Cable node thermal anomaly identification and fault location method based on FBG sensing

Abstract

The invention provides a cable node thermal anomaly identification and fault location method based on FBG sensing, and relates to the technical field of power monitoring, and the method comprises the steps: generating a temperature anomaly judgment threshold value based on the environment data and external data of each monitoring node; comparing the temperature data of the monitoring node with a temperature anomaly judgment threshold value, judging whether the temperature data exceeds a limit or not, and if the temperature data exceeds the limit, executing a locking process and obtaining a locked node; the method comprises the following steps: respectively executing chain type abnormal extension identification on two sides of a locking node by taking the locking node as a starting point, sequentially detecting temperature data of monitoring nodes adjacent to the locking node, and if the adjacent monitoring nodes meet an abnormal extension judgment condition, bringing the monitoring nodes into an abnormal link and continuing to detect the next adjacent monitoring node. By setting a chain type abnormal extension mechanism, states of surrounding nodes can be dynamically tracked from an initial thermal abnormal node, and an abnormal link which truly reflects a thermal diffusion trend is formed, so that accurate identification of a thermal abnormal propagation path is realized.

Description

Technical Field

[0001] This application relates to the field of defect identification and detection technology, specifically to a method for identifying and locating thermal anomalies in cable nodes based on FBG sensing. Background Technology

[0002] According to CN201910644626.9, a method for fault location of distributed multi-node collaborative fiber Bragg grating sensors is proposed. The main technical principle of this patent is: for abnormal states of fault data in FBG sensor networks, an ant colony algorithm and support vector machine technology are combined to construct a spectral distortion demodulation model and a node redundancy prediction model, respectively, to identify changes in the center wavelength and distortion parameters of sensor nodes, thereby determining the fault type and achieving fault location. The core of the published document lies in the joint diagnosis through distortion spectral modeling and machine learning algorithms, ultimately determining the faulty node based on spectral thresholds and model output errors.

[0003] However, the above solution has the following problems in practical applications:

[0004] 1. This technical solution focuses on static diagnosis based on the abnormal signal itself, without considering the chain-like thermal anomaly evolution mechanism caused by thermal diffusion or state linkage between cable nodes. This results in coarse detection granularity and delayed response for thermal anomaly events, making it difficult to achieve dynamic and multi-node collaborative sensing.

[0005] 2. While the publicly available technology can identify single-point faults, it lacks an effective mechanism to distinguish between partial and full-segment fault sections in scenarios where multiple consecutive nodes exhibit coordinated anomalies. This limits the ability to spatially locate regional cable faults and also makes it impossible to track the dynamic changes in fault boundaries caused by the evolution of abnormal links over time. Summary of the Invention

[0006] This application provides a method for identifying and locating thermal anomalies in cable nodes based on FBG sensing, aiming to solve the problems of poor collaborative identification of thermal anomalies in cable nodes and difficulty in dividing continuous abnormal sections in related technologies.

[0007] To achieve the above objectives, this application provides a method for identifying and locating thermal anomalies in cable nodes based on FBG sensing, comprising the following:

[0008] Acquire temperature data from each monitoring node.

[0009] Based on environmental information and external data from monitoring nodes, a threshold for determining temperature anomalies is generated.

[0010] The temperature data of the monitoring node is compared with the temperature anomaly judgment threshold to determine whether it exceeds the limit. If it does, the locking process is executed and the node is locked.

[0011] Starting from the locked node, chain-like anomaly extension identification is performed to both sides of the locked node. The temperature data of the monitoring nodes adjacent to the locked node are detected in sequence. If the adjacent monitoring node meets the anomaly extension judgment condition, the monitoring node is included in the anomaly link and the next adjacent monitoring node is detected. When a monitoring node that does not meet the anomaly extension judgment condition is detected for the first time in a series of adjacent monitoring nodes, or when the detection range has been extended to the starting point or ending point of the current monitoring node, the extension in that direction is stopped, thereby forming an anomaly link composed of the locked node and the continuous monitoring nodes on both sides of the locked node.

[0012] The fault area is determined by the distribution of abnormal links along the cable. If the abnormal link does not extend to the boundary node, it corresponds to a partial fault section; if the abnormal link extends to any boundary node, it corresponds to the entire fault section.

[0013] Preferably, the temperature anomaly detection threshold for each monitoring node is generated based on environmental and external data from the monitoring nodes, specifically including:

[0014] The temperature baseline values ​​for each monitoring node are calculated based on environmental data, including the multi-year average air temperature, historical operating load, and geographical environmental parameters of the monitoring nodes. The calculation formula is as follows:

[0015] T base,i =η1·T clim,i +η2·T load,i +η3·T alt,i

[0016] In the formula, T base,i It is the temperature reference value of monitoring node i, T clim,i It is the multi-year average temperature (T) of the area where monitoring node i is located. load,i It is the average load temperature rise value of monitoring node i during its historical operation, T alt,i η1, η2, and η3 are the correction values ​​of the geographical environment temperature in the area where monitoring node i is located. The correction values ​​of the geographical environment temperature are obtained by linearly combining the altitude and latitude values ​​in the geographical environment parameters. η1, η2, and η3 are the weight coefficients of each influencing factor, and their values ​​range from η1 + η2 + η3 = 1.

[0017] The temperature baseline values ​​of each monitoring node are adaptively corrected using external data to obtain adaptive correction results; the external data includes real-time environmental parameters of each monitoring node, historical temperature curves, and temperature rise trends of neighboring nodes.

