Method for forming high-order azimuthal polarization modes

Abstract

The present application discloses a method for forming a high-order azimuthal polarization mode, which learns and depicts the timing coordination and dynamic correlation characteristics between the reflector angle and the temperature in the laser during training, and then in the actual application process, the temperature sensor is used to monitor and collect the real-time temperature data in the laser in real time when the laser is running, and a data processing and analysis algorithm based on artificial intelligence and deep learning is introduced at the back end to perform timing dynamic feature analysis and temperature compensation calculation based on timing attenuation information aggregation on the real-time temperature data in the laser, so as to automatically generate the angle data of the reflector at the next time point. In this way, the reflector angle can be adjusted adaptively based on the timing fluctuation and real-time change of the temperature, so as to automatically compensate for the temperature deviation effect. Through this intelligent laser parameter adjustment method, it is possible to ensure that a stable high-order azimuthal polarization mode can be maintained under different operating conditions, thereby ensuring the stability of the laser mode.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more specifically, to a method for forming a high-order azimuthal polarization mode. Background Art

[0002] High-order azimuthal polarization modes play an important role in laser technology, especially in applications that require high power and good beam quality. A high-order azimuthal polarization mode is a laser mode whose amplitude and phase distribution of the laser field has a specific ring structure, and the polarization direction changes along the direction of the ring. These modes have broad application potential in optical communications, laser processing, medical lasers and other fields due to their unique light field structure and polarization characteristics. Compared with the basic Gaussian mode, the high-order azimuthal polarization mode can effectively utilize a larger cross-section of the active medium, thereby achieving higher laser output power.

[0003] However, during the operation of the laser, the operational stability of the laser is affected by many factors, including environmental changes such as temperature, humidity, vibration, and aging and wear of the internal components of the laser. These factors may cause fluctuations in the laser mode, which in turn affects the stability of the laser output mode. For example, temperature fluctuations may cause the laser components to expand or contract, thereby affecting the laser mode, which limits the application of high-order azimuthal polarization modes in scenarios that require high power and good beam quality. Therefore, ensuring the stability of high-order azimuthal polarization modes is crucial to the performance of the laser system.

[0004] Therefore, an optimized scheme for forming high-order azimuthal polarization modes is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a method for forming a high-order azimuthal polarization mode, which learns and depicts the timing coordination and dynamic correlation characteristics between the reflector angle and the temperature in the laser during training, and then in the actual application process, the temperature sensor is used to monitor and collect the real-time temperature data in the laser in real time when the laser is running, and a data processing and analysis algorithm based on artificial intelligence and deep learning is introduced at the back end to perform timing dynamic feature analysis and temperature compensation calculation based on timing attenuation information aggregation on the real-time temperature data in the laser, so as to automatically generate the angle data of the reflector at the next time point. In this way, the reflector angle can be adjusted adaptively based on the timing fluctuation and real-time change of the temperature, so as to automatically compensate for the temperature deviation effect. Through this intelligent laser parameter adjustment method, it is possible to ensure that a stable high-order azimuthal polarization mode can be maintained under different operating conditions, thereby ensuring the stability of the laser mode and improving the application capability of the high-order azimuthal polarization mode in different scenarios.

[0006] According to one aspect of the present application, a method for forming a high-order azimuthal polarization mode is provided, which includes: obtaining a time queue of real-time temperature inside a laser collected by a temperature sensor; performing timing fluctuation feature extraction on the time queue of real-time temperature inside the laser to obtain a sequence of local timing correlation feature vectors of real-time temperature inside the laser; performing temperature compensation reasoning on each local timing correlation feature vector of real-time temperature inside the laser in the sequence of local timing correlation feature vectors of real-time temperature inside the laser to obtain a sequence of temperature compensated local timing decoding representation vectors; inputting the sequence of temperature compensated local timing decoding representation vectors into a feature aggregation network guided by node feature significance attenuation to obtain a temperature compensated timing attenuation aggregation representation vector; performing temperature compensation adaptive control based on the temperature compensated timing attenuation aggregation representation vector to determine a control instruction, wherein the control instruction is used to represent a decoded value of a recommended angle value of a reflector at the next time point.

[0007] Compared with the prior art, the present application provides a method for forming a high-order azimuthal polarization mode, which learns and depicts the timing coordination and dynamic correlation characteristics between the reflector angle and the temperature in the laser during training, and then in the actual application process, the temperature sensor is used to monitor and collect the real-time temperature data in the laser in real time when the laser is running, and a data processing and analysis algorithm based on artificial intelligence and deep learning is introduced at the back end to perform timing dynamic feature analysis and temperature compensation calculation based on timing attenuation information aggregation on the real-time temperature data in the laser, so as to automatically generate the angle data of the reflector at the next time point. In this way, the reflector angle can be adaptively adjusted based on the timing fluctuation and real-time change of the temperature, so as to automatically compensate for the temperature deviation effect. Through this intelligent laser parameter adjustment method, it is possible to ensure that a stable high-order azimuthal polarization mode can be maintained under different operating conditions, thereby ensuring the stability of the laser mode and improving the application capability of the high-order azimuthal polarization mode in different scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0009] Figure 1 4 is a flow chart of a method for forming a high-order azimuthal polarization mode according to an embodiment of the present application.

