A method and system for fiber connector status awareness and resource management

Abstract

The present application relates to the technical field of optical fiber, and specifically discloses a kind of optical fiber connector state sensing and resource management method and system.The present application is by real-time multidimensional collection optical power, return loss, end face geometry, four kinds of data such as environmental temperature, change discrete inspection into continuous monitoring, capture deterioration trajectory completely, realize from after repair to in-process monitoring change, based on optical power feature calculation optical decay stress coefficient, can accurately identify contact bad, micro-displacement and other implicit deterioration, realize early warning, by reflecting deterioration, stress disturbance, deterioration trend and optical decay stress weighted fusion get joint attenuation coefficient, solve the problem that existing separation evaluation, unable to predict deterioration, and lead to obvious lag in resource management of optical fiber connector in the process of operation and maintenance.

Description

Technical Field

[0001] This invention relates to the field of optical fiber technology, and in particular to a method and system for optical fiber connector status sensing and resource management. Background Technology

[0002] With the large-scale deployment and high-speed evolution of optical fiber communication networks, optical fiber connectors, as core passive devices for detachable connections in optical links, directly affect the transmission performance and stability of the entire communication system due to their physical connection quality. In traditional network operation and maintenance systems, monitoring the status of optical fiber connectors mainly relies on periodic inspections using instruments such as optical time domain reflectometers (OTDRs). However, existing periodic inspections cannot capture the gradual degradation process of connector performance (such as end-face contamination accumulation and microcrack propagation). Faults are often only located after the link has experienced severe attenuation or interruption, resulting in significant lag in resource management of fiber optic connectors during operation and maintenance. Therefore, a fiber optic connector status awareness and resource management method is needed to solve the above problems. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for optical fiber connector status awareness and resource management to solve the technical problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A method for fiber optic connector status awareness and resource management includes: Obtain a real-time status dataset of the target fiber optic connector, wherein the real-time status dataset includes optical power characteristic parameters, return loss characteristic parameters, end face geometric characteristic parameters, and ambient temperature characteristic parameters; The insertion loss fluctuation sequence and instantaneous optical power jitter rate are obtained based on the optical power characteristic parameters, and the optical attenuation stress coefficient is obtained based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate. The reflection degradation coefficient is obtained based on the return loss characteristic parameters, and the reflection-optical attenuation joint attenuation coefficient is obtained based on the reflection degradation coefficient and the optical attenuation stress coefficient. The scratch density and stain coverage are obtained based on the end face geometric feature parameters, and the end face contamination coefficient is obtained based on the scratch density and stain coverage. The ambient temperature influence factor is obtained based on the ambient temperature characteristic parameters, and the reflection-light decay joint attenuation correction coefficient is obtained based on the ambient temperature influence factor and the reflection-light decay joint attenuation coefficient. The comprehensive status value is obtained based on the reflection-optical attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and resource management is performed on the target optical fiber connector based on the comprehensive status value.

[0005] Preferably, the step of obtaining the optical attenuation stress coefficient based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate includes: The insertion loss fluctuation amplitude and frequency are obtained from the insertion loss fluctuation sequence, and the insertion loss dynamic offset is obtained from the insertion loss fluctuation amplitude and the insertion loss fluctuation frequency. The peak jitter amplitude and jitter duration are obtained based on the instantaneous optical power jitter rate, and the transient impact of optical power is obtained based on the peak jitter amplitude and the jitter duration. The initial optical decay stress factor is obtained based on the insertion loss dynamic offset and the optical power transient impact. Obtain the reference insertion loss tolerance threshold of the fiber optic connector, and obtain the optical attenuation stress coefficient based on the initial optical attenuation stress factor and the reference insertion loss tolerance threshold.

[0006] Preferably, the steps of obtaining the reflection degradation coefficient based on the return loss characteristic parameters and obtaining the reflection-optical attenuation joint attenuation coefficient based on the reflection degradation coefficient and the optical attenuation stress coefficient include: The reflection peak energy ratio and the return loss baseline drift are obtained based on the return loss characteristic parameters, and the reflection degradation coefficient is obtained based on the reflection peak energy ratio and the return loss baseline drift. The low-frequency disturbance component and the high-frequency noise component in the optical power characteristic parameters are obtained, and the stress disturbance coefficient is obtained based on the low-frequency disturbance component and the high-frequency noise component. Obtain the historical reflection degradation coefficient sequence within a preset time window, and obtain the reflection degradation trend factor based on the historical reflection degradation coefficient sequence; The reflection degradation coefficient, the stress disturbance coefficient, the reflection degradation trend factor, and the optical decay stress coefficient are weighted and fused to obtain the reflection-optical decay joint attenuation coefficient.

[0007] Preferably, the steps of obtaining the scratch density and stain coverage rate based on the end face geometric feature parameters, and obtaining the end face contamination coefficient based on the scratch density and stain coverage rate, include: Obtain the scratch image and the stain area image from the geometric feature parameters of the end face; The total length of the scratch and the area of ​​the scratch-affected region are obtained from the scratch image, and the scratch density is obtained from the total length of the scratch and the area of ​​the scratch-affected region. The total area and morphological complexity of the stain are obtained from the stain area image, and the stain coverage rate is obtained from the total area and morphological complexity of the stain. Obtain the preset effective area of ​​the reference end face, and obtain the end face contamination coefficient based on the scratch density, the stain coverage rate, and the effective area of ​​the reference end face.

[0008] Preferably, the step of obtaining the environmental temperature influence factor based on the environmental temperature characteristic parameters, and obtaining the reflection-light decay joint attenuation correction coefficient based on the environmental temperature influence factor and the reflection-light decay joint attenuation coefficient, includes: Obtain the temperature time series sequence from the ambient temperature characteristic parameters, and obtain the temperature fluctuation amplitude and temperature fluctuation frequency based on the temperature time series sequence, and obtain the temperature fluctuation intensity based on the temperature fluctuation amplitude and the temperature fluctuation frequency; The temperature change rate sequence is obtained based on the temperature time series, and the maximum positive change rate and the maximum negative change rate within a preset time window are obtained based on the temperature change rate sequence. The temperature drastic change factor is obtained based on the maximum positive rate of change and the maximum negative rate of change; Preliminary temperature influence factors are obtained based on the intensity of temperature fluctuations and the temperature drastic change factor. Obtain the historical temperature influence factor sequence within a preset historical time period, and obtain the temperature influence trend factor based on the historical temperature influence factor sequence; The environmental temperature influence factor is obtained based on the preliminary temperature influence factor and the temperature influence trend factor. The reflection-light decay joint attenuation coefficient is linearly corrected based on the environmental temperature influence factor to obtain the reflection-light decay joint attenuation correction coefficient.

[0009] Preferably, the step of obtaining a comprehensive state value based on the reflection-optical attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and performing resource management on the target fiber optic connector based on the comprehensive state value, includes: The combined state value is obtained by weighting the reflection-light decay joint attenuation correction coefficient and the end face contamination coefficient. Obtain a preset first state threshold and a second state threshold, wherein the first state threshold is greater than the second state threshold; Based on the threshold range in which the comprehensive status value falls, determine the resource management action for the target fiber optic connector: If the overall status value is greater than the first status threshold, a connector replacement instruction is generated; If the overall state value is between the second state threshold and the first state threshold, then an end face cleaning instruction is generated; If the overall status value is less than the second status threshold, a maintenance operation instruction is generated, and the health score of the target fiber optic connector is updated based on the overall status value.

[0010] This application also provides a fiber optic connector status awareness and resource management system, including: The data acquisition module is used to acquire the real-time status dataset of the target optical fiber connector, wherein the real-time status dataset includes optical power characteristic parameters, return loss characteristic parameters, end face geometric characteristic parameters, and ambient temperature characteristic parameters. The optical attenuation stress coefficient calculation module is used to obtain the insertion loss fluctuation sequence and instantaneous optical power jitter rate based on the optical power characteristic parameters, and to obtain the optical attenuation stress coefficient based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate. The attenuation coefficient calculation module is used to obtain the reflection degradation coefficient based on the return loss characteristic parameters, and to obtain the reflection-optical attenuation joint attenuation coefficient based on the reflection degradation coefficient and the optical attenuation stress coefficient. The end-face contamination coefficient calculation module is used to obtain the scratch density and stain coverage rate based on the end-face geometric feature parameters, and to obtain the end-face contamination coefficient based on the scratch density and stain coverage rate. The temperature influence correction module is used to obtain the environmental temperature influence factor based on the environmental temperature characteristic parameters, and to obtain the reflection-light decay joint attenuation correction coefficient based on the environmental temperature influence factor and the reflection-light decay joint attenuation coefficient. The status assessment module is used to obtain a comprehensive status value based on the reflection-light attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and to perform resource management on the target fiber optic connector based on the comprehensive status value.

[0011] Preferably, the optical attenuation stress coefficient calculation module includes: An offset calculation unit is used to obtain the loss fluctuation amplitude and loss fluctuation frequency based on the insertion loss fluctuation sequence, and to obtain the dynamic offset of the insertion loss based on the loss fluctuation amplitude and the loss fluctuation frequency. The impact calculation unit is used to obtain the peak jitter amplitude and jitter duration based on the instantaneous optical power jitter rate, and to obtain the transient impact of optical power based on the peak jitter amplitude and the jitter duration. The optical attenuation stress factor calculation unit is used to obtain the initial optical attenuation stress factor based on the insertion loss dynamic offset and the optical power transient impact. The optical attenuation stress calculation unit is used to obtain the reference insertion loss tolerance threshold of the optical fiber connector, and to obtain the optical attenuation stress coefficient based on the initial optical attenuation stress factor and the reference insertion loss tolerance threshold.

[0012] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0013] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0014] The beneficial effects of this application are as follows: This invention transforms discrete inspection into continuous monitoring by collecting four types of data in real time and from multiple dimensions: optical power, return loss, end-face geometry, and ambient temperature. It completely captures the degradation trajectory, realizing the transformation from post-repair to in-process monitoring. Based on the optical power characteristics, the optical attenuation stress coefficient is calculated, which can accurately identify hidden degradation such as poor contact and micro-displacement, and realize early warning. By weighted fusion of reflection degradation, stress disturbance, degradation trend and optical attenuation stress, a joint attenuation coefficient is obtained, which solves the problem of traditional separate assessment and inability to predict degradation. Machine vision is used to quantify the end-face contamination coefficient, replacing manual inspection, objectively identifying scratches and stains, improving detection rate and consistency. Temperature influence correction is introduced to remove environmental interference, avoid misjudgment, and make the assessment more realistic and reliable. Finally, through comprehensive status value and dual threshold hierarchical decision-making, replacement, cleaning, and maintenance instructions are automatically generated, realizing precise and automated resource management and reducing operation and maintenance costs and failure risks. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of a method flow according to an embodiment of this application.