[0018] Temperature anomaly detection thresholds for each monitoring node are generated based on the temperature baseline value and adaptive correction results.

[0019] Preferably, external data is used to adaptively correct the temperature baseline values ​​of each monitoring node to obtain the adaptive correction result, calculated using the following formula:

[0020]

[0021] In the formula, ΔT i It is the adaptive correction result of monitoring node i, T base,i w1·(T) represents the baseline temperature value of monitoring node i. env,i -T env,avg,i ) is the first correction term, (T) env,i -T env,avg,i () represents the real-time ambient temperature T env,i Compared with historical average ambient temperature T env,avg,i The difference is used to reflect the immediate impact of the current environment on the temperature baseline value, and w1 is the weighting coefficient of the first correction term. The second correction term represents the historical temperature change rate of monitoring node i, used to reflect the correction of the monitoring node's own temperature trend to the baseline. w2 is the weighting coefficient of the second correction term. The third correction term represents the average temperature change rate of neighboring nodes, where n is the number of neighboring nodes, used to reflect the potential impact of the surrounding environment on the baseline, and w3 is the weighting coefficient of the third correction term; the values ​​of w1, w2, and w3 are in the range of: w1 + w2 + w3 = 1.

[0022] Preferably, the formula for generating the temperature anomaly determination threshold is:

[0023] T thresh,i =T base,i +ΔT i

[0024] In the formula, T thresh,i This is the temperature anomaly detection threshold for monitoring node i, used to compare with the temperature data of the monitoring node to determine whether a temperature anomaly exists; T base,i The baseline temperature of monitoring node i is calculated based on geographical environment type parameters, historical climate data, and load distribution characteristics. It represents the baseline temperature reference value of the monitoring node under typical climatic conditions; ΔT i It is the adaptive correction result of monitoring node i.

[0025] Preferably, the locking process specifically includes:

[0026] The locking process is implemented based on a multi-condition fusion mechanism, which includes real-time temperature difference judgment, heating rate judgment, and short-term fluctuation elimination mechanism. The execution order of the locking process is as follows: first, the real-time temperature difference judgment is executed, then the heating rate judgment is executed, and the short-term fluctuation elimination mechanism is continuously executed throughout the judgment process to eliminate false abnormal signals in real time.

[0027] The real-time temperature difference judgment refers to calculating the difference between the temperature value of the current monitoring node and the average temperature of multiple neighboring monitoring nodes in normal state, and comparing the obtained temperature difference value with the set dynamic difference threshold. When the temperature difference value exceeds the dynamic difference threshold, the monitoring node is determined to meet the abnormal temperature difference amplitude condition. The monitoring nodes that meet the abnormal temperature difference amplitude condition are recorded in the candidate locked node set.

[0028] The heating rate determination involves performing time-series analysis on the temperature data of the current monitoring node over multiple consecutive sampling periods, calculating the temperature change rate per unit time, and comparing the temperature change rate with a heating rate threshold. When the change rate exceeds the heating rate threshold, the monitoring node is determined to meet the abnormal heating rate condition. Monitoring nodes that meet the abnormal heating rate condition are then selected from the candidate locked node set and confirmed as locked nodes.

[0029] The short-term fluctuation rejection mechanism performs a temperature fluctuation persistence judgment when the temperature of a certain monitoring node exceeds the temperature anomaly judgment threshold. Abnormal temperatures that only occasionally exceed the threshold within the continuous sampling period are judged as pseudo-abnormal signals, and the monitoring node is not set as a locked node in this case. Monitoring nodes that are not defined as pseudo-abnormal signals but exceed the temperature anomaly judgment threshold are handed over to the heating rate judgment, and the locked node is obtained based on the heating rate judgment.

[0030] Preferably, the abnormal extension determination conditions include the following two conditions, which must be met simultaneously:

[0031] The first condition is that the temperature value of the adjacent monitoring node to be detected exceeds the corresponding temperature anomaly judgment threshold, and the temperature difference with the previous monitoring node is within the set extended temperature difference tolerance range.

[0032] The second condition is that the heating rate of the adjacent monitoring node to be detected has a positive correlation with the locked node or adjacent node, and the rate difference does not exceed the set rate deviation threshold.

[0033] If the monitoring node to be detected does not meet any of the above abnormal extension judgment conditions, it is judged as a monitoring node that does not meet the abnormal extension judgment conditions; if the temperature value of the monitoring node is lower than the corresponding temperature abnormal judgment threshold, or the temperature rise trend is downward or discontinuous, it is considered that the extension requirements are not met, and the extension in that direction is terminated.

[0034] Preferably, a dynamic fault segment switching mechanism is also included:

[0035] In the process of defining the segment corresponding to the node range covered by the abnormal link as a partial fault segment, or defining the entire monitoring segment as a full fault segment, a dynamic conversion mechanism for fault segments is set up; after the initial judgment of the fault area, the temperature change trend of the boundary nodes at both ends of the abnormal link is continuously monitored. When the boundary nodes are in an abnormal temperature state for multiple consecutive monitoring cycles and meet the extension judgment conditions, the originally defined partial fault segment is updated to a full fault segment.

[0036] Conversely, when the abnormal link has extended to the boundary node of the monitoring node, but the temperature value of the boundary node recovers to below the temperature anomaly judgment threshold within the continuous monitoring period and does not meet the extension trend condition, the originally defined full-segment fault section is updated to a partial fault section.

