A device and method for measuring the quality of vortex light transmission in a sea-air environment

Abstract

The application discloses a kind of sea-air environment under vortex light transmission quality measuring device and method, it is related to laser communication technical field, the device includes: optical emission guide unit, for emitting signal light;Vortex light generating device, for signal light is converted into vortex light beam;Adaptive optical compensation unit, for based on sea-air vortex light beam and vortex light beam dynamic construction comprehensive performance index, and based on comprehensive performance index, using improved SPGD algorithm generation control vector, phase compensation is carried out to sea-air vortex light beam, and the vortex light beam after compensation is generated;Interference detection unit, for the interference fusion of vortex light after compensation and signal light, generates interference image;Data processing unit, for based on interference image, using improved EfficientNet-B0 model obtains the transmission quality data of vortex light.The application can improve the accuracy and efficiency of vortex light transmission quality measurement.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of laser communication, in particular to a device and method for measuring transmission quality of vortex light in a sea-air environment. BACKGROUND

[0002] In the field of modern optical communication and detection technology, vortex light has shown great application potential due to its unique orbital angular momentum (OAM). In theory, OAM has infinitely many eigenstates, which provides a new multiplexing dimension for optical communication and is expected to greatly improve the transmission capacity and spectral efficiency of the communication system. Vortex light has been widely researched and applied in many fields such as free-space optical communication, optical imaging, micro-operation, and underwater quantum optical communication.

[0003] In the field of detection technology of orbital angular momentum of vortex light, the topological charge number and the degradation degree of OAM mode are the key indicators for evaluating the transmission quality of vortex light. However, the methods in some cases have obvious limitations in measuring the topological charge number and the degradation degree of OAM mode of vortex light. For example, some schemes are based on turbulence phase compensation and beam geometric transformation, which can improve the detection range, but the system structure is complex and difficult to integrate. Some other schemes use polarization characteristics for detection, which simplifies the operation, but only applies to specific types of vortex light beams and has insufficient generalization ability. The detection method based on deep learning has high efficient recognition ability in the simulation environment, but does not consider the complex wavefront distortion and OAM degradation caused by multi-factor coupling in the real sea-atmosphere composite transmission environment, which has insufficient reliability in actual application. Therefore, most of the technologies cannot comprehensively and accurately test the topological charge number and the degradation degree of OAM mode of vortex light in the complex environment of the sea-air interface area, and cannot determine the good or bad of the transmission quality of vortex light. SUMMARY

[0004] The purpose of the present application is to provide a device and method for measuring the transmission quality of vortex light in a sea-air environment, which can improve the accuracy and efficiency of vortex light transmission quality measurement.

[0005] To achieve the above-mentioned purpose, the present application provides the following solutions.

[0006] In a first aspect, the present application provides a device for measuring the transmission quality of vortex light in a sea-air environment, comprising:

[0007] an optical emission guiding unit, a vortex light generating device, a channel simulation unit, an adaptive optical compensation unit, an interference detection unit, and a data processing unit; the channel simulation unit is used to simulate a sea-air environment;

[0008] The optical emission guiding unit is used to emit signal light.

[0009] The vortex light generating device is configured to convert the signal light into a vortex light beam; the vortex light beam generates a sea-air vortex light beam by passing through the sea-air environment.

[0010] The adaptive optical compensation unit is configured to dynamically construct a comprehensive performance index based on the sea-air vortex light beam and the vortex light beam, and generate a control vector based on the comprehensive performance index by using a modified SPGD algorithm to perform phase compensation on the sea-air vortex light beam to generate a compensated vortex light beam.

[0011] The interference detection unit is configured to perform interference fusion on the compensated vortex light and the signal light to generate an interference image.

[0012] The data processing unit is configured to obtain transmission quality data of the vortex light based on the interference image by using an improved EfficientNet-B0 model; the transmission quality data includes a topological charge number and a degradation degree of an OAM mode; and the improved EfficientNet-B0 model adds a lightweight attention module and an early classification exit to the EfficientNet-B0 model.

[0013] In a second aspect, the present application provides a method for measuring transmission quality of vortex light in a sea-air environment, comprising:

[0014] Obtaining signal light and converting the signal light into a vortex light beam; the vortex light beam generates a sea-air vortex light beam by passing through the sea-air environment.

[0015] Dynamically constructing a comprehensive performance index based on the sea-air vortex light beam and the vortex light beam, and generating a control vector based on the comprehensive performance index by using a modified SPGD algorithm to perform phase compensation on the sea-air vortex light beam to generate a compensated vortex light beam.

[0016] Performing interference fusion on the compensated vortex light and the signal light to generate an interference image.

[0017] Inputting the interference image into an improved EfficientNet-B0 model to obtain transmission quality data of the vortex light; the transmission quality data includes a topological charge number and a degradation degree of an OAM mode; and the improved EfficientNet-B0 model adds a lightweight attention module and an early classification exit to the EfficientNet-B0 model.

[0018] According to the specific embodiments provided by the present application, the present application has the following technical effects:

[0019] The application firstly simulates the sea-air environment through the channel simulation unit, generates a sea-air vortex light beam after the vortex light beam passes through the sea-air environment, and is more suitable for actual application, and then constructs a comprehensive performance index based on the sea-air vortex light beam and the vortex light beam through the adaptive optical compensation unit, generates a control vector based on the comprehensive performance index by using the improved SPGD algorithm, and compensates the phase of the sea-air vortex light beam, so that the final compensated vortex light beam is more suitable for the phase of the initial vortex light beam, thereby improving the transmission quality of the compensated vortex light beam, and finally, the improved EfficientNet-B0 model is used to detect the degradation degree of the topological charge number and the OAM mode. Since the light attention module early classification outlet is added, not only can the feature map be adaptively optimized and enhanced, but also the transmission quality data of the vortex light can be quickly outputted, the rapid detection is realized, and the accuracy and efficiency of the vortex light transmission quality measurement are improved. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0021] Figure 1 A structural schematic diagram of a sea-air environment vortex light transmission quality measurement device provided by the present application.