[0016] Figure 2 This is a schematic diagram of the system structure according to an embodiment of this application.

[0017] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] like Figure 1 As shown, this application provides a method for fiber optic connector status awareness and resource management, including: S1. Obtain the real-time status dataset of the target fiber optic connector, wherein the real-time status dataset includes optical power characteristic parameters, return loss characteristic parameters, end face geometric characteristic parameters, and ambient temperature characteristic parameters. S2. Obtain the insertion loss fluctuation sequence and instantaneous optical power jitter rate based on the optical power characteristic parameters, and obtain the optical attenuation stress coefficient based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate. S3. Obtain the reflection degradation coefficient based on the return loss characteristic parameters, and obtain the reflection-optical attenuation joint attenuation coefficient based on the reflection degradation coefficient and the optical attenuation stress coefficient; S4. Obtain the scratch density and stain coverage rate based on the end face geometric feature parameters, and obtain the end face contamination coefficient based on the scratch density and stain coverage rate. S5. Obtain the ambient temperature influence factor based on the ambient temperature characteristic parameters, and obtain the reflection-light decay joint attenuation correction coefficient based on the ambient temperature influence factor and the reflection-light decay joint attenuation coefficient. S6. Obtain a comprehensive status value based on the reflection-optical attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and perform resource management on the target optical fiber connector based on the comprehensive status value.

[0020] As described in steps S1-S6 above, existing methods rely solely on periodic OTDR inspection, which can only obtain insertion loss / return loss values ​​at a specific moment. This fails to capture the slow degradation caused by end-face contamination, microcracks, and temperature fluctuations, often leading to failures only being detected after they occur. Therefore, this invention first acquires a real-time status dataset of the target fiber optic connector. This dataset includes optical power characteristic parameters, return loss characteristic parameters, end-face geometric characteristic parameters, and ambient temperature characteristic parameters. Four types of data are synchronously collected at fixed time intervals (e.g., 1 second / time): optical power characteristics (real-time transmitted / received optical power, insertion loss value); return loss characteristics (return loss value, reflection peak data); end-face geometric characteristics (microscopic image, radius of curvature, indentation); and ambient temperature characteristics (real-time temperature around the connector). This real-time, multi-dimensional acquisition transforms discrete inspection into continuous monitoring: continuously recording optical power fluctuations, return loss drift, subtle end-face changes, and temperature fluctuations, forming a complete time-series trajectory. Even if the degradation rate is extremely slow, it can still be captured by data sequences, avoiding asymptomatic mutations, realizing the transformation from post-event repair to in-event monitoring, and at the same time saying goodbye to delayed inspections, providing original evidence for condition assessment; Secondly, existing methods rely solely on static insertion loss assessment, failing to identify dynamic stress and exhibiting a high false positive rate. Static insertion loss compliance does not guarantee connection stability; dynamic fluctuations and transient shocks are the primary causes of early failures, leading to missed detections. Therefore, this paper addresses this issue by extracting the insertion loss fluctuation sequence and instantaneous optical power jitter rate from optical power characteristic parameters, and then calculating the optical attenuation stress coefficient. Specifically, this involves calculating the loss fluctuation amplitude and frequency from the insertion loss fluctuation sequence to obtain the dynamic offset of the insertion loss. This is achieved by extracting continuous insertion loss data from the optical power characteristics and calculating the loss fluctuation amplitude (maximum value - minimum value) and fluctuation frequency (number of fluctuations per unit time). The peak jitter amplitude and duration are calculated based on the instantaneous optical power jitter rate to obtain the transient impact of optical power. Then, the coefficient is calculated by combining the initial optical attenuation stress factor and the reference insertion loss tolerance threshold. Specifically, the reference insertion loss tolerance threshold of the connector model is obtained, and the initial factor is compared with the threshold to obtain the optical attenuation stress coefficient in the range of 0-1 (the larger the value, the stronger the stress). Its physical meaning is to quantify the impact of dynamic changes in optical power on the transmission stress of the connector, accurately reflect the loss fluctuation and impact resistance of the connector in the process of optical signal transmission. By using this coefficient, latent degradation can be identified in advance, and alarms can be triggered before the static insertion loss exceeds the standard, thus achieving early warning.

[0021] Traditional methods of separately assessing optical decay and reflection, ignoring trends, and failing to predict degradation have several drawbacks: optical decay and reflection are interdependent, and assessing them separately may underestimate the risk; without trend analysis, only the current state can be observed, and future deterioration cannot be predicted. This paper addresses these shortcomings by calculating the reflection degradation coefficient based on return loss characteristic parameters (extracting the proportion of reflection peak energy and return loss baseline drift from return loss characteristics to quantify the degree of reflection degradation). Then, the reflection degradation coefficient, stress disturbance coefficient (decomposing low-frequency disturbances and high-frequency noise in optical power to reflect the interference of external stress on transmission), and reflection degradation trend factor (obtained by taking the historical reflection degradation coefficient sequence and performing slope fitting) are combined to obtain the... The reflection-optical attenuation joint attenuation coefficient is obtained by weighted fusion of the rising / falling / stable trend and the optical attenuation stress coefficient. The reflection degradation coefficient is calculated from the proportion of reflection peak energy and the return loss baseline drift. The stress disturbance coefficient is extracted and calculated from the low-frequency disturbance and high-frequency noise components of optical power. The degradation trend factor is obtained through the analysis of historical reflection degradation coefficient sequence. In this way, by combining joint modeling and trend analysis, the current state is combined with historical trends: the reflection degradation worsens, the optical attenuation stress increases, the disturbance increases, and the trend deteriorates. The joint coefficient rises rapidly, and the system can identify progressive faults in advance, avoiding accumulation until the breakpoint before processing.

[0022] Due to the inherent limitations of manual inspection of connector end faces—namely, its subjectivity and tendency to miss minute stains and scratches—manual inspection relies on experience, making it difficult to detect and quantify minor contamination and fine scratches. End face issues often lead to reduced echo and increased light attenuation. Therefore, this paper proposes a new method to calculate scratch density and stain coverage based on end face geometric parameters, thereby obtaining an end face contamination coefficient. Specifically, scratch density is determined by statistically analyzing the total scratch length and affected area of ​​end face scratch images, while stain coverage is determined by statistically analyzing the total stain area and morphological complexity of stain area images. This is then combined with the effective area of ​​the baseline end face to calculate the cleanliness coefficient. This data comes from image information acquired by an end face microscopic imaging device. Its physical meaning is to quantify the degree of physical damage and contamination on the connector end face. End face cleanliness directly determines the scattering and attenuation levels of optical signal transmission. This step accurately reflects the impact of end face condition on transmission performance. Machine vision is used to quantify end face quality: more scratches and larger stain areas result in a higher cleanliness coefficient. The system automatically determines end face deterioration, triggering a cleaning command, transforming fuzzy manual judgment into precise machine quantification, improving consistency and detection rate.

[0023] Existing methods suffer from drawbacks such as ignoring temperature effects, distorting assessment results, and misjudging extreme environments. Temperature changes cause thermal expansion and contraction of materials and fluctuations in insertion / return loss. Without temperature correction, environmental interference may be misjudged as connector malfunctions. Therefore, this method calculates an environmental temperature influence factor based on environmental temperature characteristic parameters and then uses this factor to linearly correct the reflection-optical attenuation joint attenuation coefficient, resulting in a reflection-optical attenuation joint attenuation correction coefficient. Specifically, the initial temperature influence factor is obtained by calculating the intensity of temperature fluctuations and the temperature drastic change factor through temperature time series. The final environmental temperature influence factor is determined by combining the historical temperature influence factor series. This approach considers the impact of environmental temperature fluctuations and drastic changes on the material properties and contact state of fiber optic connectors, eliminates attenuation calculation biases caused by environmental factors, and makes the attenuation assessment results more consistent with the actual working state of the connector, thereby improving the accuracy of the condition assessment.

[0024] The comprehensive status value is obtained by multiplying the reflection-optical attenuation joint attenuation correction coefficient by the end-face contamination coefficient (formula: comprehensive status value = reflection-optical attenuation joint attenuation correction coefficient × end-face contamination coefficient). Then, resource management actions are performed according to the preset first and second status thresholds. When the comprehensive status value is greater than the first status threshold, a connector replacement command is generated; when it is between the two thresholds, an end-face cleaning command is generated; and when it is less than the second status threshold, a maintenance operation command is generated and the health score is updated. Its physical meaning is to integrate the transmission attenuation correction result with the physical status of the end face to form a quantitative indicator that can directly guide operation and maintenance, realize the automation and precision of resource management, and make the operation and maintenance actions highly matched with the actual status of the connector. At the same time, through precise classification and automatic closed loop, passive emergency repair is transformed into proactive operation and maintenance: automatic cleaning for minor degradation, automatic replacement for severe degradation, continuous monitoring of health status, precise resource allocation, and reduction of costs and failure risks.

[0025] In one embodiment, step S2, which involves obtaining the optical attenuation stress coefficient based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate, includes: S21. Obtain the loss fluctuation amplitude and loss fluctuation frequency according to the insertion loss fluctuation sequence, and obtain the insertion loss dynamic offset according to the loss fluctuation amplitude and the loss fluctuation frequency. S22. Obtain the peak jitter amplitude and jitter duration based on the instantaneous optical power jitter rate, and obtain the transient impact of optical power based on the peak jitter amplitude and the jitter duration; S23. Obtain the initial optical attenuation stress factor based on the insertion loss dynamic offset and the optical power transient impact. S24. Obtain the reference insertion loss tolerance threshold of the optical fiber connector, and obtain the optical attenuation stress coefficient based on the initial optical attenuation stress factor and the reference insertion loss tolerance threshold.

