sections, and timely warns of potential faults by monitoring abnormal changes in physical and communication state parameters.

[0092] Identification of repeated failure patterns: If a section frequently experiences similar failures (such as rapid signal attenuation, fiber breakage, etc.) under specific conditions (such as high temperature or high humidity), the section is at higher risk under similar environmental conditions. Therefore, the deployment of monitoring nodes facilitates the capture of early signs of repeated failures and optimizes subsequent maintenance and management strategies.

[0093] External interference factors at the location of the optical cable:

[0094] Areas with frequent human activities: such as areas near construction sites, traffic arteries, etc., due to the interference of human activities, optical cables are more easily damaged by external influences (such as construction vibration, equipment relocation). Therefore, the present invention selects and arranges monitoring nodes in these areas, which can track the force and vibration changes of optical cables in real time, and timely warn of potential risks caused by external interference.

[0095] Sections with special environmental conditions: Some sections may be under special environmental conditions (such as high altitudes and humid areas). The optical cables in these areas are susceptible to erosion by the harsh environment, resulting in performance degradation or hidden failures. The present invention preferentially arranges monitoring nodes in these sections, and evaluates the health status of the optical cables in real time by means of changes in physical state and communication state parameters.

[0096] Segmented optimization of monitoring resources:

[0097] Intensive monitoring of key areas: For areas with higher risks, such as those with drastic environmental changes, frequent failures, or frequent human activities, the deployment of dense monitoring nodes can effectively improve the accuracy of monitoring.

[0098] Optimized layout of low-risk areas: For low-risk sections, the density of monitoring nodes can be appropriately reduced to optimize resource allocation. This segmented layout can not only reduce the deployment and maintenance costs of the monitoring system, but also concentrate resources on intensive monitoring of high-risk sections, thereby improving the overall efficiency of the monitoring system.

[0099] Step 2, using the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable to correct the historical data of the multi-dimensional state data, and using the mean and standard deviation of the corrected historical data of the multi-dimensional state data to establish a dynamic threshold range of the multi-dimensional state data;

[0100] Preferably, step 2 comprises:

[0101] The physical state perception model is as follows:

[0102]

[0103] Where:

[0104] δ is the physical state value; T is the temperature parameter, ℃; P is the pressure parameter, Pa; S is the humidity parameter, g / m 3 ; W is the bending radius parameter, mm; 1.304, 0.79, 1.005 and 2.6 are adjustment constants; dx is the constant integral operation;

[0105] The communication status awareness model is as follows:

[0106] or

[0107] Where:

[0108] η is the communication status value; G is the optical power, W; J is the attenuation rate, dB / km; C is the transmission rate, bps; 1.66, 1.661, -1 and 0.84 are all adjustment constants;

[0109] The fault judgment model is as follows:

[0110]

[0111] Where:

[0112] μ is the fault state judgment parameter; δ is the physical state value; η is the communication state value; 1.35 and 1.03 are both adjustment constants.

[0113] Preferably, in combination with the physical state value δ and the communication state value η of the key node, the multi-dimensional state feature data of the key node is weightedly calculated according to the following formula:

[0114]

[0115] Where ω is the weighted multi-dimensional state eigenvalue; k δ and k η are the weight coefficients of the physical state value and the communication state value, satisfying k δ +k η =1;

[0116] Combine the weighted eigenvalue ω and calculate its mean And standard deviation σω, determine the initial feature distribution range of key nodes:

[0117] Step 3: Establish a fitting model for multi-dimensional state data based on the working principles of different types of state sensors, and establish a fault state judgment model based on the fitting model for multi-dimensional state data; use the corrected historical data of the multi-dimensional state data to obtain a fault state judgment value based on the fault state judgment model; and use the mean and standard deviation of the fault state judgment value to establish a distribution range of the fault state judgment value;

[0118] Preferably, step 3 comprises:

[0119] Statistics are collected based on the historical operation data of each segment, including physical status data and communication status data;

[0120] Taking the initial fault judgment threshold set in step 2 as a benchmark, the deviation analysis is performed on the physical state value δ, communication state value η and fault state judgment parameter μ in the historical data.

[0121] Calculate the difference between the actual state value and the initial threshold value to identify abnormal situations that often trigger the threshold value under specific environments or conditions. For example, under conditions of high temperature or sudden change in humidity, the physical state value δ may deviate from the initial threshold value more easily.

