A surface-emitting laser transmitter operation monitoring method and system

Abstract

The present invention relates to the field of data processing technology, and in particular to a method and system for monitoring the operation of a surface-emitting laser transmitter. The method includes the following steps: collecting the through-current of the laser transmitter and the output current of the driving power supply at several moments; taking any moment as the target moment, and taking the target moment and the previous moment as the time window of the target moment, calculating the degree of abnormality of the through-current in the time window; obtaining the weight of each moment in the corresponding time window according to the degree of abnormality of the through-current in the time window, denoising the through-current in the time window according to the weight, and obtaining the denoised through-current at each moment; monitoring the operating status of the laser transmitter based on the denoised through-current at each moment. The present invention measures the degree of abnormality of the through-current in the time window by a mean shift algorithm, and denoises the through-current in different time windows to different degrees, thereby improving the accuracy of abnormality detection.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for monitoring the operation of a surface-emitting laser transmitter. Background Art

[0002] Surface-emitting laser emitters, with their unique structural advantages, demonstrate strong application potential in optical communications, smart terminals, and other fields. Their ease of array fabrication and high-density integration meet the optical communications industry's demand for high-capacity, miniaturized optical transceiver modules. They also provide a compact and efficient light source solution for smart terminal functions such as optical sensing and 3D imaging. They have already found widespread application in scenarios such as data center optical interconnects and mobile phone facial recognition modules.

[0003] However, ensuring the safe and long-term operation of surface-emitting laser transmitters presents challenges, making operational monitoring crucial. Pass current is a key electrical parameter reflecting the operating status of a laser transmitter, and its changes can directly reflect the device's operating status. Problems such as drive circuit failures and laser tube aging are often accompanied by abnormal fluctuations in pass current. However, in practical applications, pass current acquisition is susceptible to complex noise interference.

[0004] On the one hand, environmental noise has a significant impact, such as electromagnetic radiation around the equipment and power supply ripple, which will cause the collected current signal to be mixed with high-frequency interference; on the other hand, the working characteristics of the laser transmitter itself will also bring noise. For example, the heating of the laser tube causes resistance changes, causing small fluctuations in the current. If these noises are not properly handled, the current data will deviate from the true value.

[0005] When monitoring operations based on current data contaminated by noise, it is very easy to misjudge the equipment status, identify current fluctuations caused by noise as faults, or miss the real abnormality because the noise masks it. This will not only interfere with the normal use of the equipment, but frequent false alarms will also reduce operation and maintenance efficiency. Missed judgment may lead to the expansion of the fault, shorten the service life of the laser transmitter, and even cause safety hazards. Summary of the Invention

[0006] In order to solve the above-mentioned technical problem that operation monitoring based on noise-contaminated through-current data easily leads to misjudgment of the operating status of the laser transmitter, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a method for monitoring the operation of a surface-emitting laser transmitter; the method comprises the steps of:

[0008] Collect the current through the laser transmitter and the output current of the driving power supply at several moments; take any moment as the target moment, and use the target moment and the previous moment as the target moment. A time window with a moment as the target moment is used to calculate the abnormality of the current passing through the time window, including: calculating the data consistency of the current passing through the time window and the output current of the driving power supply, constructing a monitoring point at the target moment based on the data consistency, performing mean shift on the monitoring point at the target moment, obtaining the drift distance and its standard deviation in the iterative process of the mean shift of the monitoring point at the target moment, and obtaining the abnormality of the current passing through the time window according to all the drift distances and their standard deviations; calculating the denoised current passing through any data point in the time window, including: obtaining the weight of the corresponding data point in the time window by the normalized value of the product of the relative difference between the mean of any data point and all data points in the time window and the abnormality of the current passing through, denoising the weighted sum of the current passing through the corresponding data point in the time window by the weight, and obtaining the denoised current passing through the corresponding data point; calculating the denoised current passing through all data points in the time window one by one, and monitoring the operating status of the laser emitter based on the denoised current passing through all data points.