[0026] As described in steps S21-S24 above, fiber optic connector monitoring suffers from several core drawbacks. It relies solely on static insertion loss, cannot identify dynamic contact degradation, and has a high rate of missed detection for latent faults. Traditional maintenance relies on OTDRs or optical power meters to read instantaneous insertion loss. As long as the value is within the threshold range, it can be considered normal. However, it cannot identify latent problems such as periodic fluctuations in insertion loss or small, continuous jitter. These dynamic anomalies are often early signs of connector loosening, pigtail bending stress, and micro-wear on the end face. If not addressed in time, they will gradually develop into excessive insertion loss, intermittent link disconnection, or even service interruption. Furthermore, traditional solutions cannot distinguish between statically qualified but dynamically deteriorated pseudo-healthy states, resulting in a large number of faulty connectors going undetected in advance, leading to a reactive, reactive repair situation in maintenance. Therefore, this invention is based on time-series insertion loss data and relies on digital signal processing and statistical analysis methods. The specific process is as follows: First, from the optical power characteristics of the real-time status dataset, continuous insertion loss values ​​within a preset time window (e.g., 5 seconds, 10 seconds) are extracted to form an insertion loss fluctuation time series, ensuring that the sequence length is sufficient to cover typical fluctuation periods. Second, the sequence is preprocessed by using a moving average filter to remove high-frequency random noise, retaining the true fluctuation characteristics, and avoiding distortion of amplitude and frequency calculations due to environmental interference. Next, the loss fluctuation amplitude is calculated, which is the difference between the maximum and minimum insertion loss values ​​in the filtered sequence. A larger amplitude indicates a more unstable connector contact state. Then, the loss fluctuation frequency is calculated by using zero-crossing detection or peak counting to count the number of times the loss value crosses the mean line per unit time. A higher frequency indicates more frequent stress disturbances. Finally, the fluctuation amplitude and fluctuation frequency are weighted and fused to obtain the dynamic offset of the insertion loss. The weighting follows the principle of amplitude priority and frequency auxiliary. The amplitude weight is usually set to 0.6-0.7, and the frequency weight is set to 0.3-0.4. After fusion, normalization is performed so that the offset falls within the 0-1 range. The larger the value, the more severe the dynamic degradation. The entire process is completed in real time in the edge computing unit, with a computation latency of less than 100 milliseconds, meeting the requirements of online monitoring. This solves the core defects of traditional fiber optic connector monitoring, which relies solely on static insertion loss, cannot identify dynamic contact degradation, and has a high rate of missed detection of latent faults. Traditional maintenance relies on OTDR or optical power meters to read instantaneous insertion loss. As long as the value is within the threshold range, it can be judged as normal, but it cannot identify latent problems such as periodic fluctuations in insertion loss and small, continuous jitter. These dynamic anomalies are often early manifestations of connector loosening, pigtail bending stress, and micro-wear on the end face. If not dealt with in time, they will gradually develop into excessive insertion loss, link interruption, or even service interruption. In addition, traditional solutions cannot distinguish between statically qualified but dynamically degraded pseudo-healthy states, resulting in a large number of faulty connectors not being detected in advance, and maintenance is in a passive situation of reactive repair after the fact. For example, in an intercity communication trunk line equipment room, the FC / PC fiber optic connector experienced unstable ferrule contact due to slight vibrations in the cabinet. The static insertion loss remained at 0.28 dB, far below the acceptable threshold of 0.5 dB, and traditional inspections determined it to be in good condition after three consecutive tests. However, during monitoring, the system extracted a 10-second insertion loss sequence, filtered it, and calculated a loss fluctuation of 0.14 dB, with a fluctuation frequency of 4 times / second. After weighted fusion, the dynamic offset of the insertion loss was 0.62, significantly higher than the safety threshold of 0.3. Based on this, the system determined that the connector had dynamic contact degradation. On-site inspection by maintenance personnel revealed that the connector was not fully tightened and the ferrule was slightly offset. After re-tightening, the fluctuation amplitude decreased to 0.02 dB, and the dynamic offset recovered to 0.08, preventing a link interruption due to increased vibration. This case demonstrates that this solution can effectively identify hidden faults that are statically acceptable but dynamically degraded, improving the operational stability of fiber optic links.

[0027] Secondly, because it only focuses on average optical power and stable insertion loss, it is insensitive to instantaneous strong impacts, which are often the direct cause of connector end face scratches, ferrule misalignment, and return loss degradation. Furthermore, traditional maintenance cannot record impact events, making it difficult to determine whether connector degradation is caused by long-term wear or instantaneous impact, resulting in poorly targeted maintenance solutions. In addition, the latent damage caused by instantaneous impacts does not immediately manifest as performance exceeding limits, and cannot be detected by traditional inspections. Long-term accumulation can suddenly trigger link failures, severely affecting communication reliability. Therefore, this invention uses instantaneous optical power jitter rate to complete feature extraction and impact calculation, with a rigorous and real-time process: First, instantaneous optical power jitter rate data is extracted from optical power characteristic parameters. The jitter rate is calculated from the real-time received optical power, reflecting the rate of change of optical power per unit time. Second, peak detection is performed on the jitter rate curve to locate the extreme point of instantaneous jitter, and the peak jitter amplitude is calculated, i.e., the instantaneous optical power and steady-state... The maximum difference in optical power is calculated, with a larger amplitude indicating a higher impact intensity. Next, the jitter duration is calculated from the moment the peak occurs until the optical power returns to a steady state; a longer duration indicates a more persistent impact on the link. Then, an impact calculation model is constructed, combining the peak jitter amplitude and jitter duration using a product fusion method. The formula simplifies to: Transient impact of optical power = K × Peak jitter amplitude × Jitter duration, where K is a standardization coefficient used to map the result to the 0-1 range. Finally, the impact is amplitude-limited to eliminate extreme outliers and ensure numerical stability and reliability. The entire calculation process is executed in real time, outputting results within 50 milliseconds after the jitter occurs, meeting online status awareness requirements. This precise capture and quantification of transient features enables full-process monitoring, quantification, and traceability of instantaneous impacts. When a transient disturbance occurs in the link, the system quickly extracts the peak amplitude and duration, calculating the transient impact of optical power. A higher value indicates greater instantaneous stress on the connector and a higher risk of damage. The system can determine whether the impact poses a threat to the connector based on an impact threshold; if the threshold is exceeded, the impact event is automatically recorded and the connector status is marked. For multiple small-amplitude impacts, the system can accumulate the impact amount to assess the long-term cumulative effect; for a single large-amplitude impact, it directly triggers an early warning, reminding maintenance personnel to check the end face and connection status. This process transforms invisible instantaneous stress into quantifiable indicators, thereby helping maintenance personnel accurately locate the impact source and promptly eliminate potential hazards.

[0028] Next, the initial optical attenuation stress factor is obtained based on the insertion loss dynamic offset and the optical power transient impact. Specifically, the insertion loss dynamic offset and the optical power transient impact are linearly superimposed to obtain the initial optical attenuation stress factor without threshold calibration. First, the insertion loss dynamic offset and the optical power transient impact are normalized, with a value range of 0-1. Second, weighting coefficients are determined based on the link rate and application scenario. For high-speed links (such as 100G and 400G), the transient impact has a greater impact, and the impact weight is adjusted accordingly. The initial optical attenuation stress factor is set to 0.5-0.6, and the dynamic offset weight is set to 0.4-0.5. For low-speed stable links, the impact of long-term fluctuations is more significant, so the dynamic offset weight is set to 0.6-0.7, and the impact weight is set to 0.3-0.4. Then, a weighted summation calculation is performed, using the formula: Initial optical attenuation stress factor = W1 × Insertion loss dynamic offset + W2 × Optical power transient impact, where W1 + W2 = 1. Finally, the calculation results are limited and smoothed to ensure that the factor value is within the range of 0-1, avoiding outliers from affecting subsequent calculations. The fusion process requires no manual intervention; the system automatically matches weights according to the link type, ensuring that the evaluation results fit the actual scenario. In this way, through multi-dimensional feature fusion, a comprehensive optical attenuation stress evaluation system is constructed, eliminating the one-sidedness of single-feature evaluation. The system integrates long-term dynamic degradation and short-term instantaneous impact into its calculations, ensuring that the initial optical decay stress factor contains both steady-state and transient information, accurately reflecting the comprehensive stress experienced by the connector. When both dynamic offset and impact are high, the factor value increases significantly, indicating that the connector is under high stress and has a very high risk of degradation. When one is high and the other is normal, the factor increases moderately, suggesting a potential single-factor defect. When both are low, the factor remains low, indicating a healthy condition. This process transforms the assessment from a partial judgment to a holistic evaluation, improving the accuracy of optical decay stress assessment and providing a reliable basis for subsequent standardized calculations and condition-based decisions.

[0029] Finally, due to key shortcomings in optical attenuation assessment, such as the lack of a unified standard, incomparability across different models, subjective threshold settings, and inconsistent decision-making criteria, traditional operation and maintenance uses the same judgment threshold for different connector models. This leads to false alarms for high-tolerance connectors and missed detections for low-tolerance connectors. At the same time, the lack of standardized output makes it impossible to compare the status of different links and different devices horizontally, making it difficult to achieve unified management of network resources. Furthermore, manually setting thresholds relies on experience, is highly subjective, and can easily lead to over- or under-maintenance, affecting management efficiency and reliability. To address this, a baseline insertion loss tolerance threshold for the fiber optic connector is obtained, and the final optical attenuation stress coefficient is calculated based on the initial optical attenuation stress factor and this threshold. The specific process is as follows: First, the model, specifications, and manufacturer's technical manual of the target fiber optic connector are retrieved from the equipment information database. The baseline insertion loss tolerance threshold is extracted; this threshold represents the maximum allowable insertion loss value for normal connector operation, typically provided by industry standards or the manufacturer (e.g., the baseline insertion loss threshold for single-mode FC / PC connectors is often 0.3dB-0.5dB). Second, the initial optical attenuation stress factor output by S23 is read to confirm its validity and absence of anomalies. Then, a normalized calculation model is constructed, linking the initial factor to the baseline threshold using the formula: Optical attenuation stress coefficient = Initial optical attenuation stress factor × (Real-time insertion loss / Baseline insertion loss tolerance threshold). The coefficient is normalized to ensure it falls within the 0-1 range. When the calculated result is greater than 1, it is taken as 1, indicating that the optical attenuation stress has exceeded the tolerance limit, and the risk of failure is extremely high. After calculation, the coefficients are uploaded to the status assessment module in real time for subsequent comprehensive status calculations. This standardization and normalization using benchmark thresholds establishes a unified optical attenuation stress assessment benchmark across the entire network. The system uses the manufacturer's benchmark insertion loss tolerance threshold as an objective basis, avoiding subjectivity caused by manual intervention. By combining the initial optical attenuation stress factor with the threshold, the optical attenuation stress coefficient reflects the relative risk of stress intensity / tolerance capability, rather than an absolute value, enabling comparability of risk levels for connectors of different models and in different scenarios. When the coefficient is close to 1, it indicates that the stress is approaching the tolerance limit and immediate action is required; a lower coefficient indicates a safe state. This coefficient supports refined and unified management of fiber optic connector resources across the entire network, improving the scientific nature of operation and maintenance decisions.