[0122] Analyze the statistical distribution of deviations, identify patterns or trends that frequently exceed the initial threshold, and provide specific parameter adjustment basis for threshold optimization;

[0123] Combine the environmental characteristics of each section with the state fluctuation trend to identify the state change rules under the set conditions and adjust the threshold settings;

[0124] For each section, the parameters of the physical state perception model and the communication state perception model are adjusted based on historical data to optimize the thresholds of the physical state value δ and the communication state value η, and on this basis, the threshold of the fault state judgment parameter μ is adjusted;

[0125] The optimized dynamic thresholds of each section are output, including the optimized thresholds of the physical state value δ, the communication state value η and the fault state judgment parameter μ, so that these optimized thresholds become the judgment criteria in real-time monitoring.

[0126] Preferably, step 3 comprises:

[0127] Calculate the mean of the physical state values and standard deviation σ δ , to determine the distribution range of the physical state value Calculate the mean of the communication status values and standard deviation σ η , to determine the distribution range of the communication status value

[0128] The physical state correction value is calculated using the following relationship:

[0129] δ adj =δ+k 1 ΔT+k 2 ·S+k 3 ΔP

[0130] In the formula, δ adj is the physical state correction value, ΔT is the temperature change, S is the humidity value, ΔP is the pressure change, k 1 , k 2 , k 3 All are weight coefficients;

[0131] The communication status correction value is calculated using the following relationship:

[0132] η adj =η+k 4 ·E+k 1 ΔT+k 2 ·S

[0133] Where η a dj is the communication status correction value, E is the electromagnetic interference intensity, k 4 is the weight coefficient;

[0134] Among them, each weight coefficient satisfies the following relationship:

[0135]

[0136] In the formula, ki represents the weight coefficient of the i-th characteristic variable, i = 1, 2, 3, 4 and satisfies y is the correlation coefficient between the sum of all physical state correction values ​​and communication state correction values ​​of key nodes and the failure risk y; It represents the sum of the absolute values ​​of the correlation coefficients of all variables;

[0137] The dynamic threshold range of the physical state value of the key node is generated as follows:

[0138]

[0139] The dynamic threshold range of the key node communication status value is generated as follows:

[0140]

[0141] In the formula, is the mean of the physical state correction values, is the standard deviation of the physical state correction value, is the mean value of the communication status correction value, is the standard deviation of the communication status correction value, α and β are dynamic adjustment coefficients respectively. If the abnormal proportion of key nodes is >10%, α=1.5, β=1.5; if the abnormal proportion of key nodes is <2%, α=2.5, β=2.5; the abnormal proportion of key nodes includes the proportion of data points whose physical state values ​​and / or communication state values ​​of key nodes exceed the dynamic threshold range to the total monitored data points.

[0142] Preferably, the physical state correction value and the communication state correction value are used to calculate the fault state judgment value based on the fault state judgment model; and the mean of the fault state judgment value is calculated. and standard deviation σμ to determine the distribution range of the fault state judgment value

[0143] Extract the fault judgment parameter μ calculated by the fault judgment model from the historical fault data, and calculate the distribution characteristics of normal and abnormal states by segment; in the normal state, calculate the mean value of μ and standard deviation σμ, to construct a distribution model for normal operation; for abnormal conditions, extract μ value samples that exceed the normal range, calculate the mean and standard deviation respectively, and construct a distribution model for abnormal conditions;

[0144] Based on the distribution characteristics of μ values ​​under normal conditions, the upper and lower limits of the global preset judgment value are set, covering more than 95% of the normal operating data, and setting the global preset judgment value.

[0145] Step 4, when the physical state value of the key node collected in real time is not within the dynamic threshold range of the physical state value and / or the communication state value is not within the dynamic threshold range of the communication state value, the key node is marked as an abnormal section, and the physical state value and communication state value of the key node collected in real time are used to obtain the real-time value of the fault state judgment of the key node based on the fault state judgment model; when the real-time value of the fault state judgment of the key node is not within the distribution range of the fault state judgment value, it is determined that a fault has occurred at the key node.

[0146] Preferably, step 4 comprises:

[0147] Compare the real-time collected physical status values ​​and communication status values ​​with the dynamic threshold optimized in step 3, segment by segment. If the physical or communication status value exceeds the corresponding optimized threshold, mark the segment as a possible abnormality, triggering the next step of fault judgment and analysis.

