Intelligent identification method and system for early failure of optical module

Abstract

The invention relates to the technical field of early failure identification of optical modules, in particular to an intelligent identification method and system for early failure of an optical module, which comprises the following steps of: confirming an optical module to be detected and a detection time period set, sequentially extracting detection time periods from the detection time period set, and executing the following operations on the extracted detection time periods: confirming adjacent detection time periods, and obtaining a comprehensive risk score, setting a dynamic early warning threshold value, if the comprehensive risk score is greater than the dynamic early warning threshold value, taking the comprehensive risk score as an abnormal score, taking the adjacent detection time period as a detection time period, and returning to the step of obtaining the comprehensive risk score based on the optical module early failure detection instruction, the detection time period and the optical module to be detected. And when the detection time periods in the detection time period set are all extracted, summarizing the abnormal scores to obtain an abnormal score set, calculating the number of the abnormal scores in the abnormal score set, and formulating early warning measures. The real-time performance and the accuracy of the optical module can be improved.

Description

Technical Field

[0001] This invention relates to the field of early failure identification technology for optical modules, and in particular to an intelligent identification method and system for early failure of optical modules. Background Technology

[0002] Optical modules are key components in modern communication networks, used to convert electrical signals into optical signals or vice versa, thereby enabling high-speed data transmission. Early failure refers to the performance degradation or functional malfunction of an optical module before its expected lifespan under normal operating conditions. Intelligent identification refers to the automatic detection and identification of early failure signs of optical modules using advanced technologies such as machine learning, artificial intelligence, and data mining.

[0003] Traditional optical module testing methods primarily rely on regular manual inspections and maintenance, which is inefficient and makes it difficult to detect early failures in a timely manner. Furthermore, some systems use fixed thresholds to test the performance parameters of optical modules, but this method cannot adapt to changes in the optical module's environment and operating conditions, easily leading to false alarms or missed alarms. Therefore, improving the real-time performance and accuracy of optical modules is an urgent technical problem to be solved. Summary of the Invention

[0004] This invention provides an intelligent identification method for early failure of optical modules and a computer-readable storage medium, the main purpose of which is to improve the real-time performance and accuracy of optical modules.

[0005] To achieve the above objectives, the present invention provides an intelligent identification method for early failure of optical modules, comprising: Once the optical module to be tested and the set of testing time periods are identified, the testing time periods are extracted sequentially from the set of testing time periods, and the following operations are performed on each extracted testing time period: Based on the extracted detection time periods, adjacent detection time periods are identified from the set of detection time periods. Among them, adjacent detection time periods are adjacent to the detection time periods and lag behind the detection time periods. Receive early failure detection command for optical module, obtain comprehensive risk score based on early failure detection command, detection period and optical module to be tested, and set dynamic early warning threshold; If the comprehensive risk score is greater than the dynamic warning threshold, the comprehensive risk score is taken as an abnormal score, the adjacent detection period is taken as the detection period, and the step of obtaining the comprehensive risk score based on the early failure detection instruction of the optical module, the detection period and the optical module to be tested is returned until all the detection periods in the detection period set are extracted. The abnormal scores are aggregated to obtain an abnormal score set. The number of abnormal scores in the abnormal score set is calculated. If the number of abnormal scores is greater than the preset continuous abnormal threshold, the pre-built early warning system is triggered to obtain an early warning signal. Early warning measures are then formulated based on the early warning signal.

[0006] Optionally, the step of obtaining a comprehensive risk score based on the early failure detection command of the optical module, the detection period, and the optical module under test includes: Based on the early failure detection command and detection period of the optical module, the optical module to be detected is detected to obtain the real-time feature vector of the optical module. The real-time feature vector of the optical module is input into the pre-built optical module health model to obtain the deviation value, and the real-time feature vector of the optical module is input into the pre-built optical module fault migration model to obtain the fault probability. A comprehensive risk score is calculated based on preset fault probability weights, preset deviation weights, deviation values, and fault probabilities.

[0007] Optionally, the step of detecting the optical module to be detected according to the early failure detection command and the detection period to obtain the real-time feature vector of the optical module includes: According to the early failure detection command of the optical module, the optical module under test is detected during the detection period to obtain the received optical power value sequence, the laser bias current value sequence and the laser temperature sequence. The historical detection time period is obtained based on the detection time period, and the historical laser bias current value sequence is obtained based on the historical detection time period, the preset historical detection frequency and the optical module to be tested. Calculate the temperature-current correlation coefficient based on the laser temperature sequence and the laser bias current value sequence; The received optical power sample entropy is calculated based on the received optical power value sequence, and the optical module aging rate is calculated based on the historical laser bias current value sequence. The temperature-current correlation coefficient, the sample entropy of received optical power, and the aging rate of the optical module are used as the real-time feature vector of the optical module.

[0008] Optionally, the step of calculating the temperature-current correlation coefficient based on the laser temperature sequence and the laser bias current value sequence includes: The laser temperature sequence and the laser bias current value sequence are time-aligned to obtain the aligned temperature sequence and the aligned bias current value sequence. Data missing detection is performed on the aligned temperature sequence to obtain the missing detection result, where the missing detection result indicates whether there is missing data or not. If the missing data detection result indicates that there is missing data, the missing data of the aligned temperature sequence is filled to obtain the standard temperature sequence; If the missing data detection result is that there is no missing data, then the aligned temperature sequence is used as the standard temperature sequence, wherein the standard temperature sequence includes multiple standard temperatures; A standard bias current value sequence is obtained based on the aligned bias current value sequence, wherein the standard bias current value sequence includes multiple standard bias current values; Calculate the mean temperature and mean current values ​​of the standard temperature series and the standard bias current series respectively, and calculate the temperature-current correlation coefficient based on the mean temperature and mean current values.

[0009] Optionally, the formula for calculating the temperature-current correlation coefficient is as follows: ; in, This represents the temperature-current correlation coefficient. This indicates the number of standard temperatures in a standard temperature series or the number of standard bias current values ​​in a standard bias current value series. Indicates the first A standard bias current value. Indicates the average current. Indicates the first A standard temperature, This represents the average temperature.

[0010] Optionally, calculating the received optical power sample entropy based on the received optical power value sequence includes: Calculate the standard deviation of received optical power in the received optical power value sequence, and calculate the similarity tolerance threshold based on the standard deviation of received optical power and the preset similarity tolerance coefficient. The similarity tolerance threshold is the product of the standard deviation of received optical power and the similarity tolerance coefficient. The received optical power value sequence is segmented using a preset mode length to obtain a segmented optical power set, wherein the segmented optical power set includes multiple segmented optical power groups, and each segmented optical power group includes multiple segmented optical powers. Segmented optical power groups are extracted sequentially from the segmented optical power group set, and the extracted segmented optical power groups are removed from the segmented optical power group set to obtain the remaining segmented optical power group set, wherein the remaining segmented optical power group set includes multiple remaining segmented optical power groups. The number of similar combination pairs is obtained based on the segmented optical power group and the remaining segmented optical power group set. The number of similar combination pairs is summarized to obtain the number of multiple similar combination pairs corresponding to the segmented optical power group set. Calculate the number of segmented optical power groups in the segmented optical power group set, calculate the number of comprehensive similar combination pairs based on the number of multiple similar combination pairs, and calculate the average ratio based on the number of segmented optical power groups and the number of comprehensive similar combination pairs. Perform a length increment operation on the mode length to obtain the updated mode length. Calculate the updated average ratio based on the updated mode length and the received optical power value sequence. Calculate the received optical power sample entropy based on the average ratio and the updated average ratio.