[0010] Figure 2 Schematic diagram of data flow of a method for forming a high-order azimuthal polarization mode according to an embodiment of the present application. DETAILED DESCRIPTION

[0011] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0012] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0013] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0014] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0015] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0016] A high-order azimuthal polarization mode is a laser mode whose amplitude and phase distribution of the laser field has a specific ring structure, and the polarization direction changes along the direction of the ring. Specifically, these modes can be expressed as A-TEMp1* (p=0,1,2,3.), where p represents the order of the mode, reflecting the number of rings. In laser technology, such modes usually have circular symmetry, which means that they form a series of concentric rings in the cross section. The lowest-order mode is a ring, and as the order p increases, the number of rings also increases. The spot diameter of this mode is usually larger than the spot diameter of the fundamental Gaussian mode (TEM00), which means that a larger active medium cross section can be used to increase the laser power.

[0017] Based on this, in the technical solution of the present application, a method for forming a high-order azimuthal polarization mode is proposed, which can realize intelligent adjustment of laser parameters at the control end of the laser system to ensure that a stable high-order azimuthal polarization mode can be maintained under different operating conditions. Specifically, temperature fluctuations may cause the laser components to expand or contract, thereby affecting the laser mode. For example, an increase in temperature will cause the refractive index of the gain medium to change, thereby producing a thermal lens effect. The thermal lens effect changes the propagation path of light in the gain medium, thereby changing the mode of the resonant cavity. Therefore, this solution learns the relationship between the historical angle data of the reflector and the temperature changes in the laser, thereby predicting the impact of temperature changes on the stability of the high-order azimuthal polarization mode through the time-series change characteristics and trends of the real-time temperature in the laser, and making corresponding adjustments in advance to ensure the stability of the high-order azimuthal polarization mode.

[0018] In the technical solution of the present application, a method for forming a high-order azimuthal polarization mode is proposed. Figure 1 4 is a flow chart of a method for forming a high-order azimuthal polarization mode according to an embodiment of the present application. Figure 2 FIG. 2 is a data flow diagram of a method for forming a high-order azimuthal polarization mode according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the method for forming a high-order azimuthal polarization mode of an embodiment of the present application, the steps include: S1, obtaining a time queue of the real-time temperature inside the laser collected by a temperature sensor; S2, performing timing fluctuation feature extraction on the time queue of the real-time temperature inside the laser to obtain a sequence of local timing correlation feature vectors of the real-time temperature inside the laser; S3, performing temperature compensation reasoning on each local timing correlation feature vector of the real-time temperature inside the laser in the sequence of local timing correlation feature vectors of the real-time temperature inside the laser to obtain a sequence of temperature compensated local timing decoding representation vectors; S4, inputting the sequence of temperature compensated local timing decoding representation vectors into a feature aggregation network guided by node feature significance attenuation to obtain a temperature compensated timing attenuation aggregation representation vector; S5, performing temperature compensation adaptive control based on the temperature compensated timing attenuation aggregation representation vector to determine a control instruction, wherein the control instruction is used to represent a decoded value of the recommended angle value of the reflector at the next time point.

[0019] In particular, the S1 and S2 obtain the time queue of the real-time temperature in the laser collected by the temperature sensor; and extract the time series fluctuation feature of the time queue of the real-time temperature in the laser to obtain a sequence of local time series correlation feature vectors of the real-time temperature in the laser. In a specific example of the present application, the time queue of the real-time temperature in the laser is input into the temperature time series fluctuation feature extractor based on the 1D-CNN model to obtain a sequence of local time series correlation feature vectors of the real-time temperature in the laser. That is, by inputting the time queue of the real-time temperature in the laser into the temperature time series fluctuation feature extractor based on the 1D-CNN model for feature mining, the local time series implicit feature information of the real-time temperature in the laser in the time dimension is extracted, thereby obtaining a sequence of local time series correlation feature vectors of the real-time temperature in the laser. It is worth mentioning that the one-dimensional convolutional neural network (1D-CNN) is a convolutional neural network specifically used to process one-dimensional time series data (such as sequences, signals, and text). It extracts time series features by applying convolution operations on the one-dimensional space of the input data.