[0037] The beneficial effects of this application are as follows:

[0038] 1. By setting up a chain-like anomaly extension identification mechanism, this application can dynamically track the heat diffusion path starting from the locked node, identify abnormal links, and dynamically adjust the extension range in combination with the changes in the state of the boundary nodes, which significantly improves the system's perception continuity and fault response speed for the cable heat anomaly propagation process.

[0039] 2. This application proposes a dynamic conversion mechanism for fault segments based on the abnormal link structure. It can determine whether a local fault has evolved into a full-segment fault based on the continuous state of the abnormal boundary node. It supports upgrading the fault area from "partial segment" to "full segment" or adjusting it in the opposite direction, thereby achieving accurate determination and dynamic maintenance of the fault space boundary and improving the applicability of the overall diagnosis.

[0040] 3. Based on the multi-year average air temperature, historical operating load and geographical environmental parameters of the monitoring nodes, a temperature benchmark value for the monitoring nodes is constructed. Based on the temperature benchmark value and the adaptive correction result, a temperature anomaly judgment threshold for each node is generated, which solves the misjudgment problem caused by the uniform setting of thresholds in the existing technology and realizes adaptive adjustment of judgment for different regional environments.

[0041] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart illustrating a method for identifying and locating thermal anomalies in cable nodes based on FBG sensing, provided in an embodiment of this application. Detailed Implementation

[0044] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0045] Please see Figure 1 , Figure 1 This document provides a flowchart of a method for identifying and locating thermal anomalies in cable nodes based on FBG sensing, as an embodiment of this application.

[0046] In this embodiment, a method for identifying and locating thermal anomalies in cable nodes based on FBG sensing includes steps S10, S20, S30, 40, and 50.

[0047] Step S10: Obtain temperature data from each monitoring node, specifically including:

[0048] Several FBG temperature sensors are deployed along the cable, dividing the cable route into multiple monitoring nodes. Temperature data for each monitoring node is acquired using the FBG temperature sensors. When deploying the FBG temperature sensors, the monitoring nodes are first determined based on the cable's physical structure and total length, with one monitoring node every 50 meters. The spacing is adjusted according to the cable's importance, environmental complexity, and monitoring accuracy requirements. For critical cables or complex environments, the spacing is shortened to 30 meters. FBG temperature sensors are then installed at each monitoring node according to the determined spacing. During installation, ensure the FBG temperature sensors are in close contact with the cable surface and secure them using a dedicated clamping device to guarantee accurate temperature conduction. After installation, the FBG temperature sensors are connected in series via optical fiber to ensure real-time transmission and processing of temperature data. After installation, all sensors are calibrated using a standard temperature source, and the deviation between the initial measurement value and the standard value is recorded to ensure measurement accuracy is controlled within ±0.5℃. The calibrated data is then used for subsequent data processing.

[0049] It should be noted that FBG is the abbreviation for Fiber Bragg Grating.

[0050] Step S20: Based on the environmental data and external data of each monitoring node, generate a temperature anomaly judgment threshold, specifically including:

[0051] The temperature baseline values ​​for each monitoring node are calculated based on environmental data, including the multi-year average air temperature, historical operating load, and geographical environmental parameters of the monitoring nodes. The calculation formula is as follows:

[0052] T base,i =η1·T clim,i +η2·T load,i +η3·T alt,i

[0053] In the formula, T base,i It is the temperature reference value of monitoring node i, T clim,i It is the multi-year average temperature (T) of the area where monitoring node i is located. load,i It is the average load temperature rise value of monitoring node i during its historical operation, T alt,i η1, η2, and η3 are the correction values ​​of the geographical environment temperature in the area where monitoring node i is located. The correction values ​​of the geographical environment temperature are obtained by linearly combining the altitude and latitude values ​​in the geographical environment parameters. η1, η2, and η3 are the weight coefficients of each influencing factor, and their values ​​range from η1 + η2 + η3 = 1.

[0054] It should be noted that the correction value for geographical ambient temperature is calculated as follows:

[0055] T alt,i =α·(altitude) i -altitude 基准 )+β·(latitude) i -latitude 基准 )

[0056] In the formula, α represents the altitude influence coefficient, β represents the latitude influence coefficient, and altitude... 基准 and latitude 基准 These represent the baseline reference values ​​for altitude and latitude, respectively, obtained by taking the average altitude and latitude of the monitoring nodes.

[0057] The temperature baseline values ​​of each monitoring node are adaptively corrected using external data to obtain the adaptive correction results; the external data includes the real-time environmental parameters of each monitoring node, historical temperature curves, and temperature rise trends of neighboring nodes.

[0058] Temperature anomaly detection thresholds for each monitoring node are generated based on the temperature baseline value and adaptive correction results.

[0059] It should be noted that the calculation formula for adaptive correction is as follows:

[0060]

[0061] In the formula, ΔT i It is the adaptive correction result of monitoring node i, T base,i w1·(T) represents the baseline temperature value of monitoring node i. env,i -T env,avg,i ) is the first correction term, (T) env,i -T env,avg,i () represents the real-time ambient temperature T env,i Compared with historical average ambient temperature T env,avg,i The difference is used to reflect the immediate impact of the current environment on the temperature baseline value, and w1 is the weighting coefficient of the first correction term. The second correction term represents the historical temperature change rate of monitoring node i, used to reflect the correction of the monitoring node's own temperature trend to the baseline. w2 is the weighting coefficient of the second correction term. The third correction term represents the average temperature change rate of neighboring nodes, where n is the number of neighboring nodes, used to reflect the potential impact of the surrounding environment on the baseline, and w3 is the weighting coefficient of the third correction term; the values ​​of w1, w2, and w3 are in the range of: w1 + w2 + w3 = 1.