[0022] Figure 2 A flowchart of a sea-air environment vortex light transmission quality measurement method provided by the present application.

[0023] Figure 3 A phase diagram of a vortex light beam provided by the present application.

[0024] Figure 4 A phase diagram of a sea-air vortex light beam provided by the present application.

[0025] Figure 5 A phase diagram of a compensated vortex light beam provided by the present application.

[0026] Figure 6 A structural schematic diagram of an improved EfficientNet-B0 model provided by the present application.

[0027] Figure 7 A schematic diagram of an interference pattern when the topological charge number is-3 provided by the present application.

[0028] Figure 8This is a schematic diagram of the interference pattern provided in this application when the topological charge number is -2.

[0029] Figure 9 This is a schematic diagram of the interference pattern provided in this application when the topological charge number is -1.

[0030] Figure 10 This is a schematic diagram of the interference pattern provided in this application when the topological charge number is 0.

[0031] Figure 11 This is a schematic diagram of the interference pattern provided in this application when the topological charge number is 1.

[0032] Figure 12 This is a schematic diagram of the interference pattern provided in this application when the topological charge number is 2.

[0033] Figure 13 This is a schematic diagram of the interference pattern when the topological charge number is 3, as provided in this application.

[0034] Reference numerals: Optical emission guiding unit-1, laser-11, collimating beam expander-12, polarizer-13, first beam splitter prism-14, reflector-15, vortex light generating device-2, spatial light modulator-21, spatial light modulation controller-22; channel simulation unit-3, adaptive optics compensation unit-4, second beam splitter prism-41, reflective deformable mirror-42, third beam splitter prism-43, first CCD detector-44, second CCD detector-45, optical compensation controller-46, interferometric detection module-5, data processing unit-6. Detailed Implementation

[0035] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0036] When vortex light is transmitted in the sea-air environment, it will face many severe challenges. The sea-air environment is extremely complex, and the flow of seawater, temperature gradient, salinity gradient, and aerosols, dust, and other factors in the atmosphere will have a significant impact on the transmission of vortex light. In particular, the quasi-ordered structure existing in the ocean turbulence will cause the symmetry of the OAM spectrum distribution of the vortex light to break down, and then the OAM will change. Studies have shown that in the ocean turbulence environment, serious cross-talk will occur between different OAM modes, which severely limits the effective application of OAM modes in optical communication. Moreover, the salt particles, aerosols and other factors in the marine environment have scattering and absorption effects on vortex light, causing the energy of the light signal to attenuate, further exacerbating the OAM degradation of vortex light. However, most of the technologies in detecting the OAM degradation of vortex light only consider the influence of a single factor, and cannot comprehensively and accurately simulate and test the OAM degradation caused by the coupling of multiple factors in the sea-air environment, making it difficult to meet the actual application requirements. Therefore, the present application proposes a vortex light transmission quality measurement device and method in the sea-air environment. First, the multi-factor coupling channel that causes the OAM mode degradation is reproduced in a controllable environment (i.e., simulating the sea-air environment); then, the vortex light with complex distortion after transmission is interfered with the ideal reference light, and the phase distortion and mode cross-talk information that cannot be directly observed are encoded into an interference image; finally, the improved EfficientNet-B0 model is used to decode the interference image, thereby realizing the robust identification of the topological charge number of the vortex light and the quantitative evaluation of the mode degradation degree, and improving the accuracy and efficiency of the vortex light transmission quality measurement.

[0037] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0038] In one exemplary embodiment, as shown in Figure 1 A vortex light transmission quality measurement device in the sea-air environment is provided, which includes an optical emission guiding unit 1, a vortex light generating device 2, a channel simulation unit 3, an adaptive optical compensation unit 4, an interference detection unit, and a data processing unit 5; the channel simulation unit 3 is used to simulate the sea-air environment.

[0039] The optical emission guiding unit 1 is used to emit signal light.

[0040] As an implementable way, the optical emission guiding unit 1 includes a laser 11, a collimating and expanding mirror 12, a polarizer 13, a first beam splitting prism 14, and a reflecting mirror 15 arranged in sequence.

[0041] The laser emitted by the laser 11 is first expanded and collimated by the collimating beam expander 12, and then transmitted by the polarizing plate 13 to obtain polarized light; the polarized light is divided into two signal lights by the first beam splitter prism 14; one of the signal lights is incident into the vortex light generating device 2; the other signal light is incident into the interference detection unit 5 through the reflecting mirror 15.

[0042] Specifically, the laser 11 selects a 532nm continuous wave ND:YAG laser, the collimating beam expander 12 is connected to the right of the laser 11, the output end of the laser 11 is aligned with the collimating beam expander 12, and the center of the light beam is ensured to be incident into the center of the collimating beam expander 12. The collimating beam expander 12 expands the diameter of the light beam output by the laser 11 to 9mm to meet the working aperture of the vortex light generating device 2, and then the light beam is incident into the polarizing plate 13 to adjust the polarization state of the light beam to match the spatial light modulator 21. The first beam splitter prism 14 is a non-polarized beam splitter prism with a splitting ratio of 1:1, which is located on the output light path of the right end of the polarizing plate 13, and divides the light beam into two signal lights: one signal light (i.e. the transmitted light beam) enters the vortex light generating device 2, and the other signal light (i.e. the reflected light beam) is guided to the interference and detection module 5 through the reflecting mirror 15 as reference light.

[0043] The vortex light generating device 2 is used for converting the signal light into a vortex light beam; the vortex light beam passes through the sea-air environment to generate a sea-air vortex light beam.