[0030] In one embodiment, step S3, which involves obtaining the reflection degradation coefficient based on the return loss characteristic parameter and obtaining the reflection-optical attenuation joint attenuation coefficient based on the reflection degradation coefficient and the optical attenuation stress coefficient, includes: S31. Obtain the proportion of reflection peak energy and the amount of return loss baseline drift based on the return loss characteristic parameters, and obtain the reflection degradation coefficient based on the proportion of reflection peak energy and the amount of return loss baseline drift. S32. Obtain the low-frequency disturbance component and the high-frequency noise component in the optical power characteristic parameters, and obtain the stress disturbance coefficient based on the low-frequency disturbance component and the high-frequency noise component; S33. Obtain the historical reflection degradation coefficient sequence within a preset time window, and obtain the reflection degradation trend factor based on the historical reflection degradation coefficient sequence; S34. The reflection degradation coefficient, the stress disturbance coefficient, the reflection degradation trend factor and the optical decay stress coefficient are weighted and fused to obtain the reflection-optical decay joint attenuation coefficient.

[0031] As described in steps S31-S34 above, since maintenance relies on OTDR to periodically detect return loss, it is considered normal as long as the average value is within the acceptable range. However, it cannot identify local strong reflection spikes caused by small stains, shallow scratches, micro-cracks, etc. on the end face, nor can it capture the gradual degradation of slow baseline drift. These hidden anomalies are important causes of high-speed link bit errors, optical module damage, and service interruptions. If not handled in time, they will gradually develop into excessive return loss and link interruption. Therefore, the energy ratio of reflection spikes and the amount of return loss baseline drift are obtained according to the return loss characteristic parameters, and the reflection degradation coefficient is calculated accordingly. Based on real-time return loss characteristic time series data, signal analysis and statistical calculation are used to complete the process. The specific process is as follows: First, the original waveform and time series value sequence of return loss within the preset time window are obtained from the data acquisition module. The data is preprocessed by denoising, filtering, and baseline correction to eliminate instrument noise and link background noise interference. Next, the proportion of reflection peak energy is calculated. An abnormal peak in the return loss waveform is located using a peak detection algorithm. The number of peaks exceeding the baseline threshold, the energy value of a single peak, and the total reflected energy are counted. The proportion is obtained by dividing the total peak energy by the total reflected energy; a higher value indicates a more severe local defect. Then, the return loss baseline drift is calculated. Using the return loss value at the initial calibration time as the baseline, the difference between the average return loss within the current time window and the baseline is calculated to obtain the absolute drift. Normalization is then applied to map the drift to the 0-1 range; a larger drift indicates a more significant overall degradation of reflection performance. Finally, a reflection degradation coefficient fusion model is constructed, which weights and sums the proportion of reflection peak energy and the return loss baseline drift. Typically, the weight of the peak energy proportion is set to 0.6, and the weight of the baseline drift is set to 0.4. After fusion, amplitude limiting is applied to ensure the reflection degradation coefficient is within the 0-1 range; a larger value indicates a higher degree of reflection degradation. The entire process is completed in real time at the edge computing unit, with a computation latency of less than 150 milliseconds, meeting the requirements for online monitoring. This approach, through dual-feature extraction and quantification of reflection degradation, fundamentally addresses the shortcomings of static return loss monitoring. When stains, scratches, cracks, or poor contact appear on the fiber optic connector end face, they will form obvious spikes in the return loss waveform. Furthermore, long-term operation will cause the baseline to slowly drift downwards. By capturing these subtle anomalies through time-series waveform analysis, invisible local defects and gradual degradation are transformed into calculable spike proportions and drift characteristics. Weighted fusion is then used to obtain the reflection degradation coefficient, making latent anomalies explicit. The system can determine whether the connector's reflection performance is degraded based on the magnitude of the reflection degradation coefficient. Even if the average return loss is still within acceptable limits, it can issue an early warning and provide accurate data support for preventative maintenance.

[0032] Low-frequency disturbance components and high-frequency noise components are extracted from optical power characteristic parameters, and stress disturbance coefficients are calculated accordingly. Feature extraction and coefficient calculation are completed using frequency domain decomposition technology based on optical power time-series signals. The process is rigorous and highly real-time: First, continuous time-series optical power data is extracted from the optical power characteristic parameters, using a sliding window to extract stable data segments to avoid data truncation errors. Second, the optical power signal is decomposed in the frequency domain using wavelet transform or Fast Fourier Transform (FFT), separating the signal into low-frequency disturbance components (typically 0-10Hz) and high-frequency noise components (typically above 10Hz). The effective amplitude and energy values ​​of the two types of components are extracted. Then, the low-frequency disturbance component and the high-frequency noise component are normalized, and the energy value is mapped to the 0-1 interval. The higher the component energy, the greater the corresponding disturbance intensity. Next, a stress disturbance coefficient calculation model is constructed, and the two types of components are combined using a weighted fusion method. Considering that low-frequency mechanical disturbances cause more significant damage to connectors, the weight of the low-frequency component is set to 0.6, and the weight of the high-frequency component is set to 0.4. Finally, the fusion result is smoothed and limited to ensure that the stress disturbance coefficient falls within the 0-1 interval, eliminating extreme outlier interference and ensuring numerical stability and reliability. The entire calculation process is executed in real time and can output results within 200 milliseconds, meeting the requirements of online status awareness. In this way, through frequency domain decomposition and disturbance quantization of optical power signals, the entire process of external stress can be monitored, quantified, and attributed. When the link is subjected to low-frequency stresses such as cabinet vibration and fiber optic cable tension, the energy of the low-frequency disturbance component increases significantly. When subjected to high-frequency interference such as electromagnetic interference and transient impacts, the energy of the high-frequency noise component increases significantly. S32 weights and fuses these two types of disturbance energy into a stress disturbance coefficient. The higher the value, the stronger the external stress the connector is subjected to and the higher the risk of performance fluctuations. The system can determine the type and source of disturbance based on the stress disturbance coefficient and component characteristics: low-frequency dominant disturbances prompt checks on mechanical fixation and connection status, while high-frequency dominant disturbances prompt checks on environmental interference and optical module status. This process transforms invisible external stress into quantifiable indicators and provides a basis for operations and maintenance to accurately distinguish between self-degradation and external interference, avoiding ineffective maintenance and promptly eliminating potential stress hazards.

[0033] The historical reflection degradation coefficient sequence within a preset time window is obtained, and the reflection degradation trend factor is obtained through trend fitting analysis. Based on the historical reflection degradation coefficients in the time series database, this is accomplished through statistical fitting and trend analysis algorithms. The steps are clear: First, the historical reflection degradation coefficient sequence within a preset time window (e.g., 24 hours, 7 days) is retrieved from the time series database to ensure data continuity, absence of missing values, and absence of abnormal jumps; second, the historical sequence is cleaned, missing values ​​are filled using interpolation, and outliers are removed using the 3σ principle to ensure the accuracy of trend analysis; then, a... A trend fitting model is constructed, and the historical coefficient sequence is linearly fitted using the least squares method to obtain the slope of the fitted line, which directly reflects the rate of degradation. Next, the reflection degradation trend factor is calculated, and the fitted slope is normalized: a positive slope with a larger absolute value indicates a trend factor closer to 1, signifying accelerated degradation; a slope of 0 corresponds to a trend factor of 0.5, indicating stable performance; a negative slope with a larger absolute value indicates a trend factor closer to 0, signifying gradual performance improvement. Finally, the trend factor is smoothed to eliminate short-term fluctuations and ensure stable and reliable output. The entire process is executed periodically (e.g., updated every 5 minutes) to provide the latest trend data for real-time evaluation. This approach, through trend mining of historical time-series data, constructs a reflection degradation prediction system, eliminating the lag in static evaluation. The system continuously records historical data on reflection degradation coefficients and uses fitting analysis to obtain the slope of degradation changes, which is then converted into a trend factor. When the trend factor continuously rises and exceeds a threshold, it indicates that the connector's reflection performance has entered an accelerated degradation phase, triggering an early warning even if the current state is acceptable. When the trend factor stabilizes around 0.5, it indicates a stable state that can maintain normal operation. When the trend factor decreases, it indicates improved performance, allowing for a reduction in monitoring frequency. This process transforms static assessment into dynamic trend prediction, providing maintenance personnel with ample time to handle sudden failures and turning unexpected faults into predictable and preventable gradual degradation, significantly improving network reliability.