[0011] Optionally, obtaining the number of similar combination pairs based on the segmented optical power groups and the remaining segmented optical power group sets includes: The segmented optical power group is combined with each remaining segmented optical power group to obtain a set of combination pairs. The set of combination pairs includes multiple combination pairs, and each combination pair includes one segmented optical power group and one remaining segmented optical power group. Extract combination pairs sequentially from the combination pair set, calculate multiple absolute differences between combination pairs, and determine whether multiple absolute differences are all less than the similarity tolerance threshold; If multiple absolute differences are all less than the similarity tolerance threshold, then the combined pair is regarded as a similar combined pair. Summarize similar pairs to obtain a set of similar pairs, and count the number of similar pairs in the set.

[0012] Optionally, calculating the optical module aging rate based on the historical laser bias current value sequence includes: The historical laser bias current value sequence is grouped according to the historical detection frequency to obtain a bias current value set. The bias current value set includes multiple bias current value groups, and each bias current value group includes multiple historical laser bias current values. Obtain the median bias current value of each bias current value group in the bias current value group set to obtain the median bias current value set, and obtain the grouping sequence number set in the bias current value group set; The median bias current value set is indexed using the group sequence number set to create a bias current coordinate sequence. The bias current coordinate sequence includes multiple bias current coordinates, and the horizontal axis of the bias current coordinate is the group sequence number, while the vertical axis of the bias current coordinate is the bias current value. Linear regression is performed on the bias current coordinate sequence to obtain a fitted straight line. The slope of the fitted straight line is obtained, and the aging rate of the optical module is calculated based on the slope of the straight line.

[0013] Optionally, setting the dynamic early warning threshold includes: Obtain the historical risk score array of the optical module to be tested, and obtain the individual historical baseline value from the historical risk score array, where the individual historical baseline value is the largest historical risk score in the historical risk score array; Obtain the optical module type of the optical module to be tested, and retrieve a set of optical modules of the same type from the pre-built online testing system based on the optical module type. The set of optical modules of the same type includes one or more optical modules of the same type. Based on the detection period and the set of similar optical modules, a risk score set of similar optical modules is obtained, and the group baseline value is calculated based on the risk score set of similar optical modules. The maximum baseline value is obtained from the individual historical baseline value and the group baseline value. The dynamic early warning threshold is calculated based on the maximum baseline value and the preset safety factor.

[0014] To achieve the above objectives, the present invention also provides an intelligent identification system for early failure of optical modules, comprising: The detection period extraction module is used to identify the optical module to be tested and the detection period set, extract the detection period from the detection period set in sequence, and perform the following operation on each extracted detection period: based on the extracted detection period, identify adjacent detection periods from the detection period set, wherein the adjacent detection period is adjacent to the detection period and lags behind the detection period; The comprehensive risk score acquisition module is used to receive early failure detection instructions for optical modules, acquire a comprehensive risk score based on the early failure detection instructions, the detection period, and the optical module to be tested, and set dynamic early warning thresholds. The loop detection module is used to treat the comprehensive risk score as an abnormal score and the adjacent detection period as a detection period if the comprehensive risk score is greater than the dynamic early warning threshold. It then returns to the step of obtaining the comprehensive risk score based on the early failure detection instruction of the optical module, the detection period and the optical module to be tested, until all the detection periods in the detection period set have been extracted. The early warning measure formulation module is used to summarize the abnormal scores, obtain the abnormal score set, calculate the number of abnormal scores in the abnormal score set, and if the number of abnormal scores is greater than the preset continuous abnormal threshold, the pre-built early warning system is triggered to obtain an early warning signal, and early warning measures are formulated based on the early warning signal.

[0015] To address the above problems, the present invention also provides an electronic device, the electronic device comprising: Memory, storing at least one instruction; The processor executes the instructions stored in the memory to implement the intelligent identification method for early failure of the optical module described above.

[0016] To address the aforementioned problems, the present invention also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor in an electronic device to implement the aforementioned intelligent identification method for early failure of optical modules.

[0017] To address the problems described in the background art, this invention identifies the optical module to be tested and the set of testing time periods. It then sequentially extracts testing time periods from the set and performs the following operations on each extracted testing time period: This invention determines the set of testing time periods and sequentially extracts the testing time periods, rationally dividing the entire testing process according to time. This allows for phased and planned testing of the optical module, facilitating data analysis and comparison across different time periods and aiding in the orderly execution of subsequent steps. Based on the extracted testing time periods, adjacent testing time periods are identified from the set of testing time periods. These adjacent testing time periods are adjacent to the current testing time period but lag behind it. By identifying adjacent testing time periods, this invention enables [the invention to] [achieve] [a certain level of efficiency] in the time dimension. By establishing connections between different detection periods, and recognizing the potential continuity and correlation between data from adjacent periods, comparing detection results from adjacent periods can more acutely reveal trends in optical module performance, helping to detect potential early failure risks in advance. This invention receives early failure detection commands for optical modules, and based on these commands, the detection period, and the optical module under test, obtains a comprehensive risk score and sets dynamic warning thresholds. This invention quantifies the risk of early failure of optical modules by obtaining a comprehensive risk score, using a specific numerical value to represent the degree of risk of the optical module within a specific detection period. This makes the detection results more intuitive and objective, facilitating subsequent analysis and judgment. Setting dynamic warning thresholds allows for adjustments based on changes in performance. The detection period and the actual situation of the optical module are flexibly adjusted. If the comprehensive risk score is greater than the dynamic warning threshold, the comprehensive risk score is marked as an abnormal score, and the adjacent detection period is marked as the detection period. The process returns to the step of obtaining the comprehensive risk score based on the early failure detection command of the optical module, the detection period, and the optical module to be tested, until all detection periods in the detection period set have been extracted. When the comprehensive risk score is greater than the dynamic warning threshold, the present invention marks it as an abnormal score, which can promptly detect abnormal situations of the optical module in certain periods. By continuously cyclically detecting adjacent periods, the state changes of the optical module are continuously monitored to ensure that no possible early failure signs are missed. The abnormal scores are summarized to obtain the abnormal... The invention uses a constant score set and calculates the number of abnormal scores in the abnormal score set. If the number of abnormal scores exceeds a preset continuous abnormality threshold, a pre-built early warning system is triggered, generating an early warning signal. Based on this signal, early warning measures are formulated. This invention summarizes abnormal scores to obtain an abnormal score set and calculates the number of abnormal scores, allowing for a comprehensive assessment of the optical module's anomalies across multiple detection periods. An anomaly in a single period may be due to accidental factors, while anomalies across multiple consecutive periods better reflect the possibility of early failure of the optical module. Formulating early warning measures based on the early warning signal allows for intervention before the optical module actually fails, thereby reducing losses caused by optical module failure and ensuring the stable operation of the communication system. Therefore, this invention can improve the real-time performance and accuracy of optical modules. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating an intelligent identification method for early failure of an optical module provided in an embodiment of the present invention. Figure 2 This is a functional block diagram of an intelligent identification system for early failure of optical modules provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device that implements the intelligent identification method for early failure of the optical module according to an embodiment of the present invention.