[0044] As an implementable manner, the vortex light generating device 2 comprises:

[0045] The spatial light modulation controller 22 is connected with the spatial light modulator 21 and the data processing unit 6 respectively, and is used for receiving the control instruction of the data processing unit 6 to control the spiral phase pattern of the spatial light modulator 21.

[0046] Specifically, by inputting a specific phase pattern program through the spatial light modulation controller 22, the spatial light modulation controller 22 loads the spiral phase pattern on the spatial light modulator 21. The spatial light modulator 21 is an SLM spatial light modulator.

[0047] The spatial light modulator 21 is used for converting the signal light into a vortex light beam through the spiral phase pattern.

[0048] Specifically, the spatial light modulator 21 adopts a TSLM07U-A transmissive modulator with a resolution of 1920x1080, an effective working area of 16.3mmx9.18mm, and a spectral range of 420nm-1200nm, which is located on the transmitted light path of the optical emission and guiding unit 1, the surface of which is perpendicular to the incident light axis, and the spatial light modulation controller 22 and the spatial light modulator 21 are connected through a data line, and the spiral phase pattern can be updated in real time.

[0049] As an implementable manner, the channel simulation unit 3 comprises a sea fog box and a turbulence pool, the sea fog box is used to generate simulated sea fog in a sea-air environment, and the turbulence pool is used to generate turbulence in the sea-air environment.

[0050] Specifically, the channel simulation unit 3 integrates a sea fog box, a turbulence pool, a salinity adjustment system, a temperature control system, and a turbulence simulation device, and can accurately control key parameters such as salinity, temperature, and turbulence intensity. It provides reliable simulation conditions for studying the transmission of vortex light in a real sea-air environment.

[0051] The sea fog box is a plexiglass water tank. The main task of the channel simulation unit 3 is to simulate the sea-air channel through the sea fog box and the turbulence pool. The sea fog box adopts a multi-layer sea fog simulation box structure. In each fog body maintenance system, the principle of ultrasonic vibration is used to change the water body into 0.5-1 μm water mist particles to simulate the thin sea fog environment in the sea-air environment. Different concentrations of salt water can be generated into 1-2 μm salt mist particles through ultrasonic vibration to simulate the thick sea fog environment in the sea-air environment. The turbulence pool adopts a box structure, with light holes on the left and right to support laser incidence and emission, and condenser plates and heat plates at the top and bottom, respectively. The sea fog simulation environment in the box forms turbulence through the temperature difference between the top and bottom. The size of the turbulence is controlled by controlling the temperature of the heat plate, and the specific turbulence environment is formed by changing the variables according to the experimental requirements.

[0052] The adaptive optical compensation unit 4 is used to dynamically construct a comprehensive performance index based on the sea-air vortex beam and the vortex beam, and generate a control vector based on the comprehensive performance index, to perform phase compensation on the sea-air vortex beam, and generate a compensated vortex beam.

[0053] Specifically, the adaptive optical compensation unit 4 is used to correct the wavefront distortion of the sea-air vortex beam generated in the channel transmission in real time, to reduce or correct the wavefront distortion, improve the beam quality of the optical signal entering the interference link, and improve the fidelity of the OAM mode corresponding to the target preset topological charge number.

[0054] As an implementable manner, the adaptive optical compensation unit 4 comprises:

[0055] The second beam splitting prism 41 is arranged on the exit light path of the channel simulation unit, and is used to divide the sea-air vortex beam into two sea-air vortex beams. One sea-air vortex beam is incident into the first CCD detector, and the other sea-air vortex beam is incident into the reflective deformable mirror.

[0056] The first CCD detector 44 is used to convert the sea-air vortex beam into a first image.

[0057] The reflective deformable mirror 42 is used to perform phase compensation on the sea-air vortex beam, and generate a compensated vortex beam.

[0058] The third beam splitter prism 43 is set in the output light path of the reflective deformable mirror 42 and is used to split the compensated sea-air vortex light into two compensated vortex beams; one compensated vortex beam is injected into the second CCD detector 45; and the other compensated vortex beam is injected into the interferometric detection unit 5.

[0059] The second CCD detector 45 is used to convert the compensated vortex beam into a second image.

[0060] The optical compensation controller 46 is connected to the data processing unit 6, the first CCD detector 44, the reflective deformable mirror, and the second CCD detector 45, respectively. It is used to receive control commands from the data processing unit 6, dynamically construct comprehensive performance indicators based on the ocean-atmosphere vortex beam and the vortex beam, and generate control vectors based on the comprehensive performance indicators using the improved SPGD algorithm to perform phase control on the reflective deformable mirror.

[0061] Specifically, the ocean-air vortex beam emitted from the channel simulation unit 3 is split into two beams by the second beam splitter 41. Most of the energy is incident on the reflective deformable mirror 42, which compensates for the wavefront distortion. After the compensated ocean-air vortex beam is split by the third beam splitter 43, most of the energy goes to the interferometric detection unit 5, and a small portion of the energy goes to the second CCD detector 45 to detect the beam quality and provide real-time feedback to the controller to execute the improved SPGD algorithm.

[0062] Specifically, such as Figure 1 As shown, the second beam-splitting prism 41 is located at the right end of the channel simulation unit 3, receiving the sea-air vortex beam with a splitting ratio of 1:9. 10% of the energy enters the first CCD detector 44 to acquire image information, and 90% of the energy enters the reflective deformable mirror 42. The optical compensation controller 46 is an industrial control computer running a modified SPGD algorithm. After receiving the data, it generates a drive signal to control the reflective deformable mirror 42 to achieve closed-loop compensation, thereby realizing the braking of the reflective deformable mirror 42 and thus implementing phase modulation of the sea-air vortex beam. The reflective deformable mirror 42 is an ALPAO electromagnetic small-aperture reflective deformable mirror DM97-15 model with a working aperture of 13.5mm and a tilt deformation of 60μm. It is located on the right end of the transmission light path of the second beam splitter prism 41 and receives 90% of the energy beam from the beam splitter prism 41. The third beam splitter prism 43 is located on the reflection light path of the reflective deformable mirror 42 with a splitting ratio of 1:9. 10% of the energy is incident to the second CCD detector 45 for measuring compensation data and feeding it back to the optical compensation controller 46. Under the drive of the optical compensation controller 46, the compensation surface shape is regenerated and iterated continuously. The remaining energy is emitted to the interferometric detection unit 5.