[0062] It should be noted that the formula for generating the temperature anomaly detection threshold is as follows:

[0063] T thresh,i =T base,i +ΔT i

[0064] In the formula, T thresh,i This is the temperature anomaly detection threshold for monitoring node i, used to compare with the temperature data of the monitoring node to determine whether a temperature anomaly exists; T base,i It is the temperature reference value of monitoring node i, ΔT i It is the adaptive correction result of monitoring node i.

[0065] Step S30: Compare the temperature data of the monitoring node with the temperature anomaly judgment threshold to determine whether it exceeds the limit. If it exceeds the limit, execute the locking process and lock the node, which specifically includes:

[0066] The temperature data of each monitoring node is compared and analyzed with the corresponding temperature anomaly judgment threshold. When the temperature data of a certain monitoring node exceeds the corresponding temperature anomaly judgment threshold, the monitoring node is determined to be a temperature anomaly node, and the temperature anomaly node is handed over to the locking process. The result of the locking process determines whether to set it as a locked node.

[0067] It should be noted that the locking process specifically includes:

[0068] The locking process is implemented based on a multi-condition fusion mechanism, which includes real-time temperature difference judgment, heating rate judgment, and short-term fluctuation elimination mechanism. The execution order of the locking process is as follows: first, the real-time temperature difference judgment is executed, then the heating rate judgment is executed, and the short-term fluctuation elimination mechanism is continuously executed throughout the judgment process to eliminate false abnormal signals in real time.

[0069] Real-time temperature difference judgment refers to calculating the difference between the temperature value of the current monitoring node and the average temperature of multiple neighboring monitoring nodes in normal condition, and comparing the resulting temperature difference value with a set dynamic difference threshold. When the temperature difference value exceeds the dynamic difference threshold, the monitoring node is determined to meet the abnormal temperature difference amplitude condition; the monitoring nodes that meet the abnormal temperature difference amplitude condition are recorded in the candidate locked node set.

[0070] It should be noted that the dynamic difference threshold is adaptively generated based on the node's environment and load status, and the calculation formula is as follows:

[0071]

[0072] In the formula, δ0 represents the dynamic difference threshold for monitoring node i, in °C; δ0 represents the set base difference threshold, which is set to 2 °C here; σ env,i σ represents the standard deviation of ambient temperature in the area where monitoring node i is located, calculated based on environmental data from the past 24 hours; load,i μ represents the historical standard deviation of the load corresponding to monitoring node i. i v i These represent the environmental disturbance coefficient and the load disturbance coefficient, respectively, reflecting the degree of influence of environmental changes and load changes on the threshold. They are set by regression optimization based on historical data, and the value range here is (0.1~0.5).

[0073] The temperature rise rate determination involves performing time series analysis on the temperature data of the current monitoring node over multiple consecutive sampling periods, calculating the temperature change rate per unit time, and comparing the temperature change rate with a temperature rise rate threshold. When the change rate exceeds the temperature rise rate threshold, the monitoring node is determined to meet the abnormal temperature rise rate condition. Monitoring nodes that meet the abnormal temperature rise rate condition are then selected from the candidate locked node set and confirmed as locked nodes.

[0074] It should be noted that the process for obtaining the heating rate threshold is as follows:

[0075] A1. Extract the basic standard value. Based on the design parameters of the cable along the route, extract the rated temperature rise response curve. Then, obtain the maximum allowable temperature rise rate per unit time under stable load conditions from the rated temperature rise response curve. The unit of the temperature rise rate is ℃ / h. Define the maximum temperature rise rate per unit time as the basic standard value for the temperature rise rate threshold, denoted as V.base V base It is a safety benchmark based on cable design specifications, providing a basic reference value for subsequent calculation of the heating rate threshold, ensuring that the heating rate threshold will not be lower than the design safety standard.

[0076] A2. Analyze the historical temperature rise waveform characteristics and determine the rate fluctuation range. Perform time series analysis on the temperature data of the monitoring nodes over past operating cycles to extract historical temperature rise waveform characteristics, including the following parameters:

[0077] The average temperature rise slope reflects the typical temperature rise rate of the monitoring node under normal operating conditions; the peak rate of change reflects the maximum temperature rise rate that the monitoring node may experience under sudden load changes; the fluctuation range reflects the normal range of temperature rise rate variation. Using a piecewise linear fitting method, the rate fluctuation range of the monitoring node under typical working load conditions is calculated based on the average temperature rise slope, peak rate of change, and fluctuation range, and denoted as [V]. min V max Historical temperature rise waveform characteristics are used to characterize the actual operating behavior of monitoring nodes, with the rate fluctuation range [V]. min V max This provides historical data support for the heating rate threshold, which can then be used as a reference range for dynamic adjustment in subsequent steps, ensuring that the heating rate threshold reflects the normal operating characteristics of the monitoring node.

[0078] A3. Construct an association template and use K-means clustering analysis to classify the historical temperature rise waveform characteristics of monitoring nodes, identifying typical temperature rise behaviors under different operating modes. Typical temperature rise behaviors include normal operation, load abrupt changes, and seasonal variations. Each operating mode corresponds to an association template, which records the typical temperature rise rate characteristics under that mode, including the average temperature rise slope, peak change rate, and fluctuation range. The association template links historical temperature rise waveform characteristics with specific operating modes, providing a patterned historical reference benchmark for subsequent dynamic adjustments, enabling targeted adjustments to the temperature rise rate threshold based on the current operating mode.