[0020] In particular, the S3 performs temperature compensation reasoning on each local timing-related feature vector of the real-time temperature inside the laser in the sequence of local timing-related feature vectors of the real-time temperature inside the laser to obtain a sequence of temperature-compensated local timing decoding representation vectors. In a specific example of the present application, each local timing-related feature vector of the real-time temperature inside the laser in the sequence of local timing-related feature vectors of the real-time temperature inside the laser is input into a decoder-based temperature compensation reasoner to obtain a sequence of temperature-compensated local timing decoding representation vectors. It should be understood that the local timing implicit feature information of the real-time temperature inside the laser in the time dimension can be extracted by the temperature timing fluctuation feature extractor based on the 1D-CNN model, but this local timing feature only describes the temperature change over time, and does not directly indicate how the laser should be adjusted to compensate for these temperature fluctuations. Therefore, in order to infer how to compensate for the influence of temperature change on the laser mode from the extracted temperature local timing characteristics, so as to adaptively perform the subsequent reflector angle control, in the technical solution of the present application, each laser internal real-time temperature local timing associated feature vector in the sequence of the laser internal real-time temperature local timing associated feature vector is further input into the decoder-based temperature compensation reasoner to obtain a sequence of temperature compensation local timing decoding representation vectors. The decoder-based temperature compensation reasoner can process each laser internal real-time temperature local timing associated feature vector respectively, so as to use the laser internal real-time temperature local timing characteristics in each local time period to perform temperature compensation reasoning, thereby generating a sequence of temperature compensation local timing decoding representation vectors. It is worth mentioning that each temperature compensation local timing decoding representation vector here contains information on how to adjust laser parameters (such as the angle of the reflector) in the local timing to offset the influence of temperature changes and fluctuations.

[0021] In particular, in S4, the sequence of the temperature-compensated local timing decoding representation vectors is input into a feature aggregation network guided by the node feature significance attenuation to obtain a temperature-compensated timing attenuation aggregation representation vector. It should be understood that since each temperature-compensated local timing decoding representation vector in the sequence of the temperature-compensated local timing decoding representation vectors respectively contains information about the temperature compensation decoding timing in each local time period in order to ensure the stability of the laser mode. However, in the process of laser control, the temperature changes and fluctuations in different local time periods and the corresponding compensation measures may have different importance. Moreover, if you want to compensate for the temperature fluctuations at the next time point, the temperature compensation information that is closer to the timing distance of the next time point is more important, and the importance and attenuation of this timing need to be comprehensively considered. Therefore, in order to identify which local time periods have greater impact on the laser mode due to temperature changes and fluctuations, and taking into account the timing attenuation mechanism, so as to pay more attention to these important data information in the process of temperature compensation timing information aggregation to improve the accuracy of subsequent laser parameter adjustment, in the technical solution of the present application, the sequence of the temperature compensation local timing decoding representation vector is further input into a feature aggregation network guided by node feature significance attenuation to obtain a temperature compensation timing attenuation aggregation representation vector. The feature aggregation network guided by node feature significance attenuation can extract the most valuable node feature information from the temperature compensation local timing decoding representation of multiple time nodes, and generate a comprehensive and more significant temperature compensation timing attenuation aggregation representation vector. The temperature compensation timing attenuation aggregation representation vector contains optimized key information, which can be used to guide the angle adjustment of the laser reflector to achieve the best temperature compensation effect, thereby ensuring the stability of the high-order azimuthal polarization mode. This not only improves the accuracy of the control system, but also enhances the system's adaptability and robustness.

[0022] In an embodiment of the present application, the sequence of the temperature compensated local temporal decoding representation vectors is input into a feature aggregation network guided by node feature significance attenuation to obtain a temperature compensated temporal attenuation aggregation representation vector, including: first, calculating the feature significance description factor of each temperature compensated local temporal decoding representation vector in the sequence of the temperature compensated local temporal decoding representation vector, wherein the feature significance description factor is related to the mean and variance of each temperature compensated local temporal decoding representation vector; here, the mean reflects the central position of the eigenvalues ​​in the temperature compensated local temporal decoding representation vector, and the variance represents the degree of dispersion of these eigenvalues, and the two together determine the significance of a specific temperature compensated local temporal decoding representation feature. By calculating the feature significance descriptor factor, the model can identify the node features that are statistically representative and discriminative, namely, the temperature-compensated local timing decoding representation, which provides an important priority indicator for subsequent feature selection and aggregation; then, based on the distance span between each temperature-compensated local timing decoding representation vector in the sequence of the temperature-compensated local timing decoding representation vector and the current temperature-compensated local timing decoding representation vector, the feature significance attenuation factor of each temperature-compensated local timing decoding representation vector is constructed; here, the construction of the feature significance attenuation factor allows the model to adjust the influence of the temperature-compensated local timing features of the node. Specifically, the temperature-compensated local timing features exhibited by time nodes with closer distances are considered to be more relevant to the temperature offset compensation and laser stability control at the next time point, so they will be given higher weights during the aggregation process.In this way, the node features that are closer to the current node in the feature space can obtain a higher weight, thereby strengthening the temporal correlation of these local features; further, based on the feature significance attenuation factors of the temperature compensated local temporal decoding representation vectors, the feature significance description factors of the temperature compensated local temporal decoding representation vectors are modulated to obtain a sequence of feature significance attenuation description factors; in particular, the sequence of modulated feature significance attenuation description factors more accurately reflects the relative importance of each time node feature in the sequence of temperature compensated local temporal decoding representation vectors, wherein features that are more consistent or close to the current node features will obtain a greater weight, while other features will be weakened accordingly; this is helpful for the subsequent aggregate representation of temperature compensated temporal information and the control of laser parameters at the next time point; further, the sequence of feature significance attenuation description factors is input into the gated mask module to obtain the mask feature significance attenuation weight The method comprises the following steps: a first step is to determine the sequence of the temperature-compensated local time series decoding representation vectors, and a second step is to determine the sequence of the temperature-compensated local time series decoding representation vectors. The first step is to determine the sequence of the temperature-compensated local time series decoding representation vectors, and the second step is to determine the sequence of the temperature-compensated local time series decoding representation vectors. The second ...