[0063] Interference detection unit 5 is used to perform interference fusion of the compensated vortex light and the signal light to generate an interference image.

[0064] Specifically, the interference detection unit 5 collects and transmits the interference image with high precision, and assists in the calibration and optimization of the auxiliary light path, thereby providing a basis for the analysis of the OAM degradation degree of the vortex light. The interference detection unit 5 is in a Mach-Zehnder structure. The reference arm of the interference detection unit 5 receives the reflected light of the optical emission guide unit 1, and the signal arm of the interference detection unit 5 receives the compensated vortex light beam output by the adaptive optical compensation unit 4. The optical path difference between the two arms is compensated to be less than or equal to 5 microns through a fiber delay line, and the interference fringe contrast is greater than or equal to 80%. After the interference detection unit 5 collects the interference fringe image, the interference fringe image is transmitted to the data processing unit 6 through a GigE interface.

[0065] The data processing unit 6 is configured to obtain transmission quality data of the vortex light based on the interference image by using an improved EfficientNet-B0 model. The transmission quality data includes a topological charge number and an OAM mode degradation degree. The improved EfficientNet-B0 model adds a lightweight attention module and an early classification exit to the EfficientNet-B0 model.

[0066] The data processing unit 6 is an industrial computer equipped with a GPU.

[0067] Specifically, the improved EfficientNet-B0 model is used. The innovation lies in the introduction of a lightweight attention module and an early classification exit. The dynamic inference path is generated through the early classification exit, which significantly improves the real-time response speed of the model while maintaining high precision. The lightweight attention module is embedded between the network levels. Through mixed attention calculation of the channel and spatial dimensions of the feature map, the key areas related to OAM degradation in the interference fringe image are quickly focused, and background noise is suppressed, thereby improving the feature extraction efficiency. The dynamic inference path sets multiple early classification exits in the middle of the network and designs a confidence evaluation function to judge the difficulty of the samples in the inference process. For high-confidence samples, the model outputs the classification and degradation evaluation results in advance, avoiding complete network forward propagation, thereby realizing conditional calculation at the system level and significantly reducing the average inference delay.

[0068] Specifically, the OAM mode degradation degree specifically refers to the distortion of the orbital angular momentum (OAM) mode after transmission, and the topological charge number is a core parameter representing the spiral phase structure of the vortex light beam. The device of the present application does not directly measure the physical value of OAM, but evaluates the degradation degree of the mode purity, wavefront phase and light intensity distribution of the vortex light beam with a preset topological charge number (i.e. the OAM mode degradation degree) after transmission by simulating a complex sea-atmosphere channel. These factors jointly determine the usability and reliability of the OAM mode in communication.

[0069] The beneficial effects of the vortex light transmission quality measuring device in the sea-air environment provided in the application mainly manifest in the following aspects:

[0070] 1. The application accurately reproduces the complex sea-air environment coupled by multiple factors such as salinity, temperature, and turbulence in a unified channel simulation unit through the channel simulation unit, integrates sea fog simulation, salinity adjustment, temperature control, and turbulence simulation device, overcomes the limitations of considering only a single factor in the prior art, and provides reliable simulation conditions for studying the transmission of vortex light in real sea-air environment.

[0071] 2. The adaptive optical compensation unit is based on the sea-air vortex light beam and the vortex light beam dynamic comprehensive performance index, and based on the comprehensive performance index, a control vector is generated by using a modified SPGD algorithm to compensate the phase of the sea-air vortex light beam, so that the final compensated vortex light beam is more consistent with the phase of the initial vortex light beam, thereby improving the transmission quality of the compensated vortex light beam, and the efficiency is higher and the accuracy is better compared with no compensation or traditional compensation method.

[0072] 3. The application forms a complete automatic closed loop. The data processing unit adopts an improved EfficientNet-B0 model with a lightweight attention module early classification outlet, which can not only adaptively optimize and enhance the feature map, but also quickly output the transmission quality data of the vortex light, realize rapid detection, and improve the accuracy and efficiency of the vortex light transmission quality measurement. It meets the practical needs of real-time and efficient monitoring of vortex light OAM state, and effectively promotes the vortex light technology from theoretical research to engineering application.

[0073] 4. During the test operation process, the application realizes automatic and efficient testing. The data processing unit combines deep learning, digital image processing and other technologies to receive data from the CCD camera and environmental sensors, automatically outputs the vortex light OAM degradation degree result and stores the data after preprocessing, deep learning inference and correlation analysis, and can also display the result in real time through the display. This reduces manual intervention and greatly improves test efficiency and accuracy, providing strong support for the practicality of vortex light technology and promoting related technologies from theoretical research to practical application.

[0074] Based on the same inventive concept, the application also provides a vortex light transmission quality measuring method in a sea-air environment. The method is applied to the vortex light transmission quality measuring device in the sea-air environment, and an exemplary embodiment is shown in Figure 2 , which provides a vortex light transmission quality measuring method in a sea-air environment, including steps S1 to S4:

[0075] Step S1: obtaining signal light and converting the signal light into a vortex beam, the vortex beam passing through the sea-air environment to generate a sea-air vortex beam.

[0076] Specifically, when the signal light is converted into the vortex beam, the helical phase pattern is adjusted by a spatial light modulator, so that the polarization state of the vortex beam can match the working state of the spatial light modulator. The electric field expression of the vortex beam is:

[0077] .