[0079] A4. Calculate the heating rate threshold and select the corresponding associated template from the existing associated templates based on the current cable operating mode.

[0080] Extract historical reference values ​​and obtain typical heating rate characteristics under this mode from the selected associated templates, denoted as historical reference value V. hist .

[0081] Based on the temperature change indicators of the current monitoring nodes, calculate the adjustment factor z to obtain the adjustment factor z;

[0082] Based on adjustment factor z and historical reference value V hist Basic standard value Vbase The heating rate threshold is calculated using the following formula:

[0083] V thresh =V base +z·(V hist -V base )

[0084] In the formula, z represents the adjustment factor, and V hist V represents the historical reference value. base V represents the base reference value. thresh V represents the threshold for the heating rate. thresh The value must be within the rate fluctuation range [V] min V max If the value exceeds the specified limit, the boundary value will be used.

[0085] It should be noted that the adjustment factor ranges from [0,1]. The following indicators should be referenced when determining the specific value:

[0086] The first indicator is the frequency of short-term temperature fluctuations. If the fluctuation frequency is high, it indicates an anomaly, and it is necessary to reduce z to increase the sensitivity of the heating rate threshold.

[0087] The second indicator is the load disturbance intensity. When the load changes significantly, a more relaxed heating rate is allowed, and z increases accordingly.

[0088] The third indicator is the synchronicity of adjacent nodes. If adjacent nodes also show a similar heating trend, it may be a normal phenomenon. In this case, z can be increased to relax the heating rate threshold.

[0089] Following steps A1 to A4, the temperature rise rate threshold is dynamically calculated by integrating design standards, historical data, and current operating status. This ensures that the temperature rise rate threshold meets safety requirements and adapts to dynamic changes during actual operation. New temperature data is collected weekly, and historical temperature rise waveform characteristics, rate fluctuation ranges, and associated templates are updated. The temperature rise rate threshold is then recalculated to ensure that it reflects long-term changes in cable operating characteristics and maintains dynamic adaptability.

[0090] The short-term fluctuation rejection mechanism performs a temperature fluctuation persistence judgment when the temperature of a certain monitoring node exceeds the temperature anomaly judgment threshold. Abnormal temperatures that only occasionally exceed the threshold within the continuous sampling period are judged as pseudo-abnormal signals, and the monitoring node is not set as a locked node in this case. Monitoring nodes that are not defined as pseudo-abnormal signals but exceed the temperature anomaly judgment threshold are handed over to the heating rate judgment, and the locked node is determined based on the heating rate judgment.

[0091] It should be noted that the short-term fluctuation elimination mechanism is based on the continuity criterion and implements the suppression of false alarms of high-frequency jitter or instantaneous spikes. The short-term fluctuation elimination mechanism sets the minimum number of duration periods to 3. If there is only one instance of exceeding the threshold in the last 3 sampling periods (each sampling period is 5 minutes) and there are no continuous anomalies, it is regarded as a short-term fluctuation and a sliding window is used to eliminate it within the time window.

[0092] Step S40: Starting from the locked node, perform chain-like anomaly extension identification to both sides of the locked node, sequentially detecting the temperature data of monitoring nodes adjacent to the locked node. If an adjacent monitoring node meets the anomaly extension judgment condition, the monitoring node is included in the anomaly link and the detection of the next adjacent monitoring node continues. When a monitoring node that does not meet the anomaly extension judgment condition is detected for the first time in a series of adjacent monitoring nodes, or when the detection range has expanded to the starting point or ending point of the current monitoring node, the extension in that direction is stopped, thereby forming an anomaly link composed of the locked node and the continuous monitoring nodes on both sides of the locked node, specifically including:

[0093] The criteria for determining abnormal extension include the following two conditions, which must be met simultaneously:

[0094] The first condition is that the temperature value of the adjacent monitoring node to be detected exceeds the corresponding temperature anomaly judgment threshold, and the temperature difference with the previous monitoring node is within the set extended temperature difference tolerance range; the extended temperature difference tolerance range is set based on the temperature level of the locked node, the regional environmental gradient fluctuation value, and historical extended link statistics.

[0095] It should be noted that the formula for determining the extended temperature difference tolerance range is:

[0096] |T (j) -T (j-1) |≤δ (i)

[0097] In the formula, T (j) It is the temperature value of the adjacent monitoring node j to be determined, T (j-1) It is the temperature value of the adjacent node that has already been included in the abnormal link; δ (i) The extended temperature difference tolerance threshold is generated with the locked node i as the reference node, and is determined based on the temperature level of the locked node, the temperature difference fluctuation range of the area where the cable is located, and historical link extension data.

[0098] The second condition is that the heating rate of the adjacent monitoring node to be detected has a positive correlation with the locked node or adjacent node, and the rate difference does not exceed the set rate deviation threshold.

[0099] It should be noted that the rate difference refers to the difference in the heating rate between the locked node and its adjacent nodes. The calculation method is to compare the temperature change rate of the current node to be detected in a continuous number of sampling periods with the temperature change rate of the reference node in the same number of periods, and take the average difference as the rate difference index. The heating rate is uniformly expressed as the temperature increase per unit time. The sampling period is 5 minutes, and 6 consecutive periods are taken to form a time series.