[0148] For the sections marked as possibly abnormal, their real-time physical state values ​​and communication state values ​​are input into the fault judgment model to calculate the fault state judgment parameters to assess whether there is a fault risk;

[0149] The calculated fault status judgment parameter is compared with the globally fixed preset judgment values, including 3.98, 3.981 and 3.982. If μ exceeds the preset judgment value, it is determined that there is a fault risk, otherwise the section status continues to be monitored;

[0150] When the μ value exceeds the preset judgment value, a fault warning is triggered, abnormal information is recorded and an alarm signal is sent to the monitoring center or maintenance personnel. The alarm content includes the section location, physical and communication status values ​​exceeding the threshold, fault status judgment parameters and specific information exceeding the judgment value.

[0151] Preferably, the fault state judgment parameter μ is calculated as follows:

[0152] μ=1.35·δ+1.03·η

[0153] Where:

[0154] δ represents the physical state value; η represents the communication state value; 1.35 and 1.03 are the best weight coefficients in the simulation experiment.

[0155] Example 2 of the present invention provides a multi-dimensional state monitoring and fault prediction system for optical cables, including: a data acquisition module, a data correction module, a model building module and a fault determination module;

[0156] The data acquisition module is used to obtain the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable, select key nodes from each section of the optical cable, and deploy different types of status sensors at the key nodes to collect multi-dimensional status data of the optical cable, including physical status value δ and communication status value η;

[0157] A data correction module is used to correct the historical data of the multi-dimensional state data by using the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable, and to establish a dynamic threshold range of the multi-dimensional state data by using the mean and standard deviation of the corrected multi-dimensional state data;

[0158] A model building module is used to build a fitting model of multi-dimensional state data based on the working principles of different types of state sensors, and to build a fault state judgment model based on the fitting model of the multi-dimensional state data; at the same time, using the corrected multi-dimensional state data, the fault state judgment value is calculated based on the fault state judgment model, and the distribution range of the fault state judgment value is established based on the mean and standard deviation of the fault state judgment value;

[0159] A fault judgment module is used to collect the physical state value and communication state value of the key node in real time. When the physical state value is not within the dynamic threshold range of the physical state value and / or the communication state value is not within the dynamic threshold range of the communication state value, the key node is marked as an abnormal section, and the real-time value of the fault state judgment of the key node is obtained based on the fault state judgment model using the real-time collected physical state value and communication state value of the key node; when the real-time value of the fault state judgment is not within the distribution range of the fault state judgment value, it is determined that a fault occurs at the key node.

[0160] Specifically, Figure 2 As shown, the present invention provides an intelligent monitoring system for optical cables, comprising the following components:

[0161] Each monitoring terminal 100 is arranged at each key node of the optical cable line and is used to collect the physical state parameters and communication state parameters of the optical cable in real time;

[0162] The data transmission module 200 is wirelessly connected to each monitoring terminal 100 and uses Ethernet to transmit the collected data to the central server 300;

[0163] The central server 300 is connected to the data transmission module 200 by wireless signal data, obtains the physical state parameters and the communication state parameters, respectively constructs the physical state perception model and the communication state perception model, and outputs the corresponding physical state value and the communication state value corresponding to the input physical state parameters and the communication state parameters;

[0164] The fault warning module 400 is connected to the central server 300 via wireless signal data, obtains physical status values ​​and communication status values, builds a fault judgment model, outputs fault status judgment parameters, compares preset judgment values, detects the real-time status of the optical cable, and completes the fault alarm after judgment.

[0165] Specifically, the collected physical state parameters include: a temperature parameter T obtained by a configured temperature sensor, a pressure parameter P obtained by a configured pressure sensor, a humidity parameter S obtained by a configured humidity sensor, and a bending radius parameter W obtained by a strain gauge;

[0166] It should be noted that the temperature sensor (configured on the surface), the pressure sensor (configured on the surface), the humidity sensor (configured on the surface) and the strain gauge (configured on the surface) are all applications of existing mature technologies and will not be elaborated on here.

[0167] Specifically, the collected communication status parameters include: the intensity of the optical signal in the optical cable measured by the configured optical power sensor, that is, the optical power G, the attenuation rate J measured by the configured optical attenuator in conjunction with the optical power sensor, and the transmission rate C collected by the configured high-speed data acquisition card.