[0078] wherein, is the electric field of the vortex beam, is a radial distance, i.e., a straight-line distance from the center of the vortex beam to a point, is an angular coordinate, used to describe the azimuth angle of a point relative to the center of the beam, is an axis, is the electric field of the vortex beam at the initial position, describing the intensity spatial distribution of the beam at the initial position, represents a helical phase structure, is a virtual part of the phase, is a topological charge number, is a wave number, related to the wavelength of the beam , and the specific relationship is , used to describe the propagation characteristics of light.

[0079] At the same time, the phase distribution of the vortex beam is:

[0080] .

[0081] wherein, is the phase of the vortex beam.

[0082] Step S2: dynamically constructing a comprehensive performance index based on the sea-air vortex beam and the vortex beam, and generating a control vector based on the comprehensive performance index by using a modified SPGD algorithm to perform phase compensation on the sea-air vortex beam, to generate a compensated vortex beam.

[0083] As an implementable manner, step S2 specifically includes steps S21 to S24:

[0084] Step S21: dynamically constructing a comprehensive performance index based on the sea-air vortex beam and the vortex beam.

[0085] As an implementable manner, step S21 specifically includes steps S211 to S2111:

[0086] Step S211: based on the vortex beam, determining a vortex light field and a vortex annular area of the vortex beam.

[0087] Step S212: Based on the ocean-atmosphere vortex beam, determine the vortex light field and phase of the ocean-atmosphere vortex beam.

[0088] Step S213: Calculate the sea fog intensity and turbulence intensity based on the vortex light field of the vortex beam, the vortex light field of the sea-air vortex beam, and the phase of the sea-air vortex beam.

[0089] Specifically, sea fog intensity and turbulence intensity The calculation formula is:

[0090] .

[0091] in, This refers to the vortex light field of a vortex beam. This represents the vortex light field of an ocean-atmosphere vortex beam. This represents the phase of the ocean-atmosphere vortex beam. For gradient.

[0092] Step S214: Calculate the baseline weights based on sea fog intensity and turbulence intensity.

[0093] As an feasible approach, the formula for calculating the benchmark weight is:

[0094] .

[0095] in, As the benchmark weight for the model purity index, As the initial weights, The first turbulence sensitivity coefficient, For the first Turbulence intensity estimation at the next iteration The first sea fog sensitivity coefficient, For the first Sea fog intensity estimation at the next iteration The benchmark weight for the performance index of phase topology integrity, The initial weights for the performance index of phase topology integrity, This is the second turbulence sensitivity coefficient. The second sea fog sensitivity coefficient, The benchmark weights for performance indicators representing the spatial distribution characteristics of beam intensity. The initial weights for the performance index of the spatial distribution characteristics of the beam intensity are given. This is the third turbulence sensitivity coefficient.

[0096] Step S215: Normalize the baseline weights to obtain adaptive weights.

[0097] .

[0098] .

[0099] in, For the adaptive weights of the model purity index, An adaptive weight for the performance index of phase topology integrity. Adaptive weights for performance indicators representing the spatial distribution characteristics of beam intensity. This is the sum of the baseline weights.

[0100] Step S216: Calculate the mode purity index based on the vortex light field of the ocean-atmosphere vortex beam.

[0101] .

[0102] .

[0103] .

[0104] .

[0105] .

[0106] in, As an indicator of model purity, The mode coefficients of the target vortex beam. The topological charge number is The mode coefficients of the vortex beam. The denominator is a constant; to prevent the denominator from being zero, The vortex beam after compensation, i.e., the vortex beam after reflection by the deformable mirror. Phase compensation applied to the reflective deformable mirror The phase of the disturbance caused by turbulence. The amplitude attenuation and phase perturbation function when traversing the air-sea environment. For the first The control vector of each actuator For the first The response function of an actuator It is the radial distance from the center of the reflective deformable mirror to a point on its surface.

[0107] Step S217: Calculate the estimated topological charge number based on the vortex light field of the vortex beam.

[0108] Specifically, by calculating the loop integral of the phase gradient along a closed path to predict the topological charge, and using this to construct a performance index for the phase topological integrity of the beam, the predicted topological charge can be expressed by the following formula:

[0109] .

[0110] wherein, is the phase of the vortex optical field of the vortex beam, is the gradient of the phase of the vortex optical field of the vortex beam, is the estimated topological charge.

[0111] Step S218: based on the estimated topological charge, a performance index of the phase topological structure integrity is calculated.

[0112] Specifically, the performance index of the phase topological structure integrity is:

[0113] .

[0114] wherein, is the performance index of the phase topological structure integrity, is the target topological charge, is a positive number to prevent the denominator from being zero, and also can smooth the index function.

[0115] Step S219: based on the vortex ring area, a performance index of the spatial distribution characteristics of the beam intensity is calculated.

[0116] Specifically, the ultimate purpose of compensating the wavefront is to restore the quality of the vortex light, and the quality of the vortex light is not only reflected in the phase, but also reflected in the intensity distribution, so as to construct a performance index of the spatial distribution characteristics of the beam intensity.

[0117] Let the intensity distribution of the ideal vortex beam in the ring area be , and the compensated intensity be , then:

[0118] .

[0119] wherein, is the performance index of the spatial distribution characteristics of the beam intensity, indicates the vortex ring area.

[0120] Step S2110: the mode purity index, the performance index of the phase topological structure integrity and the performance index of the spatial distribution characteristics of the beam intensity are normalized to obtain the normalized mode purity index, the normalized performance index of the phase topological structure integrity and the normalized performance index of the spatial distribution characteristics of the beam intensity.

[0121] Specifically, the normalized mode purity index , the normalized phase topological structure integrity performance index , and the normalized light intensity spatial distribution feature performance index of the light beam The expression of the comprehensive performance index is as follows:

[0122] .