[0100] It should be noted that the rate deviation threshold is used to determine whether the heating trends are consistent and the rates are similar. It is set as follows: based on the historical heating data of the target node, the standard deviation of the heating rate during the target node's stable operation period is calculated, and then linearly superimposed with the standard fluctuation range of the current ambient temperature to form a dynamic correction threshold. Specifically, the rate deviation threshold is generated by the following formula:

[0101] ε th =σ hist,j +σ env,j

[0102] In the formula, ε th σ is the rate deviation threshold of target node j; hist,j σ is the standard deviation of the heating rate of target node j over the historical monitoring period, used to reflect the inherent characteristics of the heating fluctuation of the target node; env,j It is the standard deviation of the ambient temperature in the region where the target node j is located within the current time window, used to reflect the influence of the external environment on the warming trend.

[0103] If the monitoring node to be detected does not meet any of the above abnormal extension judgment conditions, it is judged as a monitoring node that does not meet the abnormal extension judgment conditions; if the temperature value of the monitoring node is lower than the corresponding temperature abnormal judgment threshold, or the temperature rise trend is downward or discontinuous, it is considered that the extension requirements are not met, and the extension in that direction is terminated.

[0104] It should be noted that the abnormal link is formed as follows: among the nodes that meet the abnormal extension judgment conditions, a monitoring node identification index is established according to the relative position order. All monitoring nodes identified as abnormal nodes are logically associated to construct an indexed node linked list structure. With the locked node as the central node, a continuous abnormal link is formed by sequentially linking to both sides. The abnormal link is the core basic data structure for fault segment location. It does not involve physical connection and only realizes the construction of logical link through data structure association.

[0105] Step S50: Determine the fault area based on the distribution of abnormal links along the cable. If the abnormal link does not extend to the boundary node, it corresponds to a partial fault section; if the abnormal link extends to any boundary node, it corresponds to a full fault section.

[0106] It should be noted that a boundary node refers to the starting or ending node of a monitoring node.

[0107] A method for identifying and locating thermal anomalies in cable nodes based on FBG sensing, which also includes a dynamic fault section switching mechanism:

[0108] In the process of defining the segment corresponding to the node range covered by the abnormal link as a partial fault segment, or defining the entire monitoring segment as a full fault segment, a dynamic conversion mechanism for fault segments is set up; after the initial judgment of the fault area, the temperature change trend of the boundary nodes at both ends of the abnormal link is continuously monitored. When the boundary nodes are in an abnormal temperature state for multiple consecutive monitoring cycles and meet the extended judgment conditions, the originally defined partial fault segment is updated to a full fault segment.

[0109] Conversely, when the abnormal link has extended to the boundary node of the monitoring node, but the temperature value of the boundary node recovers to below the temperature anomaly judgment threshold within the continuous monitoring period and does not meet the extension trend condition, the originally defined full-segment fault section is updated to a partial fault section.

[0110] It should be noted that the dynamic fault segment conversion mechanism is based on a strategy combining continuous monitoring and status judgment. After the initial fault determination, a "boundary node status confirmation window" is introduced to dynamically track the boundary nodes at both ends of the abnormal link. The core logic of this mechanism is: if a boundary node remains in an abnormal temperature state for multiple consecutive sampling periods and meets the chain-like extension judgment criteria, it can be considered that the fault range is still expanding, thus automatically upgrading the original partial fault segment to a full-segment fault segment; the aforementioned sampling period must be ≥3 periods, with a sampling frequency of once every 5 minutes.

[0111] Conversely, if a boundary node has been included in the abnormal link but its temperature status returns to normal during continuous monitoring, and the rising trend is interrupted or shows a downward trend, then the fault is considered to have stabilized or been alleviated and no longer has the conditions for further extension. At this time, the defined full-segment fault section is downgraded to a partial fault section. The entire conversion process is executed through embedded management and uses a state triggering mechanism to update the segment definition.

[0112] In addition, to verify the difference between the technical solution of this application and the one published in CN201910644626.9 in terms of fault location performance of fiber Bragg grating sensors, the following comparative experimental procedure was established:

[0113] B1. Construct a fiber optic network test platform containing several FBG sensor nodes, designing multiple fault scenarios, including single monitoring node failure and multi-node failure; collect spectral or reflection signal data of each monitoring node under normal and fault conditions, and repeat the test multiple times to obtain stable statistical samples. In the experiment, the methods of this application and publication number CN201910644626.9 are used in parallel for fault location to ensure comparability of the two methods under the same data and environment; all test environments are kept consistent to eliminate the influence of external interference.

[0114] B2. Key indicators for comparative evaluation include:

[0115] Fault node identification accuracy measures the proportion of correctly identified fault nodes out of all predicted fault nodes in the fault location results; accuracy reflects the precision of the solution in identifying faults, and high accuracy means fewer false alarms.

[0116] Fault location response delay records the time required from the occurrence of a fault to the output of the location result, in milliseconds, to evaluate the real-time response capability of the solution in handling fault events;

[0117] The response chain integrity ratio reflects the ability to completely identify the fault propagation chain in a multi-node failure scenario; it is defined as the ratio of the length of the faulty link identified by the solution to the actual length of the link that occurred. The higher the value, the more consecutive faulty nodes the solution can cover.

[0118] Trend matching backtracking rate assesses the solution's ability to track the trend of fault evolution and propagation. The trend matching backtracking rate measures the proportion of the solution that can correctly backtrack and match the actual fault propagation sequence when a chain failure occurs, reflecting the degree to which the technical solution supports the fault development trend.

[0119] B3. Build an optical fiber test platform in a laboratory environment, including a multi-level branched optical network and FBG sensing nodes; simulate the complex deployment scenario of a passive optical network, install FBG reflectors at each splitter port, and identify reflection events of each branch using an OTDR. Set up multiple fault injection schemes, including randomly triggering single-node failure or simultaneous failure of multiple adjacent nodes, to examine the stability of the technical solution in complex environments; comparative experiments are conducted on the same equipment and datasets, with repeated data acquisition and averaging of results to ensure fair comparison.