[0009] The present invention collects the current passing through and the output current of the driving power supply, and first calculates the abnormality of the current passing through the window in units of time windows. It then constructs a target moment in combination with data consistency, and obtains the drift distance and standard deviation through mean shift to quantify the abnormality. Then, a dynamic weight is obtained based on the abnormality, and the current is denoised to effectively suppress noise interference and improve data quality. The present invention uses mean shift to participate in the measurement of the abnormality of the current passing through the time window, which can better capture abnormal data, thereby filtering the current passing through different time windows with different degrees of smoothing, achieving smoothing of noise while avoiding excessive smoothing of non-noise data, thereby improving the accuracy of abnormal alarms.

[0010] Preferably, the abnormality degree of the current passing through the time window satisfies the relationship: ;in, For the The abnormal degree of current passing through each monitoring point in the corresponding time window, For the The maximum drift distance during the mean drift of each monitoring point, For the The minimum drift distance during the mean drift of each monitoring point, For the The standard deviation of all drift distances generated during the mean drift of the monitoring points.

[0011] This method accurately quantifies the degree of abnormal deviation of current flowing through a time window by constructing an anomaly calculation model that integrates the maximum drift distance, minimum drift distance, and drift distance standard deviation. By leveraging the dynamic characteristics of the mean-shift iterative process, it breaks through the limitations of traditional static threshold judgments and can keenly capture subtle fluctuation anomalies. This provides a more practical basis for subsequent denoising weight allocation, improves the accuracy and sensitivity of anomaly identification, and fundamentally enhances the reliability of operational monitoring.

[0012] Preferably, the weights of corresponding data points in the time window satisfy the relationship: ;in, For the The monitoring point corresponds to the The weight of the data point, It is The abnormal degree of current passing through each monitoring point in the corresponding time window, For the The monitoring point corresponds to the A current flows through For the The average value of the current passing through the monitoring point in the time window, For the The maximum absolute difference between each current passing through the monitoring point and the mean current passing through the time window is, It is an absolute value.

[0013] This invention implements a weighting formula based on the degree of abnormality, using the difference between the current value and the mean and the maximum difference as adjustment factors to achieve dynamic weight allocation. The more significant the abnormality and the further the current deviates from the mean, the lower the weight, effectively suppressing noise interference and highlighting the true current characteristics. This denoising process accurately filters reliable data, optimizes signal quality, and provides purer and more representative current data support for operating status monitoring, helping to improve the accuracy of status judgment.

[0014] Preferably, calculating the data consistency between the passing current and the output current of the driving power supply in the time window includes: recording the sequence composed of the output current data of the driving power supply in the time window as the output current sequence; recording the sequence composed of the passing current data in the time window as the passing current sequence; and calculating the data consistency between the passing current and the output current of the driving power supply in the time window based on the similarity between the passing current sequence and the output current sequence.

[0015] Preferably, the data consistency satisfies the relationship: ;in, For the The data consistency of the current passing through at each moment and the output current of the driving power supply is For the The output current sequence at each moment, For the The current sequence at each moment, For the The residual sequence between the current sequence at each moment and the fitting curve of the current sequence is For the The residual sequence between the output current sequence at each moment and the output current sequence fitting curve, is the similarity calculation function.

[0016] This method constructs a consistency model that fuses the output current sequence, the through current sequence, and the residual sequence of the two, and uses a similarity function to quantify the data matching degree from multiple dimensions. This method breaks through the limitations of single sequence comparison, takes into account both the overall data trend and local residual characteristics, and accurately captures subtle correlation differences between currents. This provides a more comprehensive and reliable basis for anomaly degree calculation, enhances the depth and precision of data feature extraction in operation monitoring, and improves the accuracy of state judgment.

[0017] Preferably, obtaining the residual sequence between the passing current sequence and the passing current sequence fitting curve includes: performing curve fitting on the passing current sequence to obtain the passing current sequence fitting curve; obtaining the residual sequence between the passing current sequence and the passing current sequence fitting curve; obtaining the residual sequence between the output current sequence and the output current sequence fitting curve includes: performing curve fitting on the output current sequence to obtain the output current sequence fitting curve, and obtaining the residual sequence between the output current sequence and the output current sequence fitting curve.

[0018] Preferably, the similarity calculation function is a Pearson correlation coefficient calculation function or a cosine similarity function.