[0023] Among them, the process of calculating the feature significance description factor of each temperature compensated local temporal decoding representation vector in the sequence of the temperature compensated local temporal decoding representation vector, wherein the feature significance description factor is related to the mean and variance of each temperature compensated local temporal decoding representation vector, includes: respectively calculating the mean and variance of the temperature compensated local temporal decoding representation vector to obtain the temperature compensated local temporal feature mean and the temperature compensated local temporal feature variance; after calculating the fourth power of the positional difference between the temperature compensated local temporal decoding representation vector and the temperature compensated local temporal feature mean, calculating the expected value of the fourth power modulated semantic deviation vector to obtain the temperature compensated local temporal feature expectation factor; calculating the division between the temperature compensated local temporal feature expectation factor and the square of the temperature compensated local temporal feature variance to obtain the feature significance description factor of the temperature compensated local temporal decoding representation vector.

[0024] More specifically, based on the distance span between each temperature compensated local timing decoding representation vector in the sequence of the temperature compensated local timing decoding representation vector and the current temperature compensated local timing decoding representation vector, the process of constructing the feature significance attenuation factor of each temperature compensated local timing decoding representation vector includes: calculating the distance span between other temperature compensated local timing decoding representation vectors in the sequence of the temperature compensated local timing decoding representation vector and the current temperature compensated local timing decoding representation vector to obtain the feature distance span; extracting the maximum value in the current temperature compensated local timing decoding representation vector to obtain the current temperature compensated local timing feature maximum value; extracting the maximum value in the other temperature compensated local timing decoding representation vectors to obtain other temperature compensated local timing feature maximum values; calculating the difference between the current temperature compensated local timing feature maximum value and the other temperature compensated local timing feature maximum values ​​and dividing the difference by the feature distance span to obtain the feature significance attenuation factor.

[0025] More specifically, the process of modulating the feature significance description factors of the respective temperature compensated local timing decoding representation vectors based on the feature significance attenuation factors of the respective temperature compensated local timing decoding representation vectors to obtain a sequence of feature significance attenuation description factors includes: calculating the positional product between the feature significance attenuation factors of the respective temperature compensated local timing decoding representation vectors and the feature significance description factors of the respective temperature compensated local timing decoding representation vectors to obtain a sequence of feature significance attenuation description factors.

[0026] More specifically, the process of inputting the sequence of feature significance attenuation descriptors into the gated mask module to obtain a sequence of masked feature significance attenuation weight factors includes: taking the negative number of each feature significance attenuation descriptor in the sequence of feature significance attenuation descriptors as an exponential power, calculating the exponential function value with the natural constant e as the base to obtain a sequence of feature significance attenuation class support descriptors; calculating the reciprocal of the sum of each feature significance attenuation class support descriptor in the sequence of feature significance attenuation class support descriptors and the constant one to obtain a sequence of feature significance attenuation weight factors; based on a predetermined threshold, masking the sequence of feature significance attenuation weight factors to obtain a sequence of masked feature significance attenuation weight factors, wherein, in response to each feature significance attenuation weight factor in the sequence of feature significance attenuation weight factors being greater than the predetermined threshold, the feature significance attenuation weight factor greater than the predetermined threshold is taken to have the original value, otherwise it is set to zero.