[0120] B4. Preprocess the acquired raw spectral or time-domain signals, including filtering, denoising, and time-series calibration. Before comparative analysis, normalize the results of the two schemes and unify the units for direct comparison; use appropriate ratio calculation methods for link coverage and trend backtracking indicators.

[0121] B5. Execute the steps from B1 to B4 uniformly, obtain comparative data based on the execution results, and compare the technical solution with the average values ​​of this application on each indicator. The comparison results are shown in the table below, where the values ​​are derived from the statistical results of 50 independent experiments. As can be seen from the table, the fault node identification accuracy, response chain integrity ratio, and trend matching backtracking rate of this application are all more than 30% higher than those of the prior art.

[0122] Indicator Name unit CN201910644626.9 This application Fault Node Identification Accuracy ％ 70 95 Fault location response delay ms 75 62 Response chain integrity ratio ％ 60 88 Trend matching backtracking rate ％ 58 87

[0123] The values ​​in the table above represent the performance comparison results. In the experiment, known faulty nodes were manually set, and the identification results of the two schemes were compared and the identification accuracy was calculated. The experimental results show that the identification accuracy of this application can reach 95%, which is higher than the 70% of the prior art, representing an improvement of about 35%. This application adopts an improved fault confirmation method, which reduces false alarms and thus improves the accuracy.

[0124] The fault location response delay metric records the time from fault triggering to the output of the location result. During testing, the fault events are recorded in sequence. This application has optimized the process, and the average response delay is about 62ms, which is lower than the 75ms of the prior art. The response speed is improved accordingly, which means that this application can locate the fault node more quickly and help to take timely repair measures.

[0125] The response chain integrity ratio reflects the ability to completely identify consecutively faulty links in a multi-node failure scenario. This application further utilizes link analysis to comprehensively identify multi-node faulty links, thus covering and identifying more actual faulty chains. The integrity ratio of this application reaches 88%, higher than the 60% of the prior art, indicating that this application has a stronger link integrity identification capability in complex link scenarios.

[0126] The trend-matching backtracking rate metric evaluates the ability to track fault evolution paths. In experiments, by continuously injecting chain faults, the proportion of actual fault propagation sequences matched by the algorithm backtracking is statistically analyzed. This application supports fault propagation trend analysis and combines historical data for backtracking, resulting in a more comprehensive tracking of fault chain propagation. Therefore, the backtracking rate can reach 87%, which is higher than the 58% of traditional schemes. This result shows that this application has significant advantages in fault development trend analysis and can provide more complete trajectory tracking during fault evolution.

[0127] Thus, a method for identifying and locating thermal anomalies in cable nodes based on FBG sensing has been completed.

[0128] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0129] Those skilled in the art will recognize that the algorithms or steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0130] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for identifying and locating thermal anomalies in cable nodes based on FBG sensing, characterized in that, Including the following: Acquire temperature data from each monitoring node; Based on environmental and external data from each monitoring node, a temperature anomaly detection threshold is generated. The temperature data of the monitoring node is compared with the temperature anomaly judgment threshold to determine whether it exceeds the limit. If it exceeds the limit, the locking process is executed and the node is locked. Starting from the locked node, chain-like anomaly extension identification is performed to both sides of the locked node. The temperature data of the monitoring nodes adjacent to the locked node are detected in sequence. If the adjacent monitoring node meets the anomaly extension judgment condition, the monitoring node is included in the anomaly link and the next adjacent monitoring node is detected. When a monitoring node that does not meet the anomaly extension judgment condition is detected for the first time in a series of adjacent monitoring nodes, or when the detection range has been extended to the starting point or ending point of the current monitoring node, the extension in that direction is stopped, thus forming an anomaly link consisting of the locked node and the continuous monitoring nodes on both sides of the locked node. The fault area is determined by the distribution of abnormal links along the cable. If the abnormal link does not extend to the boundary node, it corresponds to a partial fault section; if the abnormal link extends to any boundary node, it corresponds to the entire fault section.

2. The method for identifying and locating thermal anomalies in cable nodes based on FBG sensing as described in claim 1, characterized in that, Based on environmental and external data from each monitoring node, a temperature anomaly detection threshold is generated, specifically including: The temperature baseline values ​​for each monitoring node are calculated based on environmental data, including the multi-year average air temperature, historical operating load, and geographical environmental parameters of the monitoring nodes. The calculation formula is as follows: T base,i =η1·T clim,i +η2·T load,i +η3·T alt,i In the formula, T base,i It is the temperature reference value of monitoring node i, T clim,i It is the multi-year average temperature (T) of the area where monitoring node i is located. load,i It is the average load temperature rise value of monitoring node i during its historical operation, T alt,i η1, η2, and η3 are the correction values ​​of the geographical environment temperature in the area where monitoring node i is located. The correction values ​​of the geographical environment temperature are obtained by linearly combining the altitude and latitude values ​​in the geographical environment parameters. η1, η2, and η3 are the weight coefficients of each influencing factor, and their values ​​range from η1 + η2 + η3 = 1. The temperature baseline values ​​of each monitoring node are adaptively corrected using external data to obtain the adaptive correction results; the external data includes the real-time environmental parameters of each monitoring node, historical temperature curves, and temperature rise trends of neighboring nodes. Temperature anomaly detection thresholds for each monitoring node are generated based on the temperature baseline value and adaptive correction results.