[0019] Preferably, the monitoring point at the target moment is constructed based on data consistency, including: collecting the ambient temperature of the laser emitter; calculating the average value of the ambient temperature within the time window of the target moment, and recording it as the actual temperature at the target moment; standardizing the passing current and temperature values ​​at the target moment to obtain the normalized value of the passing current and the normalized value of the actual temperature; combining the normalized value of the passing current, data consistency, and the normalized value of the actual temperature into a monitoring point, and recording it as the monitoring point at the target moment.

[0020] The present invention introduces ambient temperature to construct the target moment, combines the normalized value of the current, data consistency, the normalized value of the actual temperature and other characteristics, breaks through the limitations of a single electrical parameter, adapts to the influence of ambient temperature on laser emission, enriches data dimensions, and provides a more comprehensive and actual working condition-fitted basis for subsequent abnormality analysis, thereby improving the accuracy of status monitoring.

[0021] Preferably, the current passing through any data point in the time window after denoising satisfies the relationship: ;in, For the The monitoring point corresponds to the The passing current after denoising, For the The monitoring point corresponds to the The weight of the current passing through, For the The monitoring point corresponds to the A current flows through For the The average value of the current passing through each monitoring point in the time window.

[0022] In a second aspect, the present invention provides a surface-emitting laser emitter operation monitoring system, which includes a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a surface-emitting laser emitter operation monitoring method according to the first aspect of the present invention is implemented.

[0023] By adopting the above technical solution, a surface-emitting laser transmitter operation monitoring method of the first aspect of the present invention is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby making a terminal device based on the memory and the processor for easy use.

[0024] Beneficial effects of the present invention: The present invention measures the degree of abnormality of the current passing through the time window through the participation of mean shift, which can better capture abnormal data, thereby performing filtering with different degrees of smoothing on the current passing through different time windows, achieving smoothing of noise while avoiding excessive smoothing of non-noise data, thereby improving the accuracy of abnormal alarms. At the same time, by constructing an abnormality degree calculation model that integrates the maximum drift distance, the minimum drift distance and the standard deviation of the drift distance, the abnormal deviation of the current passing through the time window can be accurately quantified. By utilizing the dynamic characteristics of the mean shift iterative process, breaking through the limitations of traditional static threshold judgment, it can keenly capture subtle fluctuation anomalies, provide a more practical basis for the subsequent denoising weight allocation, improve the accuracy and sensitivity of abnormality identification, and enhance the reliability of operation monitoring from the root. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A flow chart of a method for monitoring the operation of a surface-emitting laser transmitter provided by an embodiment of the present invention;

[0026] Figure 2 A schematic diagram of the complete mean shift process for any monitoring point provided by an embodiment of the present invention;

[0027] Figure 3This is a structural block diagram of a surface-emitting laser transmitter operation monitoring system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] A first aspect of an embodiment of the present invention provides a method for monitoring the operation of a surface-emitting laser transmitter, such as Figure 1 As shown, the method includes steps S100 to S400:

[0029] Step S100: collecting the current passing through the laser emitter and the output current of the driving power supply at several moments.

[0030] It should be noted that the current flowing through a laser emitter varies based on the output current of the laser emitter driver power supply. At the same time, the optical power of a laser emitter varies with the current flowing through the laser emitter. There is a certain correlation between the current flowing through the laser emitter, the output current of the laser emitter driver power supply, and the optical power of the laser emitter. Furthermore, the characteristics of the light-emitting element in a laser emitter indicate that even if the current flowing through the laser emitter remains the same, the optical power will vary when the laser emitter is at different temperatures. That is, for the same current flowing through the laser emitter, the higher the temperature, the lower the optical power. Therefore, the present invention collects the output current of the laser emitter driver power supply and the current flowing through the laser emitter.

[0031] Specifically, the current flowing through the laser emitter can be collected using sensors, such as closed-loop Hall sensors or current transformers, to collect the current at several moments. The output current of the laser emitter's driver power supply can be directly measured using a power meter or power supply tester.

[0032] So far, the output current of the driving power supply and the current passing through the laser emitter at several moments have been obtained.

[0033] Step S200: take any moment as the target moment, and compare the target moment with the previous moment. The time window of each moment is the target moment, and the abnormality degree of the current passing through the time window is calculated.

[0034] The calculation of the abnormality degree of the current passing through the time window includes steps S210 to S230:

[0035] Step S210: Calculate the data consistency between the current passing through the time window and the output current of the driving power supply.