[0168] It should be noted that the optical power sensor, optical attenuator and high-speed data acquisition card are all applications of existing mature technologies and will not be elaborated on here.

[0169] Furthermore, the physical state perception model constructed is specifically as follows:

[0170]

[0171] Among them, δ is the physical state value; T is the temperature parameter, ℃; P is the pressure parameter, Pa; S is the humidity parameter, g / m 3 ; W is the bending radius parameter, mm; 1.304, 0.79, 1.005 and 2.6 are adjustment constants; dx is the constant integral operation.

[0172] It should be noted that when generating the above model, T, P, S, and W are considered as four basic influencing factors. The first three basic formulas reflect the corresponding influence of temperature, pressure, and humidity. Since the bending radius parameter is the largest influencing factor among the physical state parameters, the present invention creatively uses the integral form to express the influence of the bending radius parameter. The adjustment constant in the model is an optimized constant obtained through a large number of experiments on the simulation machine, which can further improve the expression accuracy of the physical state perception model.

[0173] Furthermore, the communication status perception model constructed is specifically as follows:

[0174] or

[0175] Wherein, η is the communication status value; G is the optical power, W; J is the attenuation rate, dB / km; C is the transmission rate, bps; 1.66, 1.661, -1 and 0.84 are all adjustment constants; dx is the constant integral operation.

[0176] It should be noted that when generating the above model, G, J, and C are considered as three basic influencing factors. The first item of the model uses an integral form to express the maximum state influence of optical power under the influence of attenuation rate, the second item of the model reflects the state influence of transmission rate under the influence of attenuation rate, and the lower three items of the model reflect the state influence of transmission rate under the influence of optical power. The adjustment constant in the model is an optimized constant obtained through a large number of experiments on the simulation machine, which can further improve the expression accuracy of the physical state perception model.

[0177] Furthermore, the fault judgment model constructed is specifically as follows:

[0178]

[0179] Among them, μ is the fault state judgment parameter; δ is the physical state value; η is the communication state value; 1.35 and 1.03 are adjustment constants; dx is the constant integral operation.

[0180] Specifically, the dynamic threshold range is used to preliminarily screen whether the physical state values ​​and communication state values ​​of key nodes are abnormal, and the abnormal section is determined by judging whether they are within or outside the range; the fault judgment model is based on the state value within the dynamic threshold range, combined with historical fault data and environmental characteristics, to calculate the fault state judgment parameter μ, and by setting several key threshold points, quantify the severity and risk probability of the fault, which is used to accurately locate the fault node.

[0181] It should be noted that the last two basic formulas of the above model are expressed in integral form to amplify the impact. At the same time, compared with the physical state parameters, the communication state parameters can have a greater impact on the fault state of the optical cable.

[0182] Specifically, the preset judgment value is defined as 3.98 or 3.981 or 3.982.

[0183] Specifically, when the fault state judgment parameter is higher than a preset judgment value, it is defined as a pre-fault of the optical cable, and the fault warning module 400 issues a fault alarm.

[0184] In order to verify the beneficial effects of the present invention, an effect experiment is carried out on a simulation machine. The specific experimental process is as follows:

[0185] 1. Test preparation stage:

[0186] -The selected urban optical cable line is 50 kilometers long, including 30 joints and 4 branch points.

[0187] -Install the optical cable intelligent monitoring system of the present invention, including a multifunctional data acquisition module, a signal processing module, a wireless communication module, a monitoring center server, an intelligent analysis module and an alarm module.

[0188] 2. Data collection and processing stage:

[0189] -Turn on the multi-function data acquisition module to monitor the operation data of the optical cable in real time.

[0190] -The signal processing module processes the received data, extracts useful information, and converts it into digital signals.

[0191] -The wireless communication module transmits the processed data to the monitoring center server.

[0192] 3. Real-time monitoring and fault diagnosis stage:

[0193] -The monitoring center server receives data and performs real-time monitoring. When an abnormal situation is detected, it performs fault diagnosis and location.

[0194] - Use specific algorithms (such as Fault Tree Analysis (FTA)) to perform fault diagnosis and compare it with the actual fault situation to evaluate the accuracy and speed of fault diagnosis.