[0123] Step S2111: Based on the normalized mode purity index, the normalized phase topological structure integrity performance index, the normalized light intensity spatial distribution feature performance index of the light beam, and the adaptive weight, a comprehensive performance index is constructed.

[0124] Specifically, the expression of the comprehensive performance index is as follows:

[0125] .

[0126] Step S22: An initial control vector is obtained.

[0127] Step S23: Based on the vortex light field of the sea-air vortex light beam, the initial control vector is iterated multiple times by the improved SPGD algorithm until the comprehensive performance index reaches a threshold value or reaches the maximum iteration number of stops, and a control vector is generated.

[0128] Specifically, during the iteration of the improved SPGD algorithm, first, the initial control vector is initialized, and let , the gain coefficient is set, and the iteration number .

[0129] The disturbance quantity is generated, the disturbance of each element is randomly taken in ±1, and finally normalized processing is performed: .

[0130] The positive and negative disturbances are applied, so that the actuators of the reflective deformable mirror act, thereby exerting an initial compensation on the vortex light beam. At this time, the image collected by the second CCD detector can obtain the performance index at the first iteration based on the disturbance , . The corresponding relationship is as follows:

[0131] .

[0132] The change value of the comprehensive performance index at the first iteration is calculated as follows: .

[0133] The control vector is updated and iterated: .​

[0134] wherein the gain coefficient in the first iteration is calculated as:

[0135] .

[0136] wherein, the gain coefficient in the first iteration, which determines the size of the update step. is the decay coefficient, which controls the rate of exponential decay of the gain with the number of iterations, the larger the decay is, the faster the decay is. is the total number of iterations, is the gradient sensitivity coefficient, which controls the degree of response of the gain to the gradient size. The gain adjustment strategy can adaptively adjust the step size at different stages to balance the convergence speed and stability.

[0137] Finally, through iteration until the comprehensive performance index reaches the threshold or reaches the maximum iteration stop, the wavefront correction (i.e. phase compensation) is completed, Figures 3 to 5 is a simulation result comparison chart under ideal conditions.

[0138] Step S24: Phase compensation is performed on the sea-air vortex beam based on the control vector to obtain a compensated vortex beam.

[0139] Step S3: Interference fusion is performed on the compensated vortex light and signal light to generate an interference image.

[0140] Specifically, the optical fiber delay line of the interference detection unit is adjusted, the interference fringe image is collected by the CCD camera, the fringe contrast is calculated, and when the contrast is ≥80%, the position of the optical fiber delay line is locked; the exposure time of the CCD camera is set to 20ms, the frame rate is set to 30 frames / second, and other parameters are set, and a standard fringe image is collected; during dynamic testing, under each environmental parameter gradient, the CCD camera continuously collects interference images according to the set parameters, and environmental data are recorded synchronously, and are transmitted to the data processing unit for storage through the GigE interface.

[0141] Step S4: The interference image is input into the improved EfficientNet-B0 model to obtain transmission quality data of the vortex light; the transmission quality data includes: topological charge number and degradation degree of OAM mode; the improved EfficientNet-B0 model adds a lightweight attention module and an early classification outlet to the EfficientNet-B0 model.

[0142] Specifically, as Figure 6 ​​As shown, the improved EfficientNet-B0 model adjusts the size of the intensity image from the initial size to the square size at the input layer, and the length-width pixel number should be an integer multiple of 32, because in the structure of the improved EfficientNet-B0 model, there are 5 times of down-sampling with a step of 2, so the total down-sampling multiple of the entire network is 2 5 =32, otherwise it will cause the network to be unable to be normally processed.

[0143] The improved EfficientNet-B0 model contains 9 stages (Stage):

[0144] Stage 1: initial convolutional layer, the 512x512 interference image is convolved with a 3x3 convolution kernel to obtain 32 feature images of 256x256, which mainly extracts basic edge and texture features.

[0145] Stage 2: the first MBConv module, the 16 256x256 images generated in this process are obtained by 3x3 convolution operation on the 32 feature images of the last step.

[0146] The following Stage2 to Stage8 are all repeated stacking of MBConv structure, and Stage9 is composed of a normal 1x1 convolution layer (including BN and activation function Swish), an average pooling layer and a fully connected layer. Among them, the pooling window size of the pooling layer is 16x16, and the 1280 nodes of the fully connected layer are connected with the pooling layer, and there are two output heads, finally output the topological charge number of vortex light and the degradation degree of OAM mode and the corresponding classification probability.

[0147] Among them, the lightweight attention module is embedded before the MBConv module of Stage3, Stage5, and Stage7, which is used to adaptively optimize and enhance the input feature map. In order to reduce the calculation amount of simple samples, three early classification exits are set after the network Stage3, Stage5, and Stage7. Each exit is composed of a global average pooling layer and a fully connected classifier, and there are two output heads, finally output the topological charge number of vortex light and the degradation degree of OAM mode and the corresponding classification probability.

[0148] As an implementable manner, the improved EfficientNet-B0 model comprises, in sequence, an initial convolutional layer, a first feature extraction module, a first early classification outlet, a second feature extraction module, a second early classification outlet, a third feature extraction module, a third early classification outlet, a first MBConv module, a convolutional layer, an average pooling layer and a fully connected layer; the first early classification outlet, the second early classification outlet and the third early classification outlet each consist of a global average pooling layer and a fully connected classifier; the first feature extraction module, the second feature extraction module and the third feature extraction module each consist of a lightweight attention module and two MBConv modules. Step S4 specifically comprises steps S41 to S46:

[0149] Step S41: inputting the interference image to the initial convolutional layer for feature extraction to obtain an initial feature map.

[0150] Step S42: inputting the initial feature map to the first feature extraction module for weighted feature extraction to obtain a first feature map.