[0036] It should be noted that since the through current of the laser emitter is positively correlated with the output current of the laser emitter driving power supply, when the output current of the driving power supply changes, the through current of the laser emitter should also show a consistent change trend. This characteristic can be used to identify the noise data of the through current. Therefore, the present invention obtains the data consistency of the laser emitter based on the through current of the laser emitter and the output current of the laser emitter driving power supply.

[0037] Specifically, the time window of the target time has been obtained in the above steps, and the time window includes the target time and the time before the target time. The data at the moment of time is recorded as the output current sequence of the driving power supply in the time window; the sequence of the through current data in the time window is recorded as the through current sequence; the data consistency of the through current in the time window and the output current of the driving power supply is calculated based on the similarity between the through current sequence and the output current sequence. The data consistency satisfies the relationship:

[0038] ;

[0039] in, For the The data consistency of the current passing through at each moment and the output current of the driving power supply is For the The output current sequence at each moment, For the The current sequence at each moment, For the The residual sequence between the current sequence at each moment and the fitting curve of the current sequence is For the The residual sequence between the output current sequence at each moment and the output current sequence fitting curve, is a similarity calculation function. The similarity calculation function is a Pearson correlation coefficient calculation function or a cosine similarity function, both of which are prior arts and will not be described in detail here.

[0040] In this formula, Representative The output current sequence at the moment The correlation between the current sequences at each moment, The larger the value, the more likely the output current sequence and the through current sequence are to have a consistent change trend, and the greater the data consistency of the laser transmitter; The smaller it is, the more likely that the output current sequence and the passing current sequence have different change trends, and the smaller the data consistency of the laser transmitter. Represents the The correlation between the residual sequence corresponding to the output current sequence at each moment and the residual sequence corresponding to the through current sequence. When the data changes of the output current sequence and the through current sequence are similar, the fitting curve of the output current sequence and the fitting curve of the through current sequence should also be similar, and the residual sequence corresponding to the output current sequence and the residual sequence corresponding to the through current sequence should also be similar; therefore The larger it is, the greater the data consistency of the laser transmitter; The smaller it is, the less consistent the data from the laser transmitter is.

[0041] It should be noted that the steps for obtaining the residual sequence between the passing current sequence and the passing current sequence fitting curve in the data consistency formula and the steps for obtaining the residual sequence between the output current sequence and the output current sequence fitting curve are the same, and the acquisition of the residual sequence between the output current sequence and the output current sequence fitting curve is taken as an example.

[0042] Specifically, the output current sequence is curve fitted by the least squares method to obtain the output current sequence fitting curve, and the difference between each value in the output current sequence and its corresponding value in the output current sequence fitting curve is obtained. This value is the residual, and the sequence composed of the residuals corresponding to all values ​​in the output current sequence is the residual sequence.

[0043] Step S220: construct monitoring points at the target time based on data consistency, and perform mean shift on the monitoring points at the target time.

[0044] It should be noted that, since the optical power of the laser emitter is different at different ambient temperatures under the same pass current, the pass current required to obtain the same optical power is different at different ambient temperatures. It is impossible to simply judge whether the pass current is noise data based on the size of the pass current. Therefore, the present invention obtains the degree of abnormality of the pass current in each time period based on the data consistency of the laser emitter, the pass current and the laser emitter temperature.

[0045] Specifically, the average temperature of the environment in which the laser emitter is located within the time range of the time window corresponding to the target moment is collected, and the average temperature of the environment is used as the temperature average at each monitoring moment, i.e., the target moment, and recorded as the actual temperature at the target moment. The through current at the target moment is divided by the theoretical maximum through current of the laser emitter, and the through current is normalized to obtain the normalized value of the through current; the temperature average at the target moment is divided by the theoretical maximum temperature of the environment in which the laser emitter is located to normalize the temperature average and obtain the normalized value of the actual temperature. The normalized value of the through current, data consistency, and normalized value of the actual temperature are combined into a monitoring point, which is recorded as the monitoring point at the target moment. That is, the monitoring point at the target moment is in the form of the normalized value of the through current, data consistency, and normalized value of the temperature average.

[0046] Step S230 : obtaining the drift distance and the standard deviation of the mean drift of the monitoring point at the target time during the iteration process, and obtaining the abnormality degree of the current passing through the time window according to all the drift distances and the standard deviations.