[0195] 4. Intelligent analysis and prediction stage:

[0196] -The intelligent analysis module analyzes and mines historical data to predict the operating status of the optical cable.

[0197] -Use machine learning algorithms and other methods to predict operating status and compare it with actual operating conditions to evaluate the accuracy of the prediction.

[0198] 5. Alarm and maintenance stage:

[0199] -When an abnormal situation is detected, the alarm module sends out an alarm signal to notify maintenance personnel.

[0200] -Record alarm conditions and evaluate the effectiveness of the alarm function.

[0201] 6. Evaluation of test results:

[0202] -Compare the fault diagnosis results with the actual fault conditions to evaluate the accuracy and speed of fault diagnosis.

[0203] -Compare the forecast results with the actual operation status to evaluate the accuracy of the forecast.

[0204] -Compare the alarm status with the actual maintenance status to evaluate the effectiveness of the alarm function.

[0205] 7. Test data table:

[0206] As shown in Table 1, the performance parameters and evaluation results of the optical cable intelligent monitoring system in different test stages include system installation, data collection, fault diagnosis, operation status prediction, alarm response and result evaluation. Specifically, the system was successfully installed and operated normally in the preparation stage; real-time monitoring of optical cable parameters was achieved in the data collection stage; the fault diagnosis accuracy rate reached 90% in the real-time monitoring stage; the operation status prediction accuracy rate was 85% in the intelligent analysis stage; the alarm response time in the alarm and maintenance stage was 2 minutes; the test results evaluation showed that the system as a whole met the requirements, showing the stability and accuracy of the monitoring system.

[0207] Table 1: Performance parameters

[0208]

[0209] As shown in Table 2, the diagnostic accuracy and response speed of the optical cable intelligent monitoring system under different fault types include information such as fault type, fault occurrence and diagnosis time, actual fault location, diagnostic error, and diagnostic accuracy. The data shows that the system has a diagnostic error of 0 meters and a diagnostic accuracy of 100% in the case of joint failure, voltage fluctuation, and current anomaly, while the diagnostic position error of temperature anomaly is 50 meters and the accuracy rate reaches 90%. The system completes the diagnosis within 1 minute after the fault occurs, indicating that the monitoring system performs well in fault detection accuracy and response speed.

[0210] Table 2: Fault diagnosis accuracy and speed evaluation table

[0211]

[0212] Compared with the prior art, the beneficial effects of the present invention include at least:

[0213] 1. The present invention arranges integrated sensor modules for physical status and communication status at key nodes of optical cables, which can collect physical parameters such as temperature, humidity, pressure, bending radius and communication parameters such as optical power and attenuation rate of optical cables in multiple dimensions, thereby realizing comprehensive monitoring of optical cables. Compared with the traditional single data monitoring method, the present invention can more comprehensively and accurately reflect the operating status of optical cables, reducing the risk of underreporting and misjudgment;

[0214] 2. For different sections of the optical cable line, the present invention sets an initial fault judgment threshold for each section based on environmental characteristics and historical fault conditions, combined with the risk weight of the section, and dynamically optimizes these thresholds in the subsequent process, so that the monitoring system can be personalized according to the actual needs of each section, avoiding the monitoring error problem of the prior art that cannot distinguish different environmental conditions;

[0215] 3. The physical state and communication state of each section are monitored in real time through the fault judgment model, and the optimized dynamic threshold is used for state judgment. When an abnormal state is detected, a secondary screening is performed by comparing the fault state judgment parameter with the preset judgment value, so as to timely identify potential faults and trigger early warnings, thus avoiding the problem of fault expansion caused by delayed monitoring or untimely response in the existing technology;

[0216] 4. The design of the present invention combines a variety of environmental and operating data. Through intelligent and partitioned dynamic threshold optimization methods, it can reduce unnecessary false alarms and maintenance frequency, significantly reduce operation and maintenance costs, and maintain monitoring accuracy and high efficiency of fault prediction when the optical cable operating environment changes, overcoming the problems of threshold fixation and poor adaptability in the prior art.