[0034] The reflection degradation coefficient, stress disturbance coefficient, reflection degradation trend factor, and optical decay stress coefficient are weighted and fused to obtain the final reflection-optical decay joint attenuation coefficient. First, the reflection degradation coefficient, stress disturbance coefficient, reflection degradation trend factor, and optical decay stress coefficient are normalized to the 0-1 range; then, a weighted calculation is performed, using the following formula: ; in, Indicates the joint attenuation coefficient. Indicates the reflection degradation coefficient. The weight value represents the reflection degradation coefficient. denoted by , and b represents the weighting value of the stress disturbance coefficient. denoted as the reflection degradation trend factor, where c represents the weight value of the reflection degradation trend factor. denoted by , where d represents the light decay stress coefficient and d represents the weighting value of the light decay stress coefficient. Finally, the fusion results are subjected to amplitude limiting, smoothing, and normalization to ensure that the reflection-light decay joint attenuation coefficient strictly falls within the 0-1 range. Higher values ​​indicate more severe overall attenuation degradation and a higher risk of failure. After calculation, the coefficient is uploaded in real-time to the temperature influence correction module for subsequent joint attenuation correction calculations. This multi-dimensional weighted fusion constructs a comprehensive attenuation assessment system, eliminating the limitations of single indicators and linear assessments. The system incorporates four core characteristics—current reflection degradation, external stress interference, future degradation trends, and light decay stress intensity—into a unified weighted calculation, simulating a real physical degradation mechanism: when multiple indicators are simultaneously high, the fused joint coefficient increases exponentially, accurately reflecting the risk amplification effect; when only one indicator is high and the others are normal, the coefficient increases moderately, indicating a single potential hazard; when all four indicators are low, the coefficient remains low, indicating a healthy state. This process transforms the judgment from scattered indicators to comprehensive risk quantification, significantly improving the accuracy and reliability of attenuation assessment and providing a unified, accurate, and high-value core basis for subsequent temperature correction, state assessment, and resource management.

[0035] In one embodiment, step S4, which involves obtaining the scratch density and stain coverage based on the end-face geometric feature parameters, and obtaining the end-face contamination coefficient based on the scratch density and stain coverage, includes: S41. Obtain the scratch image and the stain area image from the geometric feature parameters of the end face; S42. Obtain the total length of the scratch and the area of ​​the scratch-affected region based on the scratch image, and obtain the scratch density based on the total length of the scratch and the area of ​​the scratch-affected region; S43. Obtain the total area of ​​the stain and the complexity of the stain morphology based on the stain area image, and obtain the stain coverage rate based on the total area of ​​the stain and the complexity of the stain morphology. S44. Obtain the preset effective area of ​​the reference end face, and obtain the end face contamination coefficient based on the scratch density, the stain coverage rate and the effective area of ​​the reference end face.

[0036] As described in steps S41-S44 above, the present invention first acquires the scratch image and the stain area image from the geometric feature parameters of the end face. Specifically, the automated fiber optic end face microscopic imaging system completes image acquisition and separation. The overall process is highly standardized and automated: First, the system uses a mechanical positioning device to precisely fix the fiber optic connector end face within the imaging focal plane, ensuring that the ferrule center is aligned with the lens center, avoiding image distortion caused by off-axis or defocus; second, it turns on the combined illumination of the coaxial light source and the ring light source to provide a uniform, non-reflective, and shadow-free lighting environment, highlighting microscopic defects such as scratches, cracks, stains, and defects, while suppressing image overexposure caused by Fresnel reflection from the end face; then, it transmits the images through a high-resolution CMOS / CCD image transmission system. The sensor acquires the original geometric image of the end face, with a resolution typically no less than 1280×1024 and a single pixel size ≤0.5μm, ensuring that micron-level defects are clearly visible. Then, the original image is preprocessed, including grayscale conversion, Gaussian filtering, binarization, and morphological opening / closing operations to remove background noise and light spot interference, and enhance the defect edges. Finally, the preprocessed image is split into two independent layers using semantic segmentation algorithms (such as U-Net, traditional threshold segmentation + region growing): a scratch image (retaining only linear / planar physical damage such as scratches, cracks, and defects) and a stain area image (retaining only planar contamination such as oil stains, dust, fingerprints, and glue stains). The two layers are strictly non-overlapping and complete, fully covering the entire defect area. The entire acquisition and segmentation process is completed within a preset time (e.g., 1 second), supporting continuous batch detection and meeting the online monitoring needs of high-density connectors in data centers. This automated end-face imaging system acquires complete end-face geometric images in a single step, accurately separating scratch images and stain area images into two target layers, enabling independent visualization of defect areas. Secondly, by acquiring microscopic images of the end face, previously invisible micro-scratches and stains are transformed into clear digital images. Then, semantic segmentation strictly separates scratches and stains, solving the problem of mutual masking and inability to independently assess the two types of defects. The entire process is non-contact, avoiding secondary damage. The original images are stored in the database in real time, forming readily accessible original evidence, solving the problems of untraceability and lack of verification. This process upgrades traditional subjective and vague judgment to objective and precise imaging, fully revealing hidden defects invisible to humans, providing a unique, reliable, and high-precision data source for subsequent quantitative calculations, ensuring the accuracy and reliability of end-face cleanliness assessment from the source. The total scratch length and the area of ​​the scratch-affected region are obtained from the scratch image, and the scratch density is obtained based on the total scratch length and the area of ​​the scratch-affected region. Specifically, the scratch image is thinned to convert wide scratches into single-pixel skeletons for accurate length extraction, while retaining the original scratch area for area calculation. Next, the total scratch length is calculated by traversing all connected components of the scratch in the image, accumulating the pixel length of each connected component, and then converting it to the actual physical length (μm) using the camera calibration coefficient (μm / pixel) to obtain the total length of all scratches on the entire end face. Finally, the area of ​​the scratch-affected region is calculated, and the scratch skeleton is expanded by a specified width. (Typically, the area is expanded by 3-5 pixels, corresponding to an actual size of 1.5-2.5 μm, simulating the range of light field scattering influence.) The total number of pixels in the expanded area is counted, and then converted into the actual physical area (μm²) using a calibration coefficient. This area represents the actual interference range of the scratch on the light-guiding area. Next, the effective light-guiding area area of ​​the end face is obtained (single-mode fiber is typically about 125 μm in diameter, with an effective area ≈ 12300 μm²). Finally, a scratch density calculation model is constructed: Scratch density = Total scratch length ÷ Effective light-guiding area + Scratch-affected area area ÷ Effective light-guiding area. After weighted fusion, the model is normalized to the 0-1 range, with larger values ​​indicating more severe physical damage. The entire calculation process is automated, requiring no human intervention. The results are repeatable and verifiable. By using image quantification and statistics, abstract scratches are transformed into objective values, enabling measurable, comparable, and predictable damage levels. The system first accurately calculates the length and affected area of ​​all scratches, then normalizes them using the effective light-guiding area to obtain a uniform scratch density. When the density exceeds a safety threshold, it indicates that the physical damage threatens transmission quality, and even if it is not visible to humans, it must be addressed. A lower density indicates minor damage, which can be observed further. This process upgrades the vague visual judgment to a precise machine calculation, transforming hidden physical damage into clear numerical indicators. This allows the system to automatically assess damage risk, providing reliable physical damage data for subsequent cleanliness assessments and maintenance decisions, fundamentally solving the problems of missed detections, misjudgments, and lack of standards. The total area and morphological complexity of the stains are obtained from the stain area image. The stain coverage rate is then calculated based on these factors. Specifically, connected component analysis is performed on the stain image to label all independent stain regions and exclude isolated noise points (typically, areas <5 μm² are considered noise). Next, the total stain area is calculated by counting the number of pixels in all effective connected components of the stains and converting this count to the actual physical area (μm²) using calibration coefficients, thus obtaining the total coverage area of ​​all stains on the end face. Finally, the morphological complexity of the stains is calculated using a combination of fractal dimension, dispersion, and shape factor: the fractal dimension reflects... The irregularity of the stain edge, dispersion reflecting the concentration / dispersion of the stain, and shape factor reflecting the roundness / irregularity of the stain are weighted together to obtain a morphological complexity in the range of 0 to 1. A higher value indicates a more dispersed, irregular, and hazardous stain. Next, the effective light-guiding area of ​​the end face is obtained. Finally, a stain coverage model is constructed: Stain Coverage = (Total Stain Area ÷ Effective Area) × (1 + Morphological Complexity). The area is weighted and amplified by morphological complexity to achieve higher coverage for dispersed, irregular, and highly hazardous contaminants. The result is normalized to the range of 0 to 1, with a higher value indicating more severe contamination and higher hazard. The calculation process is completed in real time and supports online output. This joint quantification of area and morphology accurately reflects the degree of contamination hazard: the system first calculates the total stain area, then uses morphological complexity for weighting, significantly increasing the coverage of dispersed, irregular, and highly hazardous contaminants, while relatively reducing the coverage of concentrated, regular, and low-hazard contaminants, making the indicators highly consistent with actual optical hazards. When the stain coverage exceeds a threshold, regardless of the area size, the system will determine that the pollution risk is high and automatically trigger a cleaning command; low coverage indicates that the pollution is minor or has little harm. This process upgrades the subjective judgment of dirtiness by humans to the objective calculation of harm by machines, realizing the transformation of pollution from visible to calculable, comparable, and controllable, accurately identifying high-harm micro-pollutants, and providing the most critical pollution dimension basis for end-face health assessment; Obtain a preset effective area of ​​the reference end face, and obtain the end face contamination coefficient based on the scratch density, the stain coverage rate, and the effective area of ​​the reference end face. The effective area of ​​the reference end face is the standard area of ​​the optical fiber guide zone (single-mode fiber ≈ 12300 μm², multi-mode fiber configured according to model), serving as a unified evaluation benchmark for the entire network. Next, read the scratch density output from S42 and the stain coverage rate output from S43 to confirm that both indicators have been normalized to the 0~1 range. Then, set weighting: typically, stain coverage rate has a weight of 0.6, and scratch density has a weight of 0.4 (contamination has a more widespread and severe impact on end face degradation). High-speed chain... The weighting of scratches can be increased for roads, and the weighting of stains can be increased for outdoor dusty environments. Next, a model for calculating the end-face contamination coefficient is constructed: End-face contamination coefficient = α1 × scratch density + β1 × stain coverage (α1 + β1 = 1). The calculation results are further calibrated and compensated using the baseline effective area to ensure fair evaluation of end faces of different sizes. Finally, the end-face contamination coefficient is limited and smoothed to ensure it falls strictly within the 0-1 range: End-face contamination coefficient < 0.2 is excellent, end-face contamination coefficient = 0.2-0.4 is good, end-face contamination coefficient = 0.4-0.6 is average, and end-face contamination coefficient > 0.6 is poor, requiring immediate cleaning or replacement. The coefficients are uploaded to the status assessment module in real time for calculating the comprehensive status value. This unified benchmark and weighted fusion construct a standardized end-face assessment system across the entire network: benchmark effective area eliminates model differences, enabling unified comparison across devices, scenarios, and manufacturers; weighted fusion combines physical damage and contamination into a single coefficient, automatically comparing the end-face contamination coefficient with the threshold, allowing the system to directly output maintain / clean / replace instructions without manual judgment. When the coefficient is high, it can be analyzed in reverse to determine whether scratches or stains are the primary cause, guiding precise maintenance actions: stains are prioritized for cleaning, while scratches are prioritized for inspection, grinding, or replacement. This process upgrades from fragmented qualitative judgment to integrated quantitative assessment, supporting automated, intelligent, and refined end-face resource management, significantly improving maintenance efficiency and link reliability.