[0019] Explanation of reference numerals in the attached figures: 10. Electronic device; 11. Processor; 12. Memory; 13. Bus.

[0020] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0022] This application provides an intelligent method for identifying early failure of optical modules. The executing entity of this intelligent method for identifying early failure of optical modules includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the intelligent method for identifying early failure of optical modules can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0023] Reference Figure 1 The diagram shown is a flowchart illustrating an intelligent identification method for early failure of an optical module according to an embodiment of the present invention. In this embodiment, the intelligent identification method for early failure of an optical module includes: S1. Identify the optical module to be tested and the set of testing time periods, extract the testing time periods from the set of testing time periods in sequence, and perform the following operations on each extracted testing time period.

[0024] It should be explained that the optical module under test refers to the optical module that requires early failure detection. An optical module is a device used in fiber optic communication to convert electrical signals into optical signals, or vice versa. An optical module typically includes a transmitter (laser), a receiver (photodetector), driver circuitry, and interfaces. A detection time period set refers to a collection of multiple detection time periods. Each detection time period is a specific time segment used for performance monitoring and analysis of the optical module.

[0025] S2. Based on the extracted detection time period, identify adjacent detection time periods from the detection time period set. The adjacent detection time periods are adjacent to the detection time periods and lag behind the detection time periods.

[0026] It should be explained that adjacent detection periods refer to time periods that are adjacent to and lag behind the current detection period within the detection period cluster. They are used to compare the data of the current detection period in order to detect trends and changes.

[0027] S3. Receive the early failure detection command for the optical module, obtain a comprehensive risk score based on the early failure detection command, the detection period, and the optical module to be tested, and set a dynamic early warning threshold.

[0028] In detail, the comprehensive risk score obtained based on the early failure detection command of the optical module, the detection period, and the optical module under test includes: Based on the early failure detection command and detection period of the optical module, the optical module to be detected is detected to obtain the real-time feature vector of the optical module. The real-time feature vector of the optical module is input into the pre-built optical module health model to obtain the deviation value, and the real-time feature vector of the optical module is input into the pre-built optical module fault migration model to obtain the fault probability. A comprehensive risk score is calculated based on preset fault probability weights, preset deviation weights, deviation values, and fault probabilities.

[0029] It should be explained that the early failure detection command for optical modules refers to the command initiated by the system to begin performance monitoring and analysis of the optical module under test. The real-time feature vector of the optical module refers to the temperature-current correlation coefficient, received optical power sample entropy, and aging rate of the optical module extracted from it during a specific detection period, reflecting the performance status of the optical module under test during that period. The optical module health model is a model pre-built by collecting a large number of feature vectors of the optical modules under test under normal operating conditions and training them using machine learning algorithms (such as linear regression, support vector machines, etc.). The optical module fault transfer model is a model pre-trained using time-series data (full lifecycle data from normal operation to failure) of known failure cases, employing time-series-based transfer learning (e.g., using a one-dimensional convolutional neural network or LSTM as a feature extractor). The deviation value is the output value after inputting the real-time feature vector of the optical module into the pre-built optical module health model. The failure probability refers to the probability that the optical module under test will fail under the real-time feature vector of the optical module. In the step of calculating the comprehensive risk score based on the preset fault probability weight, deviation degree weight, deviation degree value, and fault probability, the comprehensive risk score = fault probability weight. Failure probability + deviation value weight Deviation degree value. The failure probability weight is a pre-set value used to adjust the contribution of failure probability to the overall risk score. A larger failure probability weight indicates a greater impact of failure probability on the overall risk score. The deviation degree value weight is a pre-set value used to adjust the contribution of deviation degree value to the overall risk score. A larger deviation degree value weight indicates a greater impact of deviation degree value on the overall risk score. In this invention, the sum of the deviation degree value weight and the failure probability weight is 1.

[0030] Specifically, the step of detecting the optical module to be detected according to the early failure detection command and the detection period to obtain the real-time feature vector of the optical module includes: According to the early failure detection command of the optical module, the optical module under test is detected during the detection period to obtain the received optical power value sequence, the laser bias current value sequence and the laser temperature sequence. The historical detection time period is obtained based on the detection time period, and the historical laser bias current value sequence is obtained based on the historical detection time period, the preset historical detection frequency and the optical module to be tested. Calculate the temperature-current correlation coefficient based on the laser temperature sequence and the laser bias current value sequence; The received optical power sample entropy is calculated based on the received optical power value sequence, and the optical module aging rate is calculated based on the historical laser bias current value sequence. The temperature-current correlation coefficient, the sample entropy of received optical power, and the aging rate of the optical module are used as the real-time feature vector of the optical module.

[0031] It should be explained that the received optical power value sequence refers to the time series of optical power values ​​measured at the receiver of the optical module during the detection period. This is used to detect changes in optical signal intensity and helps identify potential link problems. The laser bias current value sequence refers to the time series of bias current values ​​of the laser in the optical module during the detection period. This reflects changes in the laser's operating current and helps identify potential aging or faults. The laser temperature sequence refers to the time series of temperature values ​​of the laser in the optical module during the detection period. This helps identify potential heat dissipation problems or temperature control circuit faults. Obtaining the historical detection period based on the detection period refers to determining a period prior to the current detection period. For example, the historical detection period is 30 days prior to the current detection period. The historical laser bias current value sequence refers to the time series of bias current values ​​of the laser in the optical module under test during historical detection periods. This is used to assess long-term performance changes of the laser. The temperature-current correlation coefficient is the correlation coefficient between the laser bias current sequence and the laser temperature sequence. This is used to assess the laser's temperature control performance and detect whether the temperature control circuit is functioning correctly.

[0032] Specifically, the calculation of the temperature-current correlation coefficient based on the laser temperature sequence and the laser bias current value sequence includes: The laser temperature sequence and the laser bias current value sequence are time-aligned to obtain the aligned temperature sequence and the aligned bias current value sequence. Data missing detection is performed on the aligned temperature sequence to obtain the missing detection result, where the missing detection result indicates whether there is missing data or not. If the missing data detection result indicates that there is missing data, the missing data of the aligned temperature sequence is filled to obtain the standard temperature sequence; If the missing data detection result is that there is no missing data, then the aligned temperature sequence is used as the standard temperature sequence, wherein the standard temperature sequence includes multiple standard temperatures; A standard bias current value sequence is obtained based on the aligned bias current value sequence, wherein the standard bias current value sequence includes multiple standard bias current values; Calculate the mean temperature and mean current values ​​of the standard temperature series and the standard bias current series respectively, and calculate the temperature-current correlation coefficient based on the mean temperature and mean current values.