[0151] Specifically, taking the first feature extraction module as an example, the connection order in the first feature extraction module is: MBConv module-lightweight attention module-MBConv module. Assuming that the input feature map of the lightweight attention module is , the output is a weighted feature map .

[0152] The calculation process of the module is as follows: .

[0153] wherein, is an attention weight vector in the channel dimension, , is a Sigmoid activation function, is a 1x1 convolutional layer, is a lightweight two-layer fully connected layer, is global average pooling.

[0154] Secondly, the spatial attention weight matrix is calculated as follows: .

[0155] wherein, is an attention weight matrix in the spatial dimension, , represents a 1x1 convolutional layer for compressing the channel number and fusing spatial information.

[0156] Finally, the weighted feature map after attention is output as follows: .

[0157] wherein, The element-wise multiplication is performed. With little computational overhead, the module enables the network to quickly focus on the features most relevant to the phase singularity and interference fringe structure of the vortex light, improving the signal-to-noise ratio of useful information and laying a foundation for fast and accurate judgment of subsequent levels.

[0158] Step S43: input the first feature map into the first early classification outlet to obtain first confidence and first vortex light transmission quality data; if the first confidence is greater than the confidence threshold, it is determined that the first vortex light transmission quality data is the vortex light transmission quality data; if the first confidence is less than or equal to the confidence threshold, the first feature map is input into the second feature extraction module for weighted feature extraction to obtain the second feature map.

[0159] Specifically, taking the first early classification outlet as an example, the input of the first early classification outlet is the first feature map , and the output is the first confidence and the first vortex light transmission quality data.

[0160] The classification probability vector of the output of the first feature map is: .

[0161] wherein, is the classification probability vector, is a Softmax activation function, which ensures that the sum of all output probabilities is 1, and the probability value of each class is between 0 and 1, is the weight matrix of the classifier of the first early outlet, is the bias vector of the first early outlet.

[0162] The prediction confidence of the outlet is defined as the maximum value of the probability vector: .

[0163] wherein, is the first confidence.

[0164] An adjustable confidence threshold is set. In the inference process, the data flows through the network in turn, and when it reaches the first early outlet, if the first confidence is greater than the confidence threshold, the output of the outlet is immediately taken as the final result, and the forward calculation of the subsequent network layer is terminated.

[0165] Due to the setting of the early classification outlet, the model can respond extremely fast when facing "simple" samples with slight degradation and obvious features; only for "complex" samples with severe degradation and difficult to judge, the entire model is called for calculation.

[0166] Step S44: Input the second feature map into the second early classification exit to obtain the second confidence level and the transmission quality data of the second vortex light; if the second confidence level is greater than the confidence threshold, then determine the transmission quality data of the second vortex light as the transmission quality data of the vortex light; if the second confidence level is less than or equal to the confidence threshold, then input the second feature map into the third feature extraction module for weighted feature extraction to obtain the third feature map.

[0167] Specifically, such as Figure 6 As shown, the prediction confidence of the second early classification exit is the maximum value of the probability vector: .in, This represents the second confidence level.

[0168] Step S45: Input the third feature map into the third early classification exit to obtain the third confidence level and the transmission quality data of the third vortex light; if the third confidence level is greater than the confidence level threshold, then determine the transmission quality data of the third vortex light as the transmission quality data of the vortex light; if the third confidence level is less than or equal to the confidence level threshold, then input the third feature map into the first MBConv module for weighted feature extraction to obtain the fourth feature map.

[0169] Specifically, such as Figure 6 As shown, the prediction confidence of the third early classification export is the maximum value of the probability vector: .in, This represents the third confidence level.

[0170] Step S46: The fourth feature map is sequentially processed through convolutional layers, average pooling layers, and fully connected layers for feature extraction, dimensionality reduction, and classification to obtain the transmission quality data of the vortex light.

[0171] Specifically, such as Figures 7 to 13 As shown, the interference fringes produced by the non-coaxial interference of a vortex beam and a plane wave are illustrated. Figures 7 to 9 This represents the interferometric images with topological charges of -3, -2, and -1, respectively. Figure 10 This represents the interferometric image when the topological charge number is 0. Figures 11 to 13 The images represent the interferometric images with topological charges of 1, 2, and 3, respectively.

[0172] Specifically, to obtain the training data of the improved EfficientNet-B0 model, data acquisition is needed, and the acquisition test operation is as follows: fix a certain set of parameters of the channel simulation unit 3, such as environmental parameters of salinity, temperature, and turbulence intensity, emit a vortex light beam with a known topological charge through the vortex light generating device 2, make it pass through the channel, and collect the interference image. At the same time, through theoretical simulation, the degradation degree of the OAM mode of the received light field at this time is simulated. The interference image is paired with the degradation degree label of the corresponding OAM mode, that is, a training sample is formed. By traversing different preset topological charges, channel parameters and environmental conditions, a large-scale training data set can be constructed for training the improved EfficientNet-B0 model.

[0173] In order to eliminate the non-uniformity of the light intensity of the image, the interference image intensity normalization formula is:

[0174] .

[0175] Among them, is the interference image intensity with pixel coordinates , is the original interference image intensity with pixel coordinates , is the mean value of the interference image intensity, is the standard deviation of the interference image intensity.

[0176] The complex amplitude is extracted by using the single-frame phase shift interference demodulation method:

[0177] .

[0178] Among them, is the complex amplitude extracted by using the single-frame phase shift interference demodulation method, is the two-dimensional Fourier transform, is the frequency domain window function of the band-pass filter, is the coordinate of the spatial frequency axis, is the coordinate of the spatial frequency axis.

[0179] The implicit phase information in the interference image is explicitly expressed and converted into physical features that can be used for training of the improved EfficientNet-B0 model, thereby improving the recognition accuracy and robustness of the vortex light beam OAM mode degradation degree.