[0047] The above operations obtain the monitoring points at the target moment. By performing the above operations at each moment in the same manner, the corresponding monitoring points for each historical moment under different ambient temperatures of the laser transmitter can be obtained. The set of monitoring points corresponding to each historical moment is used as the monitoring data set. Mean shift is performed on the monitoring points at each monitoring moment to obtain the drift distance of each monitoring point during the mean shift iteration. The standard deviation of each drift distance during the mean shift process for each monitoring point is obtained. The degree of abnormality of the current passing through each time period is determined based on the drift distance and the standard deviation of the drift distance. The termination condition for mean shift can be set as follows: the length of the offset mean is less than a preset threshold. In this example, the preset threshold is 0.01, and it is set according to actual needs.

[0048] like Figure 2 As shown in the figure, it is the complete mean shift process of a monitoring point, where the black dot is the real-time monitoring point, the gray dot with dotted outline is the end point of the mean shift, the white dot with dotted outline is the end point of each drift, and the end point of the last drift is also the end point of the mean shift. The dotted line is the drift distance of each drift.

[0049] The previous section describes the construction of monitoring points and the process of mean shift. The following section describes how to determine the degree of current anomaly in each time period based on the drift distance and its standard deviation.

[0050] Specifically, the abnormality of the current passing through the target time within the time window satisfies the relationship:

[0051] ;

[0052] in, For the The abnormal degree of current passing through each monitoring point in the corresponding time window, For the The maximum drift distance during the mean drift of each monitoring point, For the The minimum drift distance during the mean drift of each monitoring point, For the The standard deviation of all drift distances generated during the mean drift of each monitoring point.

[0053] Since mean shift causes data points to drift toward the data-dense areas of the data space, and noise data is sporadic, the currents corresponding to the vast majority of monitoring points in the data space are not noise data, and thus the vast majority of monitoring points will form a dense area in the data space. When mean shift is performed on the monitoring points, the monitoring points will drift toward the data-dense areas of the data space, thus forming a dense area. The closer the monitoring point is to the dense area, the smaller the difference between the drift distances generated when the mean shift is performed. The farther the monitoring point is from the dense area, the greater the difference between the drift distances generated when the mean shift is performed. The closer the monitoring point is to the dense area, the more likely it is that there will be no noise data in the current corresponding to the monitoring point, and the smaller the degree of abnormality of the current in the time window corresponding to the monitoring point. The farther the monitoring point is from the dense area, the more likely it is that there will be noise data in the current corresponding to the monitoring point, and the greater the degree of abnormality of the current in the time window corresponding to the monitoring point.

[0054] therefore, The larger the value is, the greater the difference between the drift distances generated during the mean shift, and the greater the abnormality of the current passing through the corresponding time window of the monitoring point. The smaller it is, the smaller the difference between the drift distances generated during the mean shift is, and the smaller the abnormality of the current passing through the monitoring point corresponding to the time window is. The larger the value is, the greater the difference between the drift distances generated during the mean shift, and the greater the abnormality of the current passing through the corresponding time window of the monitoring point. The smaller it is, the smaller the difference between the drift distances generated during the mean shift is, and the smaller the abnormality of the current passing through the monitoring point corresponding to the time window is.

[0055] Divide by It is for as well as Normalize to facilitate subsequent calculations.

[0056] At this point, the abnormality level of the current passing through the time window is obtained.

[0057] Step S300: Calculate the denoised current of any data point within the time window.

[0058] It should be noted that when the degree of abnormality of the current passing through the time window is different, the possibility of noise data existing in the time period corresponding to the window is also different. When the degree of abnormality of the current passing through the time period is greater, the corresponding current data passing through the time period is more likely to contain noise data, and a larger filtering strength should be used to smooth the noise; when the degree of abnormality of the current passing through the time period is smaller, the corresponding current data passing through the time period is more likely to not contain noise data, and a smaller filtering strength should be used to avoid excessive smoothing, which results in the inability to use over-smoothed data for laser transmitter operation monitoring. Therefore, the present invention obtains the weight of each current passing through the time period based on the degree of abnormality of the current passing through the time period, and denoises the current passing through the time period using the filtering weight.