[0033] It should be explained that the aligned temperature sequence refers to the laser temperature sequence aligned with the laser bias current value sequence in time. The aligned bias current value sequence refers to the laser bias current value sequence aligned with the laser temperature sequence in time. The purpose of alignment is to ensure that the timestamps of the two sequences are consistent, so as to facilitate the subsequent calculation of the temperature-current correlation coefficient. Data missing detection refers to checking whether there are missing data points in the aligned temperature sequence. The standard temperature sequence refers to the aligned temperature sequence after data missing data imputation. Missing data imputation refers to estimating and filling in the missing data points in the aligned temperature sequence using imputation methods (such as linear interpolation, polynomial interpolation, etc.). The standard temperature refers to the temperature value in the standard temperature sequence. The standard bias current value sequence refers to the laser bias current value sequence after processing. The method for obtaining the standard bias current value sequence based on the laser bias current value sequence is the same as the method for obtaining the standard temperature sequence based on the laser temperature sequence, and will not be repeated here. The standard bias current value refers to the bias current value in the standard bias current value sequence.

[0034] In detail, the formula for calculating the temperature-current correlation coefficient is as follows: ; in, This represents the temperature-current correlation coefficient. This indicates the number of standard temperatures in a standard temperature series or the number of standard bias current values ​​in a standard bias current value series. Indicates the first A standard bias current value. Indicates the average current. Indicates the first A standard temperature, This represents the average temperature.

[0035] It should be explained that the average temperature value refers to the average value of the standard temperature series. The average current value refers to the average value of the standard bias current value series. In this invention, the standard temperature series and the standard bias current value series are time-aligned; therefore, the number of values ​​corresponding to the standard temperature series and the standard bias current value series is the same. Indicates the first Subtract the average current from the standard bias current value at each time point. If the result is positive, it means that the standard bias current value at that time point is higher than the average level. If the result is negative, it means that the standard bias current value is lower than the average level. and Similarly, this will not be elaborated upon here. This represents the sum of the products of the deviations at all points in time. If most of the time the two change in the same direction (both high or both low), the numerator will be a large positive number; if most of the time the two change in opposite directions (one high and one low), the numerator will be a large negative number. If the changes are irregular, the positive and negative values ​​will cancel each other out, and the numerator will be close to zero. This represents the product of standard deviations, which scales the molecular results to a standardized range (-1, 1), making the calculated coefficients independent of the specific units or values ​​of current and temperature, allowing correlation coefficients between different modules to be compared.

[0036] Specifically, the step of calculating the received optical power sample entropy based on the received optical power value sequence includes: Calculate the standard deviation of received optical power in the received optical power value sequence, and calculate the similarity tolerance threshold based on the standard deviation of received optical power and the preset similarity tolerance coefficient. The similarity tolerance threshold is the product of the standard deviation of received optical power and the similarity tolerance coefficient. The received optical power value sequence is segmented using a preset mode length to obtain a segmented optical power set, wherein the segmented optical power set includes multiple segmented optical power groups, and each segmented optical power group includes multiple segmented optical powers. Segmented optical power groups are extracted sequentially from the segmented optical power group set, and the extracted segmented optical power groups are removed from the segmented optical power group set to obtain the remaining segmented optical power group set, wherein the remaining segmented optical power group set includes multiple remaining segmented optical power groups. The number of similar combination pairs is obtained based on the segmented optical power group and the remaining segmented optical power group set. The number of similar combination pairs is summarized to obtain the number of multiple similar combination pairs corresponding to the segmented optical power group set. Calculate the number of segmented optical power groups in the segmented optical power group set, calculate the number of comprehensive similar combination pairs based on the number of multiple similar combination pairs, and calculate the average ratio based on the number of segmented optical power groups and the number of comprehensive similar combination pairs. Perform a length increment operation on the mode length to obtain the updated mode length. Calculate the updated average ratio based on the updated mode length and the received optical power value sequence. Calculate the received optical power sample entropy based on the average ratio and the updated average ratio.

[0037] It should be explained that the standard deviation of received optical power is a statistic that measures the degree of fluctuation in the received optical power value sequence. In sample entropy calculation, the standard deviation of received optical power is used to determine the threshold for similarity judgment. Pattern length refers to the dimension by which the received optical power value sequence is divided into groups in sample entropy calculation. For example, the pattern length is 2. For instance, if the received optical power value sequence is {a, b, c, d, e, f}, and the pattern length is 2, after segmenting the received optical power value sequence using a pattern length of 2, the resulting set of segmented optical power groups is {(a, b), (b, c), (c, d), (d, e), (e, f)}. Segmented optical power refers to the continuous optical power value segments extracted from the received optical power value sequence after the pattern length is determined. The remaining set of segmented optical power groups refers to the set of segmented optical power groups remaining after extracting one segmented optical power group from the set of segmented optical power groups. This set is used to combine with the extracted segmented optical power groups to calculate similarity. The remaining segmented optical power group refers to a segmented optical power group within the remaining segmented optical power group set. This invention calculates a similarity tolerance threshold based on the standard deviation of the received optical power and a preset similarity tolerance coefficient, achieving adaptive standardization of the similarity tolerance threshold. It exhibits good robustness and universality for optical power signals of different intensities and fluctuation characteristics, avoiding misjudgments caused by different baseline levels of optical power signals.

[0038] Importantly, the number of segmented optical power groups refers to the total number of segmented optical power groups within the segmented optical power group set. The number of comprehensive similarity pairs refers to the sum of the numbers of multiple similarity pairs. The average proportion calculated based on the number of segmented optical power groups and the number of comprehensive similarity pairs is calculated as: average proportion = number of comprehensive similarity pairs / number of segmented optical power groups. The operation of increasing the mode length refers to increasing the mode length. For example, if the mode length is 2, then adding 1 to the mode length results in an updated mode length of 3. The updated mode length refers to the new mode length after increasing the mode length during the calculation of the received optical power sample entropy. The updated average proportion is the ratio of the number of similarity pairs recalculated under the updated mode length to the number of segmented optical power groups. The method for calculating the updated average proportion based on the updated mode length and the received optical power value sequence is the same as the method for calculating the average proportion based on the mode length and the received optical power value sequence, and will not be repeated here. The number of comprehensive similarity pairs directly measures the regularity and repetition within the received optical power value sequence. The stronger the regularity, the greater the number of comprehensive similarity pairs. The purpose of performing the length increase operation on the pattern length is to capture the ability of pattern persistence. If the received optical power value sequence is highly ordered, the update average ratio will be higher than the average ratio; if the received optical power value sequence is disordered, the update average ratio will be lower than the average ratio. The received optical power sample entropy of this invention directly reflects the unpredictability and disorder of the received optical power value sequence. If the received optical power sample entropy is higher than the preset normal sample entropy, it indicates that the received optical power value sequence is disordered, unstable, and highly random, which is a sign of early failure. Normal sample entropy refers to a value that is preset by the user.