[0037] In one embodiment, step S5, which involves obtaining an environmental temperature influence factor based on the environmental temperature characteristic parameters and obtaining a reflection-light decay joint attenuation correction coefficient based on the environmental temperature influence factor and the reflection-light decay joint attenuation coefficient, includes: S51. Obtain the temperature time series sequence in the ambient temperature characteristic parameters, and obtain the temperature fluctuation amplitude and temperature fluctuation frequency according to the temperature time series sequence, and obtain the temperature fluctuation intensity according to the temperature fluctuation amplitude and the temperature fluctuation frequency. S52. Obtain the temperature change rate sequence based on the temperature time series, and obtain the maximum positive change rate and the maximum negative change rate within a preset time window based on the temperature change rate sequence. S53. Obtain the temperature drastic change factor based on the maximum positive rate of change and the maximum negative rate of change; S54. Obtain preliminary temperature influence factors based on the temperature fluctuation intensity and the temperature drastic change factor; S55. Obtain the historical temperature influence factor sequence within a preset historical time period, and obtain the temperature influence trend factor based on the historical temperature influence factor sequence. S56. Obtain the ambient temperature influence factor based on the preliminary temperature influence factor and the temperature influence trend factor; S57. The reflection-light decay joint attenuation coefficient is linearly corrected according to the ambient temperature influence factor to obtain the reflection-light decay joint attenuation correction coefficient.

[0038] As described in steps S51-S57 above, the temperature time series sequence in the ambient temperature characteristic parameters is obtained, and the temperature fluctuation amplitude and frequency are calculated based on the sequence. The temperature fluctuation intensity is then obtained based on the temperature fluctuation amplitude and frequency. This process is based on real-time acquired ambient temperature time series data and relies on time series statistical analysis and signal processing methods. The specific process is as follows: First, ambient temperature is continuously acquired from the temperature sensor in the data acquisition module at fixed time intervals (e.g., 1 second / time), forming a temperature time series sequence within a preset time window (e.g., 5 minutes, 10 minutes). This ensures that the sequence length is sufficient to cover typical temperature fluctuation cycles and accurately reflect the temperature fluctuations. The process involves several steps: First, the temperature time series is preprocessed, using a moving average filter to remove random noise and anomalous jumps, preserving the true temperature fluctuation characteristics and avoiding calculation errors caused by sensor mistakes or instantaneous interference. Second, the temperature fluctuation amplitude is calculated, which is the difference between the maximum and minimum values ​​in the filtered temperature series; a larger amplitude indicates a more intense temperature change during that period. Third, the temperature fluctuation frequency is calculated by using peak counting or zero-crossing detection to count the number of times the temperature value crosses the mean line per unit time; a higher frequency indicates more frequent temperature alternation. Finally, the fluctuation amplitude and frequency are adaptively weighted and fused to obtain the temperature fluctuation intensity. The weighting follows the principle of amplitude dominance and frequency auxiliary, with amplitude weight typically set at 0.65-0.75 and frequency weight at 0.25-0.35. After fusion, normalization is performed to ensure the intensity value falls within the 0-1 range; a larger value indicates more severe temperature fluctuations and a stronger impact on the thermal stress of the connector. The entire process is completed in real time on the edge computing unit, with a computation latency of less than 100 milliseconds, meeting the engineering requirements for online monitoring and real-time correction. This approach, through time-series temperature feature extraction and intensity quantification, fundamentally addresses the shortcomings of static temperature monitoring. When the environment where the fiber optic connector is located experiences conditions such as air conditioning start / stop, day-night temperature differences, uneven equipment heating, or abnormal ventilation, the ambient temperature will fluctuate significantly or frequently. By accurately capturing these subtle changes through continuous time-series data, the invisible thermal stress disturbances are transformed into calculable amplitude and frequency characteristics. Then, through weighted fusion, the intensity of the temperature fluctuations is obtained, making latent thermal anomalies explicit and quantifiable. The system can determine whether the thermal environment poses a potential threat to the connector based on the magnitude of the temperature fluctuation intensity, identifying thermal stress risks in advance even if the current temperature is still within the rated range. This process upgrades from single-point temperature judgment to full-dimensional analysis of time-series fluctuations, advancing the fault detection window from after a fault occurs to the early stages of thermal degradation, providing accurate data support for subsequent temperature impact correction and preventative maintenance.

[0039] The temperature change rate sequence is obtained from the temperature time series. Within a preset time window, the maximum positive and maximum negative change rates are extracted. The temperature time series ensures smooth, noise-free, and anomalous data jumps, guaranteeing accurate rate of change calculations. Next, the first-order derivative of the temperature time series is calculated to obtain the temperature change rate sequence. The formula is: Temperature change rate (°C / second) = (Current temperature - Previous temperature) / Sampling interval. This sequence reflects the rate and direction of temperature change each second; positive values ​​indicate rising temperatures, and negative values ​​indicate falling temperatures. Then, within a preset time window (e.g., 1 minute, 3 minutes), extreme value searches are performed on the temperature change rate sequence to find the maximum value, which is the maximum positive change rate, representing the most drastic rate of temperature increase during that period. Simultaneously, the minimum value is found, which is the maximum negative change rate, representing the most drastic rate of temperature decrease during that period. Finally, the two extreme values ​​are standardized and mapped to the 0-1 interval for easier subsequent fusion calculations; larger values ​​indicate a higher intensity of temperature change in the corresponding direction. The entire calculation process is executed synchronously with temperature acquisition, outputting the latest extreme value results in real time without lag or omission. This temperature change rate calculation and extreme value extraction enable full monitoring, quantification, and traceability of instantaneous thermal shocks. When environmental conditions such as sudden air conditioning startup, direct sunlight, direct cold air, or high-power equipment operation occur, the temperature will rise or fall rapidly in a short period of time. Secondly, differential calculations quickly capture the rate of change, locating the maximum temperature rise / fall extreme value. The higher the value, the greater the instantaneous thermal stress the connector experiences and the higher the risk of deformation. The system can determine whether the thermal shock poses a threat to the connector based on the extreme value. If it exceeds the safety threshold, it automatically records the drastic event and marks the connector status. For multiple small drastic changes, the system can accumulate and evaluate the long-term deformation effect; for a single large drastic change, it directly triggers an early warning, reminding maintenance personnel to check the connection status and end-face contact. This process transforms the invisible instantaneous thermal shock into a quantifiable indicator, achieving an upgrade from imperceptible to fully perceptible, helping maintenance personnel accurately locate the source of thermal shock and promptly eliminate potential thermal degradation risks.

[0040] The temperature drastic change factor is calculated based on the maximum positive and maximum negative rates of change. By weighting and fusing the absolute values ​​of these two extreme rates of change, a drastic change factor characterizing the degree of instantaneous and drastic temperature change can be obtained. First, the two standardized extreme values ​​obtained in the previous step are read to confirm that both values ​​fall within the 0-1 range without any abnormal overflow. Second, the maximum positive and maximum negative rates of change are subjected to absolute value equivalence processing and normalization to eliminate the influence of positive and negative signs and uniformly measure the intensity of drastic changes. Then, a weighted calculation is performed, with the sum of the weights equal to 1, ensuring that the temperature drastic change factor strictly falls within the 0-1 range to avoid extreme values ​​interfering with subsequent calculations. The entire process requires no manual intervention, automatically completing the unified quantification of bidirectional drastic changes, and providing stable and reliable output. Thus, through the fusion of bidirectional extreme values, a comprehensive temperature drastic change risk assessment system is constructed, eliminating the one-sidedness of unidirectional assessments. The system incorporates the fastest heating and cooling rates into its calculations, emphasizing the most dramatic changes to ensure that the temperature drastic change factor accurately reflects the maximum instantaneous thermal stress experienced by the connector. When both extreme values ​​are high, the factor value increases significantly, indicating the connector is in a high-risk state for drastic changes. When one value is high and the other is normal, the factor increases moderately, suggesting unidirectional thermal shock. When both values ​​are low, the factor remains low, indicating stable temperature changes. This process shifts from judging unidirectional drastic changes to a comprehensive two-way assessment, improving the accuracy of thermal shock evaluation and providing high-quality drastic change characteristic input for subsequent temperature influence factor calculations.

[0041] The initial temperature influence factor is obtained by weighted fusion of temperature fluctuation intensity and temperature drastic change factor. Using the output temperature fluctuation intensity and temperature drastic change factor as inputs, a linear weighted fusion is performed to generate the initial temperature influence factor. First, two normalized indicators are read to confirm their validity and absence of anomalies. Second, weighting coefficients are determined based on the application scenario: for environments prone to drastic changes such as outdoor and base station environments, the drastic change factor weight is set to 0.55-0.65, and the fluctuation intensity weight is set to 0.35-0.45; for environments dominated by fluctuations such as computer rooms and indoor environments, the fluctuation intensity weight is set to 0.55-0.65, and the drastic change factor weight is set to 0.35-0.45. Then, a weighted summation calculation is performed using the formula: Initial Temperature Influence Factor = W3 × Temperature Fluctuation Intensity + W4 × Temperature Dramatic Change Factor, where W3 + W4 = 1. Finally, the results are limited to ensure they fall within the 0-1 range, guaranteeing indicator stability. The entire calculation process is executed in real time and updated synchronously with temperature acquisition. This multi-dimensional feature fusion constructs a comprehensive temperature impact assessment system, eliminating the limitations of single-feature assessments. The system incorporates both long-term fluctuations and short-term dramatic changes into the calculation, ensuring that the initial temperature impact factor simultaneously includes both cumulative and impact effects, accurately reflecting the comprehensive thermal stress experienced by the connector: when both fluctuation intensity and dramatic change factors are high, the factor increases significantly, indicating a very strong thermal impact; when one is high and the other is normal, the factor increases moderately; when both are low, the factor remains stable at a low level. This process transforms the assessment from a partial temperature judgment to a global thermal impact assessment, improving the reliability of temperature impact assessment and providing accurate basic data for subsequent trend correction.