3. The method for identifying and locating thermal anomalies in cable nodes based on FBG sensing as described in claim 2, characterized in that, The temperature baseline values ​​of each monitoring node are adaptively corrected using external data to obtain the adaptive correction result, calculated using the following formula: In the formula, ΔT i It is the adaptive correction result of monitoring node i, T base,i w1·(T) represents the baseline temperature value of monitoring node i. env,i -T env,avg,i ) is the first correction term, (T) env,i -T env,avg,i () represents the real-time ambient temperature T env,i Compared with historical average ambient temperature T env,avg,i The difference is used to reflect the immediate impact of the current environment on the temperature baseline value, and w1 is the weighting coefficient of the first correction term. The second correction term represents the historical temperature change rate of monitoring node i, used to reflect the correction of the monitoring node's own temperature trend to the baseline. w2 is the weighting coefficient of the second correction term. The third correction term represents the average temperature change rate of neighboring nodes, where n is the number of neighboring nodes, used to reflect the potential impact of the surrounding environment on the baseline, and w3 is the weighting coefficient of the third correction term; the values ​​of w1, w2, and w3 are in the range of: w1 + w2 + w3 = 1.

4. The method for identifying and locating thermal anomalies in cable nodes based on FBG sensing as described in claim 2, characterized in that, The formula for generating the temperature anomaly detection threshold is: T thresh,i =T base,i +ΔT i In the formula, T thresh,i It is the temperature anomaly detection threshold of monitoring node i, which is used to compare with the temperature data of the monitoring node to determine whether there is a temperature anomaly. T base,i It is the temperature reference value of monitoring node i, ΔT i It is the adaptive correction result of monitoring node i.

5. The method for identifying and locating thermal anomalies in cable nodes based on FBG sensing as described in claim 1, characterized in that, The locking process specifically includes: The locking process is implemented based on a multi-condition fusion mechanism, which includes real-time temperature difference judgment, heating rate judgment and short-term fluctuation elimination mechanism. The execution order of the locking process is as follows: first, the real-time temperature difference judgment is executed, then the heating rate judgment is executed, and the short-term fluctuation elimination mechanism is continuously executed throughout the judgment process to eliminate false abnormal signals in real time. The real-time temperature difference judgment refers to calculating the difference between the temperature value of the current monitoring node and the average temperature of multiple neighboring monitoring nodes in normal state, and comparing the obtained temperature difference value with a set dynamic difference threshold. When the temperature difference value exceeds the dynamic difference threshold, the monitoring node is determined to meet the abnormal temperature difference amplitude condition; the monitoring nodes that meet the abnormal temperature difference amplitude condition are recorded in the candidate locked node set. The heating rate determination involves performing time series analysis on the temperature data of the current monitoring node over multiple consecutive sampling periods, calculating the temperature change rate per unit time, and comparing the temperature change rate with a heating rate threshold. When the change rate exceeds the heating rate threshold, the monitoring node is determined to meet the abnormal heating rate condition. Monitoring nodes that meet the abnormal heating rate condition are selected from the candidate locked node set and confirmed as locked nodes. The short-term fluctuation rejection mechanism performs a temperature fluctuation persistence judgment when the temperature of a certain monitoring node exceeds the temperature anomaly judgment threshold. Abnormal temperatures that only occasionally exceed the threshold within the continuous sampling period are judged as pseudo-abnormal signals, and the monitoring node is not set as a locked node in this case. Monitoring nodes that are not defined as pseudo-abnormal signals but exceed the temperature anomaly judgment threshold are handed over to the heating rate judgment, and the locked node is obtained based on the heating rate judgment.

6. The method for identifying and locating thermal anomalies in cable nodes based on FBG sensing as described in claim 1, characterized in that, The abnormal extension determination criteria include the following two conditions, which must be met simultaneously: The first condition is that the temperature value of the adjacent monitoring node to be detected exceeds the corresponding temperature anomaly judgment threshold, and the temperature difference with the previous monitoring node is within the set extended temperature difference tolerance range. The second condition is that the heating rate of the adjacent monitoring node to be detected has a positive correlation with the locked node or adjacent node, and the rate difference does not exceed the set rate deviation threshold. If the monitoring node to be detected does not meet any of the above abnormal extension judgment conditions, it is judged as a monitoring node that does not meet the abnormal extension judgment conditions; if the temperature value of the monitoring node is lower than the corresponding temperature abnormal judgment threshold, or the temperature rise trend is downward or discontinuous, it is considered that the extension requirements are not met, and the extension in that direction is terminated.

7. The method for identifying and locating thermal anomalies in cable nodes based on FBG sensing as described in claim 1, characterized in that, It also includes a dynamic fault section switching mechanism: In the process of defining the segment corresponding to the node range covered by the abnormal link as a partial fault segment, or defining the entire monitoring segment as a full fault segment, a dynamic conversion mechanism for fault segments is set up; after the initial judgment of the fault area, the temperature change trend of the boundary nodes at both ends of the abnormal link is continuously monitored. When the boundary nodes are in an abnormal temperature state for multiple consecutive monitoring cycles and meet the extension judgment conditions, the originally defined partial fault segment is updated to a full fault segment. Conversely, when the abnormal link has extended to the boundary node of the monitoring node, but the temperature value of the boundary node recovers to below the temperature anomaly judgment threshold within the continuous monitoring period and does not meet the extension trend condition, the originally defined full-segment fault section is updated to a partial fault section.

Images