[0039] Furthermore, the formula for calculating the sample entropy in the step of calculating the received optical power sample entropy based on the average ratio and the updated average ratio is as follows:

[0040] in, This represents the sample entropy of the received optical power. Represents the logarithmic function with base 10. Indicates average proportion. This indicates the updated average ratio.

[0041] Specifically, obtaining the number of similar combination pairs based on the segmented optical power groups and the remaining segmented optical power group sets includes: The segmented optical power group is combined with each remaining segmented optical power group to obtain a set of combination pairs. The set of combination pairs includes multiple combination pairs, and each combination pair includes one segmented optical power group and one remaining segmented optical power group. Extract combination pairs sequentially from the combination pair set, calculate multiple absolute differences between combination pairs, and determine whether multiple absolute differences are all less than the similarity tolerance threshold; If multiple absolute differences are all less than the similarity tolerance threshold, then the combined pair is regarded as a similar combined pair. Summarize similar pairs to obtain a set of similar pairs, and count the number of similar pairs in the set.

[0042] It should be explained that the multiple absolute differences of the calculated combination pair refer to the multiple absolute values ​​obtained by subtracting the optical power corresponding to the corresponding positions in the combination pair. For example, if the combination pair is {(a, b), (c, d)}, then |ac| and |bd| are used as the multiple absolute differences. A similar combination pair is a combination pair where, when calculating similarity, multiple absolute differences are all less than the similarity tolerance threshold. A similar combination pair set is the set of all similar combination pairs. The number of similar combination pairs refers to the number of similar combination pairs in the similar combination pair set. This invention, by judging whether any two waveforms in the received optical power value sequence are similar, microscopically detects whether there is a repetitive and stable pattern in the received optical power value sequence, transforming the inherent regularity of the received optical power value sequence into a quantifiable indicator.

[0043] Specifically, the calculation of the optical module aging rate based on the historical laser bias current value sequence includes: The historical laser bias current value sequence is grouped according to the historical detection frequency to obtain a bias current value set. The bias current value set includes multiple bias current value groups, and each bias current value group includes multiple historical laser bias current values. Obtain the median bias current value of each bias current value group in the bias current value group set to obtain the median bias current value set, and obtain the grouping sequence number set in the bias current value group set; The median bias current value set is indexed using the group sequence number set to create a bias current coordinate sequence. The bias current coordinate sequence includes multiple bias current coordinates, and the horizontal axis of the bias current coordinate is the group sequence number, while the vertical axis of the bias current coordinate is the bias current value. Linear regression is performed on the bias current coordinate sequence to obtain a fitted straight line. The slope of the fitted straight line is obtained, and the aging rate of the optical module is calculated based on the slope of the straight line.

[0044] It should be explained that the bias current value set refers to the collection of bias current value groups obtained by grouping the historical laser bias current value sequence according to the historical detection frequency. The historical laser bias current value refers to the current value required for the laser to operate normally at the historical detection frequency. The laser bias current value is one of the important parameters for evaluating the performance of optical modules; changes in the historical laser bias current value over time can reflect the aging of the optical module under test. The historical detection frequency refers to the frequency at which the laser bias current value was measured during past detection processes. For example, the historical detection frequency is once per hour. The median bias current value refers to the median of all bias current values ​​in a bias current value group, because the median is more resistant to the interference of instantaneous outliers than the average value. The group sequence number set refers to the set of unique group sequence numbers assigned to each bias current value group in the bias current value set. The group sequence number is a unique identifier for each bias current value group in the group sequence number set, used to identify the time sequence of each bias current value group. Time index creation refers to the operation of using the group sequence number as the time of the corresponding median bias current value. The time index is created to map the bias current value to time, facilitating subsequent linear regression. The linear regression of the bias current coordinate sequence refers to performing a linear regression using the least squares method. The fitted line is the straight line obtained through linear regression analysis, used to describe the trend of the bias current value over time. The slope of the line represents the rate of change of the bias current value over time. In the step of calculating the optical module aging rate based on the slope of the line, the optical module aging rate = slope of the line. 24 hours / day The number of days in the historical monitoring period.

[0045] Specifically, setting the dynamic early warning threshold includes: Obtain the historical risk score array of the optical module to be tested, and obtain the individual historical baseline value from the historical risk score array, where the individual historical baseline value is the largest historical risk score in the historical risk score array; Obtain the optical module type of the optical module to be tested, and retrieve a set of optical modules of the same type from the pre-built online testing system based on the optical module type. The set of optical modules of the same type includes one or more optical modules of the same type. Based on the detection period and the set of similar optical modules, a risk score set of similar optical modules is obtained, and the group baseline value is calculated based on the risk score set of similar optical modules. The maximum baseline value is obtained from the individual historical baseline value and the group baseline value. The dynamic early warning threshold is calculated based on the maximum baseline value and the preset safety factor.

[0046] It should be explained that the historical risk score set refers to the collection of comprehensive risk scores calculated for the optical module under test in various past testing periods. The historical risk score is the comprehensive risk score calculated in past testing periods, and the calculation process is the same as the method for obtaining the comprehensive risk score based on the early failure detection command of the optical module, the testing period, and the optical module under test, which will not be repeated here. Optical module type refers to the model or category of the optical module. For example, optical modules with different rates (1Gbps, 10Gbps, 25Gbps, etc.) and different transmission distances (short distance, long distance) belong to different types. The online testing system refers to a system that monitors and analyzes the performance of optical modules in real time, storing a large amount of historical and real-time data of optical modules. Same-type optical modules refer to other optical modules of the same type (model, specifications, etc.) as the optical module under test. The same-type optical module risk score set refers to the collection of comprehensive risk scores calculated for each same-type optical module in the same-type optical module set within a specific testing period. The same-type optical module risk score refers to the comprehensive risk score calculated for the same-type optical modules within a specific testing period.

[0047] Importantly, the step of calculating the group baseline value based on the risk score set of similar optical modules is as follows: The risk score set of similar optical modules is sorted in ascending order to obtain a sorted risk score set; the number of sorted risk scores in the sorted risk score set is obtained; a score index value is calculated based on the preset required percentile and the number of sorted risk scores; and the corresponding sorted risk score is identified from the sorted risk score set based on the score index value, which is the group baseline value. The sorted risk score set refers to the collection of all risk scores from the risk score set of similar optical modules arranged in ascending order. The purpose of sorting is to quickly locate the corresponding specific value for the required percentile. The number of sorted risk scores refers to the total number of elements in the sorted risk score set. The required percentile refers to the pre-set percentile that needs to be calculated in the sorted risk score set. For example, a required percentile of 90% means that 90% of the sorted risk scores in the sorted risk score set are less than or equal to this value. The score index value is the index value calculated based on the required percentile, used to find the corresponding sorted risk score from the sorted risk score set. In the step of calculating the score index value based on the preset required percentile and ranking risk score, the score index value = (required percentile / 100). The number of risk scores is ranked. The maximum baseline value refers to the maximum of the individual's historical baseline value and the group's baseline value. The safety factor is a pre-set coefficient used to adjust the dynamic warning threshold to ensure that the warning threshold is sufficiently conservative. The dynamic warning threshold is the value obtained by multiplying the dynamic warning threshold by the safety factor.