[0059] Specifically, the through current sequence corresponding to each monitoring point is obtained, the mean value of all through currents in the through current sequence is obtained, and the difference between each through current and the through current mean value is obtained. The weight of the through current is obtained according to the difference between each through current and the through current mean value and the abnormality of the through current in the time window corresponding to the monitoring point. The weight of the current passing through satisfies the relationship:

[0060] ;

[0061] in, For the The monitoring point corresponds to the The weight of the current passing through, It is The abnormal degree of current passing through each monitoring point in the corresponding time window, For the The monitoring point corresponds to the A current flows through For the The average value of the current passing through the monitoring point in the time window, For the The maximum absolute difference between each current passing through the monitoring point and the mean current passing through the time window is, It is an absolute value.

[0062] In this formula The acquisition of The average current value of the time window corresponding to each monitoring point is passed through the The monitoring points correspond to all the currents passing through the time window, and the maximum value of the absolute difference with the average value of the time window is obtained.

[0063] In this formula, The larger the The monitoring point corresponds to the The more a passing current deviates from the overall level of the passing current within the time window, The smaller the number, the The monitoring point corresponds to the The closer the individual passing current is to the overall level of the passing current within the time window. The larger the The more likely the through-current in the time window corresponding to each monitoring point is to contain noise, the more the through-current deviates from the overall level of the through-current in the time window, the more likely it is noise data. In this case, the through-current that deviates from the overall level of the through-current should be given a smaller weight to smooth the noise data. The smaller the The more likely it is that the current passing through the time window corresponding to each monitoring point is free of noise, the more the current passing through the time window deviates from the overall level of the current passing through the time window, the more likely it is that it is not noise data. In this case, there is no need to smooth the data to avoid over-smoothing.

[0064] After obtaining the weights, the corresponding denoised current satisfies the relationship:

[0065] ;

[0066] in, For the The monitoring point corresponds to the The passing current after denoising, For the The monitoring point corresponds to the The weight of the current passing through, For the The monitoring point corresponds to the A current flows through For the The average value of the current passing through each monitoring point in the time window.

[0067] In this formula, The bigger it is, The more likely it is not noise data, the closer the filtered value of the current should be to , thus avoiding excessive smoothing of the through current; The smaller the time, The more likely it is to be noise data, the closer the filtered value of the current should be to , in order to achieve filtering of the passing current.

[0068] Step S400: Calculate the denoised currents of all data points within the time window one by one, and monitor the operating status of the laser transmitter based on the denoised currents of all data points.

[0069] It should be noted that the operation of the laser transmitter can be monitored by the denoised through-current. If the through-current shows multiple abnormal changes within a continuous time period, it is determined that the laser transmitter has an abnormal operating state.

[0070] Specifically, if, for example, the current changes 10 or more times within one minute, and the amplitude of each change exceeds 20%, the laser transmitter is deemed to be operating abnormally. An alarm is issued to notify personnel to carry out repairs.

[0071] The second aspect of this embodiment provides a surface emitting laser transmitter operation monitoring system, such as Figure 3 As shown, the system includes a memory and a processor, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, a method for monitoring the operation of a surface-emitting laser transmitter according to the first aspect of the present invention is implemented.

[0072] The surface-emitting laser transmitter operation monitoring system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. Their configuration and functions are known in the art and will not be described in detail here.

[0073] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as resistive random access memory, dynamic random access memory, static random access memory, enhanced dynamic random access memory, high bandwidth memory, hybrid memory cube, etc., or any other medium that can be used to store the required information and can be accessed by an application, module, or both. Any such computer storage medium may be part of, accessible to, or connectable to the device.

[0074] The above are all preferred embodiments of the present invention, and are not intended to limit the scope of protection of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for monitoring the operation of a surface-emitting laser transmitter, characterized in that: Including steps: Collect the current through the laser transmitter and the output current of the driving power supply at several moments; Take any moment as the target moment, and take the target moment and its previous moment as the target moment. a time window with a moment as a target moment, and calculating the abnormality of the current passing through the time window, including: calculating the data consistency between the current passing through the time window and the output current of the driving power supply, constructing a monitoring point at the target moment based on the data consistency, performing a mean shift on the monitoring point at the target moment, obtaining a drift distance and a standard deviation thereof in an iterative process of the mean shift of the monitoring point at the target moment, and obtaining the abnormality of the current passing through the time window according to all the drift distances and their standard deviations; Calculating a denoised through-current for any data point in the time window, comprising: obtaining a weight of the corresponding data point in the time window by multiplying a normalized value of a relative difference between any data point and a mean of all data points in the time window by a degree of abnormality of the through-current; performing denoising on a weighted sum of the through-currents for the corresponding data points in the time window using the weights to obtain the denoised through-current for the corresponding data point; The denoised through-currents of all data points within the time window are calculated one by one, and the operating state of the laser transmitter is monitored based on the denoised through-currents of all data points.