[0042] The process involves obtaining a historical temperature influence factor sequence within a preset historical time period, and then identifying temperature influence trend factors based on this sequence. First, the historical temperature influence factor sequence for the preset historical time period (e.g., 24 hours, 7 days) is retrieved from a time-series database to ensure data continuity, absence of missing values, and absence of abnormal jumps. Second, the historical sequence is cleaned by using linear interpolation to fill in missing values ​​and removing outliers using the 3σ principle to ensure accurate trend analysis. Then, a trend fitting model is constructed, and the historical sequence is linearly fitted using the least squares method to obtain the slope of the fitted line, which directly reflects the rate of change of temperature influence. Next, the temperature influence trend factor is calculated, and the slope is normalized and mapped: a positive slope with a larger absolute value indicates a factor closer to 1, signifying accelerated deterioration of the thermal influence; a slope of 0 indicates a factor of 0.5, signifying a stable state; and a negative slope with a larger absolute value indicates a factor closer to 0, signifying a gradual weakening of the influence. Finally, the factor is smoothed to eliminate short-term fluctuation interference. The entire process is executed periodically (e.g., updated every 5 minutes), providing the latest trend data. By mining historical time-series data, a temperature impact prediction system is built, eliminating the lag in static assessments. The system continuously records historical data on temperature impact factors, and through fitting analysis, obtains the slope of change, converting it into a trend factor: when the trend factor continuously rises and exceeds a threshold, it indicates that the thermal impact has entered an accelerated deterioration phase, triggering an early warning even if the current state is acceptable; when the trend factor stabilizes around 0.5, it indicates a stable state and normal operation; when the trend factor decreases, it indicates a weakening impact, allowing for a reduction in monitoring frequency. This process transforms sudden thermal failures into predictable and preventable gradual degradation.

[0043] The ambient temperature influence factor is obtained based on the preliminary temperature influence factor and the temperature influence trend factor. Using the preliminary temperature influence factor and the temperature influence trend factor as inputs, an adaptive weighted fusion is used to generate the final ambient temperature influence factor: First, two normalized indicators are read to confirm data validity; second, weighting coefficients are set: the preliminary temperature influence factor has a weight of 0.6, and the temperature influence trend factor has a weight of 0.4, taking into account both the current state and future trends. For high-speed and critical links, the trend weight can be increased to 0.5; then, a weighted summation calculation is performed: Ambient temperature influence factor = 0.6 × preliminary temperature influence factor + 0.4 × temperature influence trend factor; finally, the result is limited and smoothed to ensure it strictly falls within the 0-1 range. A larger value indicates a stronger overall impact of temperature on the connector and a greater performance correction magnitude. After calculation, the factor is uploaded to the temperature correction execution unit in real time. This fusion constructs a comprehensive temperature influence quantification system, eliminating the problem of fragmented evaluation. The system calculates real-time thermal intensity and future degradation trends in a unified manner, ensuring that the environmental temperature impact factor truly reflects the risk throughout the entire lifecycle: when both the current impact and trend are severe, the factor increases significantly, and the correction magnitude is increased; when the current impact is high but the trend is improving, the factor decreases appropriately to avoid over-correction; when both the current situation and trend are favorable, the factor remains stable at a low level, and the correction magnitude is small. This process achieves a shift from unilateral temperature correction to precise compensation throughout the entire lifecycle, providing the most reliable basis for subsequent joint attenuation coefficient correction. The ambient temperature influence factor linearly corrects the reflection-optical attenuation joint attenuation coefficient, ultimately yielding the reflection-optical attenuation joint attenuation correction coefficient. Using the reflection-optical attenuation joint attenuation coefficient and the ambient temperature influence factor as inputs, the final compensation is achieved through a linear correction model: First, the reflection-optical attenuation joint attenuation coefficient (denoted as C) and the ambient temperature influence factor (denoted as T); second, a linear correction formula is constructed: Reflection-optical attenuation joint attenuation correction coefficient = C × (1 + λ × T), where λ is the temperature correction gain coefficient (usually taken as 0.2~0.5, which can be optimized according to the scenario), used to control the correction amplitude; since the coefficients are dimensionless, finally, the correction result needs to be normalized and limited to ensure it falls within the 0-1 range. A larger value indicates more severe optical degradation of the connector itself, independent of temperature. The correction coefficient is uploaded to the status assessment module in real time for calculating the comprehensive status value. Through linear temperature compensation, environmental interference is completely eliminated, restoring the true health state: when the temperature impact is significant, the correction coefficient is appropriately amplified to avoid masking true degradation; when the temperature impact is minor, the correction coefficient is close to the original value, accurately reflecting the status; when temperature causes artificially high performance, the correction coefficient is adjusted back to avoid missing degradation assessments. This process upgrades the assessment from temperature interference to true health assessment, completely resolving misjudgments and omissions, and supporting precise resource management.

[0044] In one embodiment, step S6, which involves obtaining a comprehensive state value based on the reflection-optical attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and performing resource management on the target fiber optic connector based on the comprehensive state value, includes: S61. The reflection-light decay joint attenuation correction coefficient and the end face contamination coefficient are weighted and calculated to obtain the comprehensive state value; S62. Obtain a preset first state threshold and a second state threshold, wherein the first state threshold is greater than the second state threshold; S63. Based on the threshold range of the comprehensive status value, determine the resource management action of the target fiber optic connector: if the comprehensive status value is greater than the first status threshold, generate a connector replacement command; if the comprehensive status value is between the second status threshold and the first status threshold, generate an end face cleaning command; if the comprehensive status value is less than the second status threshold, generate a maintenance operation command, and update the health score of the target fiber optic connector based on the comprehensive status value.

[0045] As described in steps S61-S63 above, the reflection-light decay joint attenuation correction coefficient and the end-face contamination coefficient are weighted and calculated to obtain a comprehensive state value. This weighted fusion unifies dispersed, multi-source, and heterogeneous indicators into a single, intuitive, and reliable comprehensive state value. The calculation formula is: Comprehensive State Value = Reflection-Light Decay Joint Attenuation Correction Coefficient * Weight Value of Reflection-Light Decay Joint Attenuation Correction Coefficient + End-face Contamination Coefficient * Weight Value of End-face Contamination Coefficient, achieving an upgrade from chaotic judgment based on multiple indicators to accurate assessment using a single value. The system first strongly correlates and fuses optical transmission degradation (already temperature-compensated) with end-face physical degradation, significantly amplifying the risk of dual degradation and preventing a single qualified indicator from masking a severely abnormal one. Simultaneously, the product form inherently possesses high sensitivity, showing a significant increase in the early stages of degradation, enabling early detection and early warning. Once the comprehensive state value is calculated, both the system and maintenance personnel can judge the overall risk using a single value, eliminating the need for manual comparison of multiple indicators and completely resolving the problem of ambiguous decision-making criteria. This process transforms complex and unquantifiable state judgments into simple and automated numerical calculations, giving the entire system an intelligent and automated foundation. The system acquires preset first and second state thresholds, wherein the first state threshold is greater than the second state threshold. Specifically: First, the system reads preset threshold base data from a local configuration file or cloud parameter library, including a default first state threshold (e.g., 0.35) and a default second state threshold (e.g., 0.15). These thresholds are determined by industry standards, manufacturer specifications, and laboratory test data. Second, the system acquires basic attribute information of the current target fiber optic connector, including model, speed, application scenario, importance level, service life, and installation location, for scenario-based matching. Then, the system activates a scenario adaptive adjustment mechanism to fine-tune the base thresholds: for core backbone links, high-speed links, and outdoor harsh environment links, the thresholds are appropriately lowered to improve sensitivity; for ordinary access links, low-speed links, and indoor stable environment links, the thresholds can be appropriately raised to avoid frequent alarms. Next, the adjusted thresholds are validated to ensure that the first state threshold is greater than the second state threshold and both are within the 0-1 range, preventing logical errors from causing decision failure. Finally, the final determined first threshold (T1) and second threshold (T2) are synchronized to the state judgment unit for interval comparison and action generation in step S63. The entire process is completed automatically without human intervention, and supports batch configuration and unified distribution. By defining dual thresholds, a unified, objective, and quantifiable status classification scale is established for all fiber optic connectors in the network, completely eliminating the differences in subjective human judgment. Based on the threshold range of the comprehensive status value, the resource management action of the target fiber optic connector is determined: if the comprehensive status value is greater than the first status threshold, a connector replacement command is generated; if the comprehensive status value is between the second status threshold and the first status threshold, an end-face cleaning command is generated; if the comprehensive status value is less than the second status threshold, a maintenance operation command is generated, and the health score of the target fiber optic connector is updated based on the comprehensive status value. This hierarchical automated closed-loop processing transforms traditional passive, lagging, and extensive management into proactive, forward-looking, and precise management. The system monitors, calculates, and judges in real time, identifying degradation trends and triggering corresponding actions before faults occur: mild degradation is automatically reminded to clean, preventing further deterioration; moderate degradation is addressed promptly to restore performance; severe degradation is immediately replaced to prevent interruptions. The health score is continuously and dynamically updated, fully recording the connector's entire lifecycle from installation, operation, degradation to replacement, supporting predictive maintenance. All actions are executed and recorded automatically, requiring no manual intervention, achieving remote centralized intelligent management. This process addresses operational delays, significantly reduces failure rates and operational costs, and improves resource utilization and network reliability.