[0048] S4. If the comprehensive risk score is greater than the dynamic warning threshold, the comprehensive risk score is taken as an abnormal score, the adjacent detection period is taken as the detection period, and the step of obtaining the comprehensive risk score based on the early failure detection instruction of the optical module, the detection period and the optical module to be tested is returned until all the detection periods in the detection period set are extracted.

[0049] It should be explained that the anomaly score refers to the comprehensive risk score that exceeds the dynamic warning threshold. If the comprehensive risk score is greater than the dynamic warning threshold, it indicates that the optical module under test has shown clear early failure characteristics, its health status has significantly deviated from the normal track, and the system needs to pay higher attention to it.

[0050] S5. Summarize the abnormal scores to obtain an abnormal score set, calculate the number of abnormal scores in the abnormal score set, and if the number of abnormal scores is greater than the preset continuous abnormal threshold, trigger the pre-built early warning system to obtain an early warning signal, and formulate early warning measures based on the early warning signal.

[0051] It should be explained that the anomaly score set refers to the collection of anomaly scores corresponding to consecutive detection periods. The number of anomaly scores refers to the total number of anomaly scores in the anomaly score set, used to assess the overall anomaly status of the optical module under test over multiple detection periods. The continuous anomaly threshold is a pre-set threshold used to determine whether the number of anomaly scores reaches the standard for triggering the early warning system, so that early warning measures can be taken. The early warning system refers to a pre-built system used to issue an early warning signal when an anomaly is detected. The early warning signal is a signal issued by the early warning system indicating that the optical module is at risk of early failure. For example, the early warning signal is a text message, such as: "Optical module A1234 under test detected an early failure signal at 14:00 on September 26, 2026, with an anomaly score of 5. Immediate detailed diagnosis is recommended." Developing early warning measures based on the early warning signal means increasing the detection frequency of the optical module under test to more closely monitor its performance changes or to perform detailed technical diagnosis of the optical module under test, checking the performance of its internal components (such as lasers).

[0052] To address the problems described in the background art, this invention identifies the optical module to be tested and the set of testing time periods. It then sequentially extracts testing time periods from the set and performs the following operations on each extracted testing time period: This invention determines the set of testing time periods and sequentially extracts the testing time periods, rationally dividing the entire testing process according to time. This allows for phased and planned testing of the optical module, facilitating data analysis and comparison across different time periods and aiding in the orderly execution of subsequent steps. Based on the extracted testing time periods, adjacent testing time periods are identified from the set of testing time periods. These adjacent testing time periods are adjacent to the current testing time period but lag behind it. By identifying adjacent testing time periods, this invention enables [the invention to] [achieve] [a certain level of efficiency] in the time dimension. By establishing connections between different detection periods, and recognizing the potential continuity and correlation between data from adjacent periods, comparing detection results from adjacent periods can more acutely reveal trends in optical module performance, helping to detect potential early failure risks in advance. This invention receives early failure detection commands for optical modules, and based on these commands, the detection period, and the optical module under test, obtains a comprehensive risk score and sets dynamic warning thresholds. This invention quantifies the risk of early failure of optical modules by obtaining a comprehensive risk score, using a specific numerical value to represent the degree of risk of the optical module within a specific detection period. This makes the detection results more intuitive and objective, facilitating subsequent analysis and judgment. Setting dynamic warning thresholds allows for adjustments based on changes in performance. The detection period and the actual situation of the optical module are flexibly adjusted. If the comprehensive risk score is greater than the dynamic warning threshold, the comprehensive risk score is marked as an abnormal score, and the adjacent detection period is marked as the detection period. The process returns to the step of obtaining the comprehensive risk score based on the early failure detection command of the optical module, the detection period, and the optical module to be tested, until all detection periods in the detection period set have been extracted. When the comprehensive risk score is greater than the dynamic warning threshold, the present invention marks it as an abnormal score, which can promptly detect abnormal situations of the optical module in certain periods. By continuously cyclically detecting adjacent periods, the state changes of the optical module are continuously monitored to ensure that no possible early failure signs are missed. The abnormal scores are summarized to obtain the abnormal... The invention uses a constant score set and calculates the number of abnormal scores in the abnormal score set. If the number of abnormal scores exceeds a preset continuous abnormality threshold, a pre-built early warning system is triggered, generating an early warning signal. Based on this signal, early warning measures are formulated. This invention summarizes abnormal scores to obtain an abnormal score set and calculates the number of abnormal scores, allowing for a comprehensive assessment of the optical module's anomalies across multiple detection periods. An anomaly in a single period may be due to accidental factors, while anomalies across multiple consecutive periods better reflect the possibility of early failure of the optical module. Formulating early warning measures based on the early warning signal allows for intervention before the optical module actually fails, thereby reducing losses caused by optical module failure and ensuring the stable operation of the communication system. Therefore, this invention can improve the real-time performance and accuracy of optical modules.

[0053] like Figure 2 The diagram shown is a functional block diagram of an intelligent identification system for early failure of optical modules provided in an embodiment of the present invention.

[0054] The intelligent identification system 100 for early failure of optical modules described in this invention can be installed in electronic devices. Depending on the functions implemented, the intelligent identification system 100 for early failure of optical modules may include a detection period extraction module 101, a comprehensive risk score acquisition module 102, a cyclic detection module 103, and an early warning measure formulation module 104. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device. The detection period extraction module 101 is used to identify the optical module to be detected and the detection period set, extract the detection period from the detection period set in sequence, and perform the following operation on each extracted detection period: based on the extracted detection period, identify adjacent detection periods from the detection period set, wherein the adjacent detection period is adjacent to the detection period and lags behind the detection period; The comprehensive risk score acquisition module 102 is used to receive the early failure detection command of the optical module, acquire the comprehensive risk score based on the early failure detection command of the optical module, the detection period and the optical module to be detected, and set a dynamic early warning threshold. The loop detection module 103 is used to, if the comprehensive risk score is greater than the dynamic early warning threshold, take the comprehensive risk score as an abnormal score, take the adjacent detection period as the detection period, and return to the step of obtaining the comprehensive risk score based on the early failure detection instruction of the optical module, the detection period and the optical module to be tested, until all the detection periods in the detection period set have been extracted. The early warning measure formulation module 104 is used to summarize the abnormal scores, obtain an abnormal score set, calculate the number of abnormal scores in the abnormal score set, and if the number of abnormal scores is greater than the preset continuous abnormal threshold, trigger the pre-built early warning system, obtain an early warning signal, and formulate early warning measures based on the early warning signal.