[0027] More specifically, the process of calculating the weighted sum of the sequence of temperature-compensated local temporal decoding representation vectors using the sequence of mask feature significance attenuation weight factors as the sequence of weights to obtain the temperature-compensated temporal attenuation aggregate representation vector includes: calculating the position-wise weighted sum between each other temperature-compensated local temporal decoding representation vector in the sequence of temperature-compensated local temporal decoding representation vectors using the sequence of mask feature significance attenuation weight factors as the sequence of weights to obtain the historical temperature-compensated local temporal decoding representation vector; and calculating the position-wise sum between the historical temperature-compensated local temporal decoding representation vector and the current temperature-compensated local temporal decoding representation vector to obtain the temperature-compensated temporal attenuation aggregate representation vector.

[0028] In summary, in the above embodiment, the sequence of the temperature-compensated local temporal decoding representation vectors is input into a feature aggregation network guided by node feature significance attenuation to obtain a temperature-compensated temporal attenuation aggregation representation vector, including: inputting the sequence of the temperature-compensated local temporal decoding representation vectors into the feature aggregation network guided by node feature significance attenuation to process the vectors with the following feature aggregation formula to obtain the temperature-compensated temporal attenuation aggregation representation vector; wherein the feature aggregation formula is: ;in, a sequence of representation vectors for the temperature compensated local timing decoding, The first in the sequence of the temperature compensated local timing decoding representation vector A temperature-compensated local timing decoding representation vector, and are respectively the first and A temperature-compensated local timing decoding representation vector, Indicates taking The maximum value in Indicates taking The minimum value in express and The distance span between For the said The feature significance attenuation factor corresponding to the temperature-compensated local temporal decoding representation vector, For the said The mean of the temperature-compensated local timing decoding representation vector, For the said The variance of the temperature-compensated local timing decoding representation vector, For the said The temperature-compensated local timing decoding represents the vector The eigenvalues ​​at the positions, represents the computational expectation, For the said The feature significance attenuation description factor corresponding to the temperature-compensated local temporal decoding representation vector, For the said The feature significance attenuation weight factor of the temperature-compensated local temporal decoding representation vector, For masking operation, is the predetermined threshold, For the said The mask feature significance attenuation weight factor of the temperature compensated local temporal decoding representation vector, The number of feature vectors in the sequence of the temperature compensated local timing decoding representation vector, A representation vector for the temperature compensation timing attenuation aggregation.

[0029] In particular, the S5 performs temperature compensation adaptive control based on the temperature compensation timing attenuation aggregate representation vector to determine a control instruction, and the control instruction is used to represent the decoded value of the angle recommendation value of the reflector at the next time point. In a specific example of the present application, the temperature compensation timing attenuation aggregate representation vector is input into a temperature compensation adaptive controller based on a decoder to obtain the control instruction. That is, the aggregate representation information of the temperature compensation timing node is used to perform decoding regression, so as to perform adaptive control of temperature compensation, thereby automatically generating the angle data of the reflector at the next time point. In this way, the reflector angle can be adaptively adjusted based on the timing fluctuation and real-time change of temperature, so as to automatically compensate for the temperature deviation effect. Through this intelligent laser parameter adjustment method, it is possible to ensure that a stable high-order azimuthal polarization mode can be maintained under different operating conditions, thereby ensuring the stability of the laser mode and improving the application capability of the high-order azimuthal polarization mode in different scenarios.

[0030] In a preferred example, inputting the temperature-compensated time-attenuation aggregation representation vector into a decoder-based temperature-compensated adaptive controller to obtain a control instruction includes: determining the temperature-compensated time-attenuation aggregation maximum eigenvalue and the temperature-compensated time-attenuation aggregation minimum eigenvalue of the temperature-compensated time-attenuation aggregation representation vector; calculating the temperature-compensated time-attenuation aggregation mean and the temperature-compensated time-attenuation aggregation standard deviation of the feature set of the temperature-compensated time-attenuation aggregation representation vector, and calculating the quotient of the temperature-compensated time-attenuation aggregation mean and the temperature-compensated time-attenuation aggregation standard deviation to obtain a temperature-compensated time-attenuation aggregation statistical normalization value; calculating the inverse of each eigenvalue of the temperature-compensated time-attenuation aggregation representation vector, and comparing it with the temperature-compensated time-attenuation aggregation maximum eigenvalue and the temperature-compensated time-attenuation aggregation standard deviation. After multiplying the difference between the minimum eigenvalues ​​of the temperature-compensated timing attenuation aggregation, subtract the statistical normalized value of the temperature-compensated timing attenuation aggregation to obtain the temperature-compensated timing attenuation aggregation distribution approximation vector; calculate an exponential function with the natural constant as the base, and each eigenvalue of the temperature-compensated timing attenuation aggregation distribution approximation vector as the exponent to obtain the temperature-compensated timing attenuation aggregation distribution class approximation vector; perform point addition on the temperature-compensated timing attenuation aggregation distribution class approximation vector and the statistical normalized value of the temperature-compensated timing attenuation aggregation, and calculate the base-two logarithm of the absolute value of each eigenvalue of the point addition vector to obtain the optimized temperature-compensated timing attenuation aggregation representation vector; and input the optimized temperature-compensated timing attenuation aggregation representation vector into a decoder-based temperature compensation adaptive controller to obtain a control instruction.