[0046] like Figure 2 As shown, this application also provides a fiber optic connector status awareness and resource management system, including: The data acquisition module is used to acquire the real-time status dataset of the target optical fiber connector, wherein the real-time status dataset includes optical power characteristic parameters, return loss characteristic parameters, end face geometric characteristic parameters, and ambient temperature characteristic parameters. The optical attenuation stress coefficient calculation module is used to obtain the insertion loss fluctuation sequence and instantaneous optical power jitter rate based on the optical power characteristic parameters, and to obtain the optical attenuation stress coefficient based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate. The attenuation coefficient calculation module is used to obtain the reflection degradation coefficient based on the return loss characteristic parameters, and to obtain the reflection-optical attenuation joint attenuation coefficient based on the reflection degradation coefficient and the optical attenuation stress coefficient. The end-face contamination coefficient calculation module is used to obtain the scratch density and stain coverage rate based on the end-face geometric feature parameters, and to obtain the end-face contamination coefficient based on the scratch density and stain coverage rate. The temperature influence correction module is used to obtain the environmental temperature influence factor based on the environmental temperature characteristic parameters, and to obtain the reflection-light decay joint attenuation correction coefficient based on the environmental temperature influence factor and the reflection-light decay joint attenuation coefficient. The status assessment module is used to obtain a comprehensive status value based on the reflection-light attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and to perform resource management on the target fiber optic connector based on the comprehensive status value.

[0047] In one embodiment, the optical attenuation stress coefficient calculation module includes: An offset calculation unit is used to obtain the loss fluctuation amplitude and loss fluctuation frequency based on the insertion loss fluctuation sequence, and to obtain the dynamic offset of the insertion loss based on the loss fluctuation amplitude and the loss fluctuation frequency. The impact calculation unit is used to obtain the peak jitter amplitude and jitter duration based on the instantaneous optical power jitter rate, and to obtain the transient impact of optical power based on the peak jitter amplitude and the jitter duration. The optical attenuation stress factor calculation unit is used to obtain the initial optical attenuation stress factor based on the insertion loss dynamic offset and the optical power transient impact. The optical attenuation stress calculation unit is used to obtain the reference insertion loss tolerance threshold of the optical fiber connector, and to obtain the optical attenuation stress coefficient based on the initial optical attenuation stress factor and the reference insertion loss tolerance threshold. This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0048] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0049] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0050] It should be noted that, in this document, the term "include" or any other variations thereof is intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Without further limitation, the inclusion of an element by a statement that includes a… does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0051] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for status awareness and resource management of fiber optic connectors, characterized in that, include: Obtain a real-time status dataset of the target fiber optic connector, wherein the real-time status dataset includes optical power characteristic parameters, return loss characteristic parameters, end face geometric characteristic parameters, and ambient temperature characteristic parameters; The insertion loss fluctuation sequence and instantaneous optical power jitter rate are obtained based on the optical power characteristic parameters, and the optical attenuation stress coefficient is obtained based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate. The reflection degradation coefficient is obtained based on the return loss characteristic parameters, and the reflection-optical attenuation joint attenuation coefficient is obtained based on the reflection degradation coefficient and the optical attenuation stress coefficient. The scratch density and stain coverage are obtained based on the end face geometric feature parameters, and the end face contamination coefficient is obtained based on the scratch density and stain coverage. The ambient temperature influence factor is obtained based on the ambient temperature characteristic parameters, and the reflection-light decay joint attenuation correction coefficient is obtained based on the ambient temperature influence factor and the reflection-light decay joint attenuation coefficient. The comprehensive status value is obtained based on the reflection-optical attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and resource management is performed on the target optical fiber connector based on the comprehensive status value.

2. The fiber optic connector status awareness and resource management method according to claim 1, characterized in that, The step of obtaining the optical attenuation stress coefficient based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate includes: The insertion loss fluctuation amplitude and frequency are obtained from the insertion loss fluctuation sequence, and the insertion loss dynamic offset is obtained from the insertion loss fluctuation amplitude and the insertion loss fluctuation frequency. The peak jitter amplitude and jitter duration are obtained based on the instantaneous optical power jitter rate, and the transient impact of optical power is obtained based on the peak jitter amplitude and the jitter duration. The initial optical decay stress factor is obtained based on the insertion loss dynamic offset and the optical power transient impact. Obtain the reference insertion loss tolerance threshold of the fiber optic connector, and obtain the optical attenuation stress coefficient based on the initial optical attenuation stress factor and the reference insertion loss tolerance threshold.

3. The fiber optic connector status awareness and resource management method according to claim 1, characterized in that, The steps of obtaining the reflection degradation coefficient based on the return loss characteristic parameters, and obtaining the reflection-optical attenuation joint attenuation coefficient based on the reflection degradation coefficient and the optical attenuation stress coefficient, include: The reflection peak energy ratio and the return loss baseline drift are obtained based on the return loss characteristic parameters, and the reflection degradation coefficient is obtained based on the reflection peak energy ratio and the return loss baseline drift. The low-frequency disturbance component and the high-frequency noise component in the optical power characteristic parameters are obtained, and the stress disturbance coefficient is obtained based on the low-frequency disturbance component and the high-frequency noise component. Obtain the historical reflection degradation coefficient sequence within a preset time window, and obtain the reflection degradation trend factor based on the historical reflection degradation coefficient sequence; The reflection degradation coefficient, the stress disturbance coefficient, the reflection degradation trend factor, and the optical decay stress coefficient are weighted and fused to obtain the reflection-optical decay joint attenuation coefficient.

4. The fiber optic connector status awareness and resource management method according to claim 1, characterized in that, The steps of obtaining the scratch density and stain coverage rate based on the end face geometric feature parameters, and obtaining the end face contamination coefficient based on the scratch density and stain coverage rate, include: Obtain the scratch image and the stain area image from the geometric feature parameters of the end face; The total length of the scratch and the area of ​​the scratch-affected region are obtained from the scratch image, and the scratch density is obtained from the total length of the scratch and the area of ​​the scratch-affected region. The total area and morphological complexity of the stain are obtained from the stain area image, and the stain coverage rate is obtained from the total area and morphological complexity of the stain. Obtain the preset effective area of ​​the reference end face, and obtain the end face contamination coefficient based on the scratch density, the stain coverage rate, and the effective area of ​​the reference end face.

5. The fiber optic connector status awareness and resource management method according to claim 1, characterized in that, The steps of obtaining the environmental temperature influence factor based on the environmental temperature characteristic parameters, and obtaining the reflection-light decay joint attenuation correction coefficient based on the environmental temperature influence factor and the reflection-light decay joint attenuation coefficient, include: Obtain the temperature time series sequence from the ambient temperature characteristic parameters, and obtain the temperature fluctuation amplitude and temperature fluctuation frequency based on the temperature time series sequence, and obtain the temperature fluctuation intensity based on the temperature fluctuation amplitude and the temperature fluctuation frequency; The temperature change rate sequence is obtained based on the temperature time series, and the maximum positive change rate and the maximum negative change rate within a preset time window are obtained based on the temperature change rate sequence. The temperature drastic change factor is obtained based on the maximum positive rate of change and the maximum negative rate of change; Preliminary temperature influence factors are obtained based on the intensity of temperature fluctuations and the temperature drastic change factor. Obtain the historical temperature influence factor sequence within a preset historical time period, and obtain the temperature influence trend factor based on the historical temperature influence factor sequence; The environmental temperature influence factor is obtained based on the preliminary temperature influence factor and the temperature influence trend factor. The reflection-light decay joint attenuation coefficient is linearly corrected based on the environmental temperature influence factor to obtain the reflection-light decay joint attenuation correction coefficient.

6. The fiber optic connector status awareness and resource management method according to claim 1, characterized in that, The step of obtaining a comprehensive state value based on the reflection-optical attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and performing resource management on the target fiber optic connector based on the comprehensive state value, includes: The combined state value is obtained by weighting the reflection-light decay joint attenuation correction coefficient and the end face contamination coefficient. Obtain a preset first state threshold and a second state threshold, wherein the first state threshold is greater than the second state threshold; Based on the threshold range in which the comprehensive status value falls, determine the resource management action of the target fiber optic connector: If the overall status value is greater than the first status threshold, a connector replacement instruction is generated; If the overall state value is between the second state threshold and the first state threshold, then an end face cleaning instruction is generated; If the overall status value is less than the second status threshold, a maintenance operation instruction is generated, and the health score of the target fiber optic connector is updated based on the overall status value.

7. A fiber optic connector status awareness and resource management system, characterized in that, include: The data acquisition module is used to acquire the real-time status dataset of the target optical fiber connector, wherein the real-time status dataset includes optical power characteristic parameters, return loss characteristic parameters, end face geometric characteristic parameters, and ambient temperature characteristic parameters. The optical attenuation stress coefficient calculation module is used to obtain the insertion loss fluctuation sequence and instantaneous optical power jitter rate based on the optical power characteristic parameters, and to obtain the optical attenuation stress coefficient based on the insertion loss fluctuation sequence and the instantaneous optical power jitter rate. The attenuation coefficient calculation module is used to obtain the reflection degradation coefficient based on the return loss characteristic parameters, and to obtain the reflection-optical attenuation joint attenuation coefficient based on the reflection degradation coefficient and the optical attenuation stress coefficient. The end-face contamination coefficient calculation module is used to obtain the scratch density and stain coverage rate based on the end-face geometric feature parameters, and to obtain the end-face contamination coefficient based on the scratch density and stain coverage rate. The temperature influence correction module is used to obtain the environmental temperature influence factor based on the environmental temperature characteristic parameters, and to obtain the reflection-light decay joint attenuation correction coefficient based on the environmental temperature influence factor and the reflection-light decay joint attenuation coefficient. The status assessment module is used to obtain a comprehensive status value based on the reflection-light attenuation joint attenuation correction coefficient and the end-face contamination coefficient, and to perform resource management on the target fiber optic connector based on the comprehensive status value.

8. The fiber optic connector status awareness and resource management system according to claim 7, characterized in that, The optical attenuation stress coefficient calculation module includes: An offset calculation unit is used to obtain the loss fluctuation amplitude and loss fluctuation frequency based on the insertion loss fluctuation sequence, and to obtain the dynamic offset of the insertion loss based on the loss fluctuation amplitude and the loss fluctuation frequency. The impact calculation unit is used to obtain the peak jitter amplitude and jitter duration based on the instantaneous optical power jitter rate, and to obtain the transient impact of optical power based on the peak jitter amplitude and the jitter duration. The optical attenuation stress factor calculation unit is used to obtain the initial optical attenuation stress factor based on the insertion loss dynamic offset and the optical power transient impact. The optical attenuation stress calculation unit is used to obtain the reference insertion loss tolerance threshold of the optical fiber connector, and to obtain the optical attenuation stress coefficient based on the initial optical attenuation stress factor and the reference insertion loss tolerance threshold.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images