[0055] In detail, the modules in the intelligent identification system 100 for early failure of optical modules described in this embodiment of the invention employ the same methods as described above during use. Figure 1 The method used here is the same as the intelligent identification method for early failure of optical modules described above, and can produce the same technical effect, so it will not be repeated here.

[0056] like Figure 3 The diagram shown is a structural schematic of an electronic device that implements an intelligent identification method for early failure of optical modules according to an embodiment of the present invention.

[0057] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and capable of running on the processor 10, such as a smart identification method program for early failure of optical modules.

[0058] The memory 11 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as a portable hard drive. In other embodiments, the memory 11 can be an external storage device of the electronic device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device 1. Furthermore, the memory 11 includes both internal storage units and external storage devices of the electronic device 1. The memory 11 can be used not only to store application software and various types of data installed on the electronic device 1, such as the code of a smart identification method program for early failure of optical modules, but also to temporarily store data that has been output or will be output.

[0059] In some embodiments, the processor 10 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control unit of the electronic device, connecting various components of the entire electronic device through various interfaces and lines. It executes programs or modules stored in the memory 11 (e.g., intelligent identification method program for early failure of optical modules) and calls data stored in the memory 11 to perform various functions of the electronic device 1 and process data.

[0060] The bus 12 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 can be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to realize the connection and communication between the memory 11 and at least one processor 10, etc.

[0061] Figure 3 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0062] For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) to power the various components. Preferably, the power supply can be logically connected to the at least one processor 10 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0063] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the electronic device 1 and other electronic devices.

[0064] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), and optionally, a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device 1 and to display a visual user interface.

[0065] The intelligent identification method program for early failure of the optical module stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When run in the processor 10, it can achieve the following: Once the optical module to be tested and the set of testing time periods are identified, the testing time periods are extracted sequentially from the set of testing time periods, and the following operations are performed on each extracted testing time period: Based on the extracted detection time periods, adjacent detection time periods are identified from the set of detection time periods. Among them, adjacent detection time periods are adjacent to the detection time periods and lag behind the detection time periods. Receive early failure detection command for optical module, obtain comprehensive risk score based on early failure detection command, detection period and optical module to be tested, and set dynamic early warning threshold; If the comprehensive risk score is greater than the dynamic warning threshold, the comprehensive risk score is taken as an abnormal score, the adjacent detection period is taken as the detection period, and the step of obtaining the comprehensive risk score based on the early failure detection instruction of the optical module, the detection period and the optical module to be tested is returned until all the detection periods in the detection period set are extracted. The abnormal scores are aggregated to obtain an abnormal score set. The number of abnormal scores in the abnormal score set is calculated. If the number of abnormal scores is greater than the preset continuous abnormal threshold, the pre-built early warning system is triggered to obtain an early warning signal. Early warning measures are then formulated based on the early warning signal.

[0066] Specifically, the processor 10's implementation method for the above instructions can be found in [reference needed]. Figures 1 to 3 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0067] Furthermore, if the modules / units integrated in the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0068] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following: Once the optical module to be tested and the set of testing time periods are identified, the testing time periods are extracted sequentially from the set of testing time periods, and the following operations are performed on each extracted testing time period: Based on the extracted detection time periods, adjacent detection time periods are identified from the set of detection time periods. Among them, adjacent detection time periods are adjacent to the detection time periods and lag behind the detection time periods. Receive early failure detection command for optical module, obtain comprehensive risk score based on early failure detection command, detection period and optical module to be tested, and set dynamic early warning threshold; If the comprehensive risk score is greater than the dynamic warning threshold, the comprehensive risk score is taken as an abnormal score, the adjacent detection period is taken as the detection period, and the step of obtaining the comprehensive risk score based on the early failure detection instruction of the optical module, the detection period and the optical module to be tested is returned until all the detection periods in the detection period set are extracted. The abnormal scores are aggregated to obtain an abnormal score set. The number of abnormal scores in the abnormal score set is calculated. If the number of abnormal scores is greater than the preset continuous abnormal threshold, the pre-built early warning system is triggered to obtain an early warning signal. Early warning measures are then formulated based on the early warning signal.

[0069] In the embodiments provided by this invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and actual implementations may have other classification methods.

[0070] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0071] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0072] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A smart identification method for early failure of optical modules, characterized in that, The method includes: Once the optical module to be tested and the set of testing time periods are identified, the testing time periods are extracted sequentially from the set of testing time periods, and the following operations are performed on each extracted testing time period: Based on the extracted detection time periods, adjacent detection time periods are identified from the set of detection time periods. Among them, adjacent detection time periods are adjacent to the detection time periods and lag behind the detection time periods. Receive early failure detection command for optical module, obtain comprehensive risk score based on early failure detection command, detection period and optical module to be tested, and set dynamic early warning threshold; If the comprehensive risk score is greater than the dynamic warning threshold, the comprehensive risk score is taken as an abnormal score, the adjacent detection period is taken as the detection period, and the step of obtaining the comprehensive risk score based on the early failure detection instruction of the optical module, the detection period and the optical module to be tested is returned until all the detection periods in the detection period set are extracted. The abnormal scores are aggregated to obtain an abnormal score set. The number of abnormal scores in the abnormal score set is calculated. If the number of abnormal scores is greater than the preset continuous abnormal threshold, the pre-built early warning system is triggered to obtain an early warning signal. Early warning measures are then formulated based on the early warning signal.

2. The intelligent identification method for early failure of optical modules as described in claim 1, characterized in that, The comprehensive risk score obtained based on the early failure detection command of the optical module, the detection period, and the optical module to be tested includes: Based on the early failure detection command and detection period of the optical module, the optical module to be detected is detected to obtain the real-time feature vector of the optical module. The real-time feature vector of the optical module is input into the pre-built optical module health model to obtain the deviation value, and the real-time feature vector of the optical module is input into the pre-built optical module fault migration model to obtain the fault probability. A comprehensive risk score is calculated based on preset fault probability weights, preset deviation weights, deviation values, and fault probabilities.

3. The intelligent identification method for early failure of optical modules as described in claim 2, characterized in that, The process involves detecting the optical module to be detected based on the early failure detection command and the detection period, obtaining the real-time feature vector of the optical module, including: According to the early failure detection command of the optical module, the optical module to be tested is detected during the detection period to obtain the received optical power value sequence, the laser bias current value sequence and the laser temperature sequence; The historical detection time period is obtained based on the detection time period, and the historical laser bias current value sequence is obtained based on the historical detection time period, the preset historical detection frequency and the optical module to be tested. Calculate the temperature-current correlation coefficient based on the laser temperature sequence and the laser bias current value sequence; The received optical power sample entropy is calculated based on the received optical power value sequence, and the optical module aging rate is calculated based on the historical laser bias current value sequence. The temperature-current correlation coefficient, the sample entropy of received optical power, and the aging rate of the optical module are used as the real-time feature vector of the optical module.