[0031] Among them, the temperature compensation timing attenuation aggregation representation vector is recorded as The optimization expression is: ;in, is the temperature-compensated timing attenuation aggregate representation vector, and are respectively the maximum eigenvalue of the temperature-compensated time-attenuation aggregation and the minimum eigenvalue of the temperature-compensated time-attenuation aggregation representation vector, and is the difference between the maximum characteristic value of the temperature compensation timing attenuation aggregation and the minimum characteristic value of the temperature compensation timing attenuation aggregation, is the aggregate mean of the temperature compensated timing decay, is the aggregate standard deviation of the temperature compensated timing decay, Aggregate statistical normalized values ​​for temperature compensated timing decay, is the temperature compensation time series attenuation aggregate distribution class approximation vector, represents exponential operation, represents the logarithm operation with base 2, It's a little plus. is the dot product, It's a point reduction. is the optimized temperature compensated timing decay aggregate representation vector.

[0032] Here, in the preferred example, since the sequence representation of the temperature-compensated local temporal decoding representation vector is a decoding inference representation of the one-dimensional local temporal correlation feature based on the real-time temperature in the laser, when it is input into the feature aggregation network guided by the attenuation of node feature significance, the decoding inference timing node significance difference attributed to the timing node feature distribution difference will also cause attenuation-guided aggregation difference, so that the temperature-compensated timing attenuation aggregation representation vector will also have a diversified set expression distribution of timing node aggregation features. Therefore, it is expected to improve the balance between the regression mapping accuracy and completeness when the temperature-compensated timing attenuation aggregation representation vector is input into the decoder-based temperature-compensated adaptive controller for decoding regression, thereby improving the accuracy of the obtained control instructions.

[0033] Therefore, by performing random statistical normalization on the diversified feature set of the temperature-compensated time-degradation aggregate representation vector, the confidence space constructed based on the overall eigenvalue of the temperature-compensated time-degradation aggregate representation vector is approximated by the standardized continuous probability density distribution relative to the response hypothesis test of each eigenvalue of the temperature-compensated time-degradation aggregate representation vector, thereby establishing the target reachability of the diversified feature distribution of the temperature-compensated time-degradation aggregate representation vector to a unified regression target, so as to achieve the balanced executableness between mapping accuracy and mapping integrity in the decoding regression process based on the diversified feature distribution of the temperature-compensated time-degradation aggregate representation vector, and improve the accuracy of the control instructions obtained by the temperature-compensated time-degradation aggregate feature to the temperature compensation adaptive controller based on the decoder. In this way, the reflector angle can be adjusted more accurately based on the temporal fluctuation and real-time change of temperature, so as to automatically compensate for the temperature deviation effect. Through this intelligent laser parameter adjustment method, it is possible to ensure that stable high-order azimuthal polarization modes can be maintained under different operating conditions, thereby ensuring the stability of the laser mode and improving the application capability of high-order azimuthal polarization modes in different scenarios.

[0034] It is worth mentioning that in the technical solution of the present application, before using the temperature timing fluctuation feature extractor based on the 1D-CNN model, the temperature compensation reasoner based on the decoder, the feature aggregation network guided by the node feature significance attenuation, and the temperature compensation adaptive controller based on the decoder to adaptively adjust the angle of the reflector based on the timing characteristics of the temperature inside the laser, it is also necessary to use a predetermined training strategy to enable the temperature timing fluctuation feature extractor based on the 1D-CNN model, the temperature compensation reasoner based on the decoder, the feature aggregation network guided by the node feature significance attenuation, and the temperature compensation adaptive controller based on the decoder to capture the timing coordination and dynamic correlation characteristics between the reflector angle and the temperature inside the laser.

[0035] In a specific example, the training process involves first collecting a large amount of training data, which includes the temperature changes of the laser during operation and the corresponding reflector angle settings. This data can be obtained through experiments, that is, recording the changes in the reflector angle at different temperatures, and also recording the temperature changes over time. After collecting enough data, the data needs to be preprocessed and feature extracted. This step may include removing noise, smoothing signals, extracting the trend of temperature changes over time, etc.

[0036] Then, the temperature inside the laser is used as sample data and the angle of the reflector is used as label data to train the temperature time series fluctuation feature extractor based on the 1D-CNN model, the temperature compensation reasoner based on the decoder, the feature aggregation network guided by the node feature significance attenuation, and the temperature compensation adaptive controller based on the decoder, so that the model can learn to predict the angle of the reflector to maintain the stability of the laser mode given the current and past temperature data. At the same time, after the training is completed, the model needs to be verified to check whether it can accurately predict the reflector angle on unseen data. If the model performs poorly, it is necessary to adjust the model structure or parameters, or even reconsider the selection of features.