[0217] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0218] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A multi-dimensional state monitoring and fault prediction method for an optical cable, wherein the optical cable comprises a plurality of sections, characterized in that: The following steps are involved: Step 1, obtain and use the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable to select key nodes from each section of the optical cable; deploy different types of state sensors at the key nodes to collect multi-dimensional state data of the optical cable, including physical state value δ and communication state value η; Step 2, using the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable to correct the historical data of the multi-dimensional state data, and using the mean and standard deviation of the corrected historical data of the multi-dimensional state data to establish a dynamic threshold range of the multi-dimensional state data; Step 3: Establish a fitting model for multi-dimensional state data based on the working principles of different types of state sensors, and establish a fault state judgment model based on the fitting model for the multi-dimensional state data; and obtain a fault state judgment value based on the fault state judgment model using the corrected historical data of the multi-dimensional state data; The distribution range of the fault state judgment value is established using the mean and standard deviation of the fault state judgment value; Step 4, when the physical state value of the key node collected in real time is not within the dynamic threshold range of the physical state value and / or the communication state value is not within the dynamic threshold range of the communication state value, the key node is marked as an abnormal section, and the physical state value and communication state value of the key node collected in real time are used to obtain the real-time value of the fault state judgment of the key node based on the fault state judgment model; when the real-time value of the fault state judgment of the key node is not within the distribution range of the fault state judgment value, it is determined that a fault has occurred at the key node.

2. The multi-dimensional condition monitoring and fault prediction method of an optical cable according to claim 1, characterized in that: In step 1, the environmental characteristic data of the optical cable include the temperature change rate and relative humidity; the historical fault frequency and historical fault location are extracted based on the historical fault distribution probability; the operation risk index includes the external electromagnetic interference intensity; The temperature change rate, relative humidity, historical fault frequency, external electromagnetic interference intensity and historical fault location are used to form multiple criteria. When any section of the optical cable meets at least one criterion, it is determined to be a critical node. According to the requirements of optical cable status monitoring, determine the type of status sensors arranged at each key node, including physical status sensors and communication status sensors; physical status sensors include temperature sensors, pressure sensors, humidity sensors and strain gauges, which respectively collect the temperature, pressure, humidity and bending radius of the optical cable; communication status sensors include optical power sensors, optical attenuators and high-speed data acquisition cards, which respectively measure optical power, attenuation rate and transmission rate; Physical status sensors and communication status sensors are integrated and installed on each key node. The data collected by the physical status sensors and the data collected by the communication status sensors constitute multi-dimensional status data. The multi-dimensional status data is transmitted to the central server through the data transmission module, and the initial status data set of the key nodes is established in the central server.

3. The multi-dimensional condition monitoring and fault prediction method of an optical cable according to claim 2, characterized in that: Several criteria include: 1) The temperature change rate is greater than ±20℃ / hour; 2) Relative humidity is greater than 90%; 3) The historical failure frequency is greater than once a month; 4) The external electromagnetic interference intensity is greater than 50μT; 5) The location of historical faults occurs repeatedly in the same section.

4. The multi-dimensional condition monitoring and fault prediction method of an optical cable according to claim 2, characterized in that: Based on the working principle of the physical state sensor, the fitting model of the physical state value satisfies the following relationship: In the formula, δ is the physical state value; T is the temperature; P is the pressure; S is the humidity; W is the bending radius; dx is the constant integral operation; Based on the working principle of the communication status sensor, the fitting model of the communication status value satisfies the following relationship: In the formula, η is the communication state value; G is the optical power; J is the attenuation rate; C is the transmission rate; Based on the fitting model of the physical state value and the fitting model of the communication state value, the fault state judgment model satisfies the following relationship: Where μ is the fault state judgment parameter.