4. The intelligent identification method for early failure of optical modules as described in claim 3, characterized in that, The calculation of the temperature-current correlation coefficient based on the laser temperature sequence and the laser bias current value sequence includes: The laser temperature sequence and the laser bias current sequence are time-aligned to obtain the aligned temperature sequence and the aligned bias current sequence. Data missing detection is performed on the aligned temperature sequence to obtain the missing detection result, where the missing detection result indicates whether there is missing data or not. If the missing data detection result indicates that there is missing data, the missing data of the aligned temperature sequence is filled to obtain the standard temperature sequence; If the missing data detection result is that there is no missing data, then the aligned temperature sequence is used as the standard temperature sequence, wherein the standard temperature sequence includes multiple standard temperatures; A standard bias current value sequence is obtained based on the aligned bias current value sequence, wherein the standard bias current value sequence includes multiple standard bias current values; Calculate the mean temperature and mean current values ​​of the standard temperature series and the standard bias current series respectively, and calculate the temperature-current correlation coefficient based on the mean temperature and mean current values.

5. The intelligent identification method for early failure of optical modules as described in claim 4, characterized in that, The formula for calculating the temperature-current correlation coefficient is as follows: ； in, This represents the temperature-current correlation coefficient. This indicates the number of standard temperatures in a standard temperature series or the number of standard bias current values ​​in a standard bias current value series. Indicates the first A standard bias current value. Indicates the average current. Indicates the first A standard temperature, This represents the average temperature.

6. The intelligent identification method for early failure of optical modules as described in claim 5, characterized in that, The step of calculating the received optical power sample entropy based on the received optical power value sequence includes: Calculate the standard deviation of received optical power in the received optical power value sequence, and calculate the similarity tolerance threshold based on the standard deviation of received optical power and the preset similarity tolerance coefficient. The similarity tolerance threshold is the product of the standard deviation of received optical power and the similarity tolerance coefficient. The received optical power value sequence is segmented using a preset mode length to obtain a segmented optical power set, wherein the segmented optical power set includes multiple segmented optical power groups, and each segmented optical power group includes multiple segmented optical powers. Segmented optical power groups are extracted sequentially from the segmented optical power group set, and the extracted segmented optical power groups are removed from the segmented optical power group set to obtain the remaining segmented optical power group set, wherein the remaining segmented optical power group set includes multiple remaining segmented optical power groups. The number of similar combination pairs is obtained based on the segmented optical power group and the remaining segmented optical power group set. The number of similar combination pairs is summarized to obtain the number of multiple similar combination pairs corresponding to the segmented optical power group set. Calculate the number of segmented optical power groups in the segmented optical power group set, calculate the number of comprehensive similar combination pairs based on the number of multiple similar combination pairs, and calculate the average ratio based on the number of segmented optical power groups and the number of comprehensive similar combination pairs. Perform a length increment operation on the mode length to obtain the updated mode length. Calculate the updated average ratio based on the updated mode length and the received optical power value sequence. Calculate the received optical power sample entropy based on the average ratio and the updated average ratio.

7. The intelligent identification method for early failure of optical modules as described in claim 6, characterized in that, The method of obtaining the number of similar combination pairs based on the segmented optical power group and the remaining segmented optical power group set includes: The segmented optical power group is combined with each remaining segmented optical power group to obtain a set of combination pairs. The set of combination pairs includes multiple combination pairs, and each combination pair includes one segmented optical power group and one remaining segmented optical power group. Extract combination pairs sequentially from the combination pair set, calculate multiple absolute differences between combination pairs, and determine whether multiple absolute differences are all less than the similarity tolerance threshold; If multiple absolute differences are all less than the similarity tolerance threshold, then the combined pair is regarded as a similar combined pair. Summarize similar pairs to obtain a set of similar pairs, and count the number of similar pairs in the set.

8. The intelligent identification method for early failure of optical modules as described in claim 7, characterized in that, The calculation of the optical module aging rate based on the historical laser bias current value sequence includes: The historical laser bias current value sequence is grouped according to the historical detection frequency to obtain a bias current value set. The bias current value set includes multiple bias current value groups, and each bias current value group includes multiple historical laser bias current values. Obtain the median bias current value of each bias current value group in the bias current value group set to obtain the median bias current value set, and obtain the grouping sequence number set in the bias current value group set; The median bias current value set is indexed using the group sequence number set to create a bias current coordinate sequence. The bias current coordinate sequence includes multiple bias current coordinates, and the horizontal axis of the bias current coordinate is the group sequence number, while the vertical axis of the bias current coordinate is the bias current value. Linear regression is performed on the bias current coordinate sequence to obtain a fitted straight line. The slope of the fitted straight line is obtained, and the aging rate of the optical module is calculated based on the slope of the straight line.

9. The intelligent identification method for early failure of optical modules as described in claim 8, characterized in that, Setting the dynamic early warning threshold includes: Obtain the historical risk score array of the optical module to be tested, and obtain the individual historical baseline value from the historical risk score array, where the individual historical baseline value is the largest historical risk score in the historical risk score array; Obtain the optical module type of the optical module to be tested, and retrieve a set of optical modules of the same type from the pre-built online testing system based on the optical module type. The set of optical modules of the same type includes one or more optical modules of the same type. Based on the detection period and the set of similar optical modules, a risk score set of similar optical modules is obtained, and the group baseline value is calculated based on the risk score set of similar optical modules. The maximum baseline value is obtained from the individual historical baseline value and the group baseline value. The dynamic early warning threshold is calculated based on the maximum baseline value and the preset safety factor.

10. An intelligent identification system for early failure of optical modules, characterized in that, The system includes: The detection period extraction module is used to identify the optical module to be tested and the detection period set, extract the detection period from the detection period set in sequence, and perform the following operation on each extracted detection period: based on the extracted detection period, identify adjacent detection periods from the detection period set, wherein the adjacent detection period is adjacent to the detection period and lags behind the detection period; The comprehensive risk score acquisition module is used to receive early failure detection instructions for optical modules, acquire a comprehensive risk score based on the early failure detection instructions, the detection period, and the optical module to be tested, and set dynamic early warning thresholds. The loop detection module is used to treat the comprehensive risk score as an abnormal score and the adjacent detection period as a detection period if the comprehensive risk score is greater than the dynamic early warning threshold. It then returns to the step of obtaining the comprehensive risk score based on the early failure detection instruction of the optical module, the detection period and the optical module to be tested, until all the detection periods in the detection period set have been extracted. The early warning measure formulation module is used to summarize the abnormal scores, obtain the abnormal score set, calculate the number of abnormal scores in the abnormal score set, and if the number of abnormal scores is greater than the preset continuous abnormal threshold, the pre-built early warning system is triggered to obtain an early warning signal, and early warning measures are formulated based on the early warning signal.

Images