[0037] In this way, when the model training is completed and verified, it can be deployed in the actual application environment. When the laser is running, the temperature data is collected in real time, and the trained model is used to predict the reflector angle at the next time point, so as to achieve adaptive adjustment and automatically compensate for the impact caused by temperature changes.

[0038] In summary, the formation method of the high-order azimuthal polarization mode according to the embodiment of the present application is explained, which learns and depicts the timing coordination and dynamic correlation characteristics between the reflector angle and the temperature in the laser during training, and then in the actual application process, the real-time temperature data in the laser is collected by real-time monitoring through the temperature sensor when the laser is running, and the data processing and analysis algorithm based on artificial intelligence and deep learning is introduced at the back end to perform timing dynamic feature analysis and temperature compensation calculation based on the aggregation of timing attenuation information on the real-time temperature data in the laser, so as to automatically generate the angle data of the reflector at the next time point. In this way, the reflector angle can be adaptively adjusted based on the timing fluctuation and real-time change of the temperature, so as to automatically compensate for the temperature deviation effect. Through this intelligent laser parameter adjustment method, it is possible to ensure that a stable high-order azimuthal polarization mode can be maintained under different operating conditions, thereby ensuring the stability of the laser mode and improving the application capability of the high-order azimuthal polarization mode in different scenarios.

[0039] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for forming a high-order azimuthal polarization mode, characterized in that: include: A time queue for obtaining the real-time temperature in the laser collected by the temperature sensor; The timing fluctuation characteristics of the time queue of the real-time temperature in the laser are extracted to obtain a sequence of local timing correlation feature vectors of the real-time temperature inside the laser; temperature compensation reasoning is performed on each local timing correlation feature vector of the real-time temperature inside the laser in the sequence to obtain a sequence of temperature compensated local timing decoding representation vectors; the sequence of temperature compensated local timing decoding representation vectors is input into a feature aggregation network guided by node feature significance attenuation to obtain a temperature compensated timing attenuation aggregation representation vector; temperature compensation adaptive control is performed based on the temperature compensated timing attenuation aggregation representation vector to determine a control instruction, wherein the control instruction is used to represent a decoded value of a recommended angle value of the reflector at the next time point.

2. The method for forming a high-order azimuthal polarization mode according to claim 1, characterized in that: The method comprises: extracting the timing fluctuation characteristics of the time queue of the real-time temperature in the laser to obtain a sequence of local timing correlation feature vectors of the real-time temperature in the laser, including: inputting the time queue of the real-time temperature in the laser into a temperature timing fluctuation feature extractor based on a 1D-CNN model to obtain a sequence of local timing correlation feature vectors of the real-time temperature in the laser.

3. The method for forming a high-order azimuthal polarization mode according to claim 2, characterized in that: Temperature compensation reasoning is performed on each laser internal real-time temperature local timing association feature vector in the sequence of laser internal real-time temperature local timing association feature vectors to obtain a sequence of temperature compensated local timing decoding representation vectors, including: inputting each laser internal real-time temperature local timing association feature vector in the sequence of laser internal real-time temperature local timing association feature vectors into a decoder-based temperature compensation reasoner to obtain the sequence of temperature compensated local timing decoding representation vectors.

4. The method for forming a high-order azimuthal polarization mode according to claim 3, characterized in that: The sequence of the temperature-compensated local temporal decoding representation vectors is input into a feature aggregation network guided by node feature significance attenuation to obtain a temperature-compensated temporal attenuation aggregation representation vector, including: calculating the feature significance description factor of each temperature-compensated local temporal decoding representation vector in the sequence of the temperature-compensated local temporal decoding representation vectors, wherein the feature significance description factor is related to the mean and variance of each temperature-compensated local temporal decoding representation vector; constructing the feature significance attenuation factor of each temperature-compensated local temporal decoding representation vector based on the distance span between each temperature-compensated local temporal decoding representation vector in the sequence of the temperature-compensated local temporal decoding representation vector and the current temperature-compensated local temporal decoding representation vector; and performing feature significance attenuation on the feature significance description factor of each temperature-compensated local temporal decoding representation vector based on the feature significance attenuation factor of each temperature-compensated local temporal decoding representation vector. The method comprises the steps of: modulating the feature significance attenuation description factors of the temperature compensated local temporal decoding representation vectors based on the feature significance attenuation factors of the temperature compensated local temporal decoding representation vectors to obtain a sequence of feature significance attenuation description factors; inputting the sequence of feature significance attenuation description factors into a gated mask module to obtain a sequence of mask feature significance attenuation weight factors; taking the sequence of mask feature significance attenuation weight factors as a sequence of weights, calculating the weighted sum of the sequence of temperature compensated local temporal decoding representation vectors to obtain the temperature compensated temporal attenuation aggregate representation vector; wherein, based on the feature significance attenuation factors of the temperature compensated local temporal decoding representation vectors, modulating the feature significance descriptors of the temperature compensated local temporal decoding representation vectors to obtain a sequence of feature significance attenuation description factors comprises: calculating the positional product between the feature significance attenuation factors of the temperature compensated local temporal decoding representation vectors and the feature significance descriptors of the temperature compensated local temporal decoding representation vectors to obtain a sequence of feature significance attenuation description factors.