5. The multi-dimensional condition monitoring and fault prediction method of an optical cable according to claim 4, characterized in that: Step 3 includes: Calculate the mean of the physical state values and standard deviation σ δ , to determine the distribution range of the physical state value Calculate the mean of the communication status values and standard deviation σ η , to determine the distribution range of the communication status value The physical state correction value is calculated using the following relationship: d adj =δ+k1·ΔT+k2·S+k3·ΔP In the formula, δ adj is the physical state correction value, ΔT is the temperature change, S is the humidity value, ΔP is the pressure change, k1, k2, k3 are all weight coefficients; The communication status correction value is calculated using the following relationship: or adj =η+k4·E+k1·ΔT+k2·S Where η adj is the communication status correction value, E is the electromagnetic interference intensity, and k4 is the weight coefficient; Among them, each weight coefficient satisfies the following relationship: In the formula, k i Represents the weight coefficient of the i-th characteristic variable, i = 1, 2, 3, 4 and satisfies y is the correlation coefficient between the sum of all physical state correction values ​​and communication state correction values ​​of key nodes and the failure risk y; It represents the sum of the absolute values ​​of the correlation coefficients of all variables; The dynamic threshold range of the physical state value of the key node is generated as follows: The dynamic threshold range of the key node communication status value is generated as follows: In the formula, is the mean of the physical state correction values, is the standard deviation of the physical state correction value, is the mean value of the communication status correction value, is the standard deviation of the communication status correction value, α and β are dynamic adjustment coefficients respectively. If the abnormal proportion of key nodes is >10%, α=1.5, β=1.5; if the abnormal proportion of key nodes is <2%, α=2.5, β=2.5; the abnormal proportion of key nodes includes the proportion of data points whose physical state values ​​and / or communication state values ​​of key nodes exceed the dynamic threshold range to the total monitored data points.

6. The multi-dimensional condition monitoring and fault prediction method of an optical cable according to claim 5, characterized in that: Calculating a fault state judgment value based on a fault state judgment model using the physical state correction value and the communication state correction value; Calculate the mean of the fault status judgment values and standard deviation σ μ , to determine the distribution range of the fault status judgment value Extract the fault judgment parameter μ calculated by the fault judgment model from the historical fault data, and calculate the distribution characteristics of normal state and abnormal state by section; Under normal conditions, calculate the mean value of μ and standard deviation σ μ , build a distribution model for normal operation; for abnormal conditions, extract μ value samples that exceed the normal range, calculate the mean and standard deviation respectively, and build a distribution model for abnormal conditions; Based on the distribution characteristics of μ values ​​under normal conditions, the upper and lower limits of the global preset judgment value are set, covering more than 95% of the normal operating data, and setting the global preset judgment value.

7. The multi-dimensional condition monitoring and fault prediction method of an optical cable according to claim 6, characterized in that: Step 4 includes: When the physical state value of the key node collected in real time is not within the dynamic threshold range of the physical state value and / or the communication state value is not within the dynamic threshold range of the communication state value, the key node is marked as an abnormal section, and the physical state value and communication state value of the key node collected in real time are used to obtain the real-time value of the fault state judgment of the key node based on the fault state judgment model; When the fault state judgment real-time value of the key node is not within the distribution range of the fault state judgment value, it is determined that the key node has a fault.

8. A multi-dimensional condition monitoring and fault prediction system for optical cables, comprising: Data acquisition module, data correction module, model building module and fault determination module; characterized in that: The data acquisition module is used to obtain the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable, select key nodes from each section of the optical cable, and deploy different types of status sensors at the key nodes to collect multi-dimensional status data of the optical cable, including physical status value δ and communication status value η; A data correction module is used to correct the historical data of the multi-dimensional state data by using the environmental characteristic data, historical fault distribution probability and operation risk index of the optical cable, and to establish a dynamic threshold range of the multi-dimensional state data by using the mean and standard deviation of the corrected multi-dimensional state data; A model building module is used to build a fitting model of multi-dimensional state data based on the working principles of different types of state sensors, and to build a fault state judgment model based on the fitting model of the multi-dimensional state data; at the same time, using the corrected multi-dimensional state data, the fault state judgment value is calculated based on the fault state judgment model, and the distribution range of the fault state judgment value is established based on the mean and standard deviation of the fault state judgment value; A fault judgment module is used to collect the physical state value and communication state value of the key node in real time. When the physical state value is not within the dynamic threshold range of the physical state value and / or the communication state value is not within the dynamic threshold range of the communication state value, the key node is marked as an abnormal section, and the real-time value of the fault state judgment of the key node is obtained based on the fault state judgment model using the real-time collected physical state value and communication state value of the key node; when the real-time value of the fault state judgment is not within the distribution range of the fault state judgment value, it is determined that a fault occurs at the key node.

9. An electronic device, characterized in that: The electronic device comprises: a processor; a memory for storing instructions executable by the processor; the processor is used to execute the multi-dimensional state monitoring and fault prediction method for optical cables described in any one of claims 1 to 7.

10. A readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed by a processor, the device where the storage medium is located is controlled to execute a multi-dimensional condition monitoring and fault prediction method for an optical cable as described in any one of claims 1 to 7.

Images