2. The surface emitting laser transmitter operation monitoring method according to claim 1, characterized in that: The abnormal degree of the current passing through the time window satisfies the relationship: ; in, For the The abnormal degree of current passing through each monitoring point in the corresponding time window, For the The maximum drift distance during the mean drift of each monitoring point, For the The minimum drift distance during the mean drift of each monitoring point, For the The standard deviation of all drift distances generated during the mean drift of each monitoring point.

3. The surface emitting laser transmitter operation monitoring method according to claim 1, characterized in that: The weights of the corresponding data points in the time window satisfy the relationship: ; in, For the The monitoring point corresponds to the The weight of the data point, It is The abnormal degree of current passing through each monitoring point in the corresponding time window, For the The monitoring point corresponds to the A current flows through For the The average value of the current passing through the monitoring point in the time window, For the The maximum absolute difference between each current passing through the monitoring point and the mean current passing through the time window is, It is an absolute value.

4. The surface emitting laser transmitter operation monitoring method according to claim 1, characterized in that: The calculating of data consistency between the current passing through the time window and the output current of the driving power supply includes: The sequence formed by the output current data of the driving power supply within the time window is recorded as the output current sequence; The sequence of the passing current data within the time window is recorded as the passing current sequence; The data consistency between the through current and the output current of the driving power supply within the time window is calculated according to the similarity between the through current sequence and the output current sequence.

5. The surface emitting laser transmitter operation monitoring method according to claim 4, characterized in that: The data consistency satisfies the relationship: ; in, For the The data consistency of the current passing through at each moment and the output current of the driving power supply is For the The output current sequence at each moment is: For the The current sequence at each moment, For the The residual sequence between the current sequence at each moment and the fitting curve of the current sequence is For the The residual sequence between the output current sequence at each moment and the output current sequence fitting curve, is the similarity calculation function.

6. The surface emitting laser transmitter operation monitoring method according to claim 5, characterized in that: The acquisition of the residual sequence between the passing current sequence and the passing current sequence fitting curve includes: Performing curve fitting on the passing current sequence to obtain a fitting curve of the passing current sequence; obtaining a residual sequence between the passing current sequence and the passing current sequence fitting curve; Acquiring the residual sequence between the output current sequence and the output current sequence fitting curve includes: performing curve fitting on the output current sequence to obtain the output current sequence fitting curve, and acquiring the residual sequence between the output current sequence and the output current sequence fitting curve.

7. The surface emitting laser transmitter operation monitoring method according to claim 5, characterized in that: The similarity calculation function is a Pearson correlation coefficient calculation function or a cosine similarity function.

8. The surface emitting laser transmitter operation monitoring method according to claim 1, characterized in that: The monitoring point at the target time is constructed based on data consistency, including: collecting the ambient temperature of the laser transmitter; Calculate the average value of the ambient temperature within the time window of the target moment, and record it as the actual temperature at the target moment; The through current and temperature values ​​at the target moment are normalized to obtain the normalized value of the through current and the normalized value of the actual temperature; The normalized value of the passing current, the data consistency, and the normalized value of the actual temperature are combined into a monitoring point, which is recorded as the monitoring point at the target time.

9. The surface emitting laser transmitter operation monitoring method according to claim 1, characterized in that: The current after denoising at any data point in the time window satisfies the relationship: ; in, For the The monitoring point corresponds to the The passing current after denoising, For the The monitoring point corresponds to the The weight of the current passing through, For the The monitoring point corresponds to the A current flows through For the The average value of the current passing through each monitoring point in the time window.

10. A surface emitting laser transmitter operation monitoring system, characterized in that: The system includes a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, a method for monitoring the operation of a surface-emitting laser transmitter according to any one of claims 1 to 9 is implemented.

Images