5. The method for forming a high-order azimuthal polarization mode according to claim 4, characterized in that: Calculate the feature significance description factor of each temperature compensated local temporal decoding representation vector in the sequence of the temperature compensated local temporal decoding representation vector, wherein the feature significance description factor is related to the mean and variance of each temperature compensated local temporal decoding representation vector, including: respectively calculating the mean and variance of the temperature compensated local temporal decoding representation vector to obtain the temperature compensated local temporal feature mean and the temperature compensated local temporal feature variance; after calculating the fourth power of the positional difference between the temperature compensated local temporal decoding representation vector and the temperature compensated local temporal feature mean, calculate the expected value of the fourth power modulated semantic deviation vector to obtain the temperature compensated local temporal feature expectation factor; calculate the division between the temperature compensated local temporal feature expectation factor and the square of the temperature compensated local temporal feature variance to obtain the feature significance description factor of the temperature compensated local temporal decoding representation vector.

6. The method for forming a high-order azimuthal polarization mode according to claim 5, characterized in that: Based on the distance span between each temperature compensated local timing decoding representation vector in the sequence of the temperature compensated local timing decoding representation vector and the current temperature compensated local timing decoding representation vector, a feature significance attenuation factor of each temperature compensated local timing decoding representation vector is constructed, including: calculating the distance span between other temperature compensated local timing decoding representation vectors in the sequence of the temperature compensated local timing decoding representation vector and the current temperature compensated local timing decoding representation vector to obtain a feature distance span; extracting the maximum value in the current temperature compensated local timing decoding representation vector to obtain the current temperature compensated local timing feature maximum value; extracting the maximum value in the other temperature compensated local timing decoding representation vectors to obtain other temperature compensated local timing feature maximum values; calculating the difference between the current temperature compensated local timing feature maximum value and the other temperature compensated local timing feature maximum values ​​and dividing the difference by the feature distance span to obtain the feature significance attenuation factor.

7. The method for forming a high-order azimuthal polarization mode according to claim 6, characterized in that: The sequence of feature significance attenuation descriptors is input into a gated mask module to obtain a sequence of masked feature significance attenuation weight factors, including: taking the negative number of each feature significance attenuation descriptor in the sequence of feature significance attenuation descriptors as an exponential power, calculating an exponential function value with a natural constant e as the base to obtain a sequence of feature significance attenuation class support descriptors; calculating the reciprocal of the sum of each feature significance attenuation class support descriptor in the sequence of feature significance attenuation class support descriptors and a constant one to obtain a sequence of feature significance attenuation weight factors; based on a predetermined threshold, masking the sequence of feature significance attenuation weight factors to obtain a sequence of masked feature significance attenuation weight factors, wherein, in response to each feature significance attenuation weight factor in the sequence of feature significance attenuation weight factors being greater than the predetermined threshold, the feature significance attenuation weight factor greater than the predetermined threshold is taken to have an original value, otherwise it is set to zero.

8. The method for forming a high-order azimuthal polarization mode according to claim 7, characterized in that: Taking the sequence of the mask feature significance attenuation weight factors as the sequence of weights, calculating the weighted sum of the sequence of the temperature-compensated local temporal decoding representation vectors to obtain the temperature-compensated temporal attenuation aggregate representation vector, including: taking the sequence of the mask feature significance attenuation weight factors as the sequence of weights, calculating the position-wise weighted sum between each other temperature-compensated local temporal decoding representation vector in the sequence of the temperature-compensated local temporal decoding representation vector to obtain the historical temperature-compensated local temporal decoding representation vector; calculating the position-wise sum between the historical temperature-compensated local temporal decoding representation vector and the current temperature-compensated local temporal decoding representation vector to obtain the temperature-compensated temporal attenuation aggregate representation vector.

9. The method for forming a high-order azimuthal polarization mode according to claim 8, characterized in that: Temperature compensation adaptive control is performed based on the temperature compensation timing attenuation aggregate representation vector to determine a control instruction, wherein the control instruction is used to represent a decoded value of the recommended angle value of the reflector at the next time point, including: inputting the temperature compensation timing attenuation aggregate representation vector into a decoder-based temperature compensation adaptive controller to obtain the control instruction.

Images