[0180] The method for measuring the transmission quality of vortex light in a sea-air environment has the following beneficial effects: when the improved SPGD algorithm is used for phase compensation, the weights of the comprehensive performance index are dynamically and adaptively constructed based on the vortex light beam and the sea-air vortex light beam, so that the sea-air vortex light beam is more consistent with the vortex light beam during phase compensation, thereby improving the transmission quality of the vortex light beam. Finally, the improved EfficientNet-B0 model accurately tests the topological charge number (i.e., the identification of the OAM mode) and the OAM mode degradation degree (i.e., the comprehensive degradation of the mode purity and the beam quality) of the vortex light beam, thereby realizing accurate measurement of the transmission quality of the vortex light in the sea-air environment. The accurate identification of the topological charge number is the fundamental prerequisite for the OAM multiplexing communication system to realize error-free demodulation of information. The quantitative evaluation of the OAM mode degradation degree comprehensively reflects the systematic damage caused by the decrease in mode purity and wavefront distortion due to complex environmental factors, thereby providing quantitative data support for the actual system design of vortex light spatial laser communication.

[0181] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0182] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to a memory, a database or other medium used in the embodiments provided in the present application can include at least one of a non-volatile and a volatile memory. The non-volatile memory can include a read-only memory (ROM), a magnetic tape, a floppy disk, a flash memory, an optical storage, a high-density embedded non-volatile memory, a resistive random access memory (ReRAM), a magnetoresistive random access memory (MRAM), a ferroelectric random access memory (FRAM), a phase change memory (PCM), a graphene memory, etc. The volatile memory can include a random access memory (RAM) or an external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as a static random access memory (SRAM) or a dynamic random access memory (DRAM), etc.

[0183] The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a blockchain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0184] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist contradictory, they should be considered as the scope of the present application.

[0185] The principles and implementation modes of the present application are described by applying specific examples in the present application. The above-mentioned embodiments are only used to help understand the method and its core idea of the present application; meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range can be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A device for measuring the quality of vortex light transmission in a sea-air environment, characterized in that, The device for measuring the quality of vortex light transmission in the sea-air environment comprises an optical emission guide unit, a vortex light generating device, a channel simulation unit, an adaptive optical compensation unit, an interference detection unit and a data processing unit; the channel simulation unit is used for simulating the sea-air environment; The optical emission guide unit is used for emitting signal light; The vortex light generating device is used for converting the signal light into a vortex light beam; the vortex light beam generates a sea-air vortex light beam by passing through the sea-air environment; The adaptive optical compensation unit is used for dynamically constructing a comprehensive performance index based on the sea-air vortex light beam and the vortex light beam, and generating a control vector by using a modified SPGD algorithm based on the comprehensive performance index, so as to perform phase compensation on the sea-air vortex light beam and generate a compensated vortex light beam; The interference detection unit is used for performing interference fusion on the compensated vortex light and the signal light, so as to generate an interference image; The data processing unit is used for obtaining transmission quality data of the vortex light by using an improved EfficientNet-B0 model based on the interference image; the transmission quality data comprises a topological charge number and an OAM mode degradation degree; the improved EfficientNet-B0 model adds a lightweight attention module and an early classification exit to the EfficientNet-B0 model; The adaptive optical compensation unit comprises: A second beam splitting prism is arranged on an exit light path of the channel simulation unit and is used for splitting the sea-air vortex light beam into two sea-air vortex light beams; one sea-air vortex light beam is incident into a first CCD detector; the other sea-air vortex light beam is incident into a reflective deformable mirror; The first CCD detector is used for converting the sea-air vortex light beam into a first image; The reflective deformable mirror is used for performing phase compensation on the sea-air vortex light beam, so as to generate a compensated vortex light beam; A third beam splitting prism is arranged on an exit light path of the reflective deformable mirror and is used for splitting the compensated sea-air vortex light beam into two compensated vortex light beams; one compensated vortex light beam is incident into a second CCD detector; the other compensated vortex light beam is incident into the interference detection unit; The second CCD detector is used for converting the compensated vortex light beam into a second image; An optical compensation controller is connected with the data processing unit, the first CCD detector, the reflective deformable mirror and the second CCD detector respectively, is used for receiving a control instruction of the data processing unit, dynamically constructing a comprehensive performance index based on the sea-air vortex light beam and the vortex light beam, and generating a control vector by using a modified SPGD algorithm based on the comprehensive performance index, so as to perform phase regulation and control on the reflective deformable mirror; The adaptive optical compensation unit dynamically constructs a comprehensive performance index based on the sea-air vortex light beam and the vortex light beam, specifically comprising: Based on the vortex light beam, the vortex light field and the vortex ring area of the vortex light beam are determined; Based on the sea-air vortex light beam, the vortex light field and the phase of the sea-air vortex light beam are determined; calculating sea fog intensity and turbulence intensity based on the vortex light field of the vortex light beam, the vortex light field of the sea-air vortex light beam and the phase of the sea-air vortex light beam; calculating a reference weight based on the sea fog intensity and the turbulence intensity; normalizing the reference weight to obtain an adaptive weight; calculating a mode purity index based on the vortex light field of the sea-air vortex light beam; calculating an estimated topological charge based on the vortex light field of the vortex light beam; calculating a performance index of phase topological structure integrity based on the estimated topological charge; calculating a performance index of spatial distribution characteristics of light intensity of the beam based on the vortex ring area; normalizing the mode purity index, the performance index of phase topological structure integrity and the performance index of spatial distribution characteristics of light intensity of the beam to obtain a normalized mode purity index, a normalized performance index of phase topological structure integrity and a normalized performance index of spatial distribution characteristics of light intensity of the beam; constructing a comprehensive performance index based on the normalized mode purity index, the normalized performance index of phase topological structure integrity, the normalized performance index of spatial distribution characteristics of light intensity of the beam and the adaptive weight.

2. The device for measuring the quality of vortex light transmission in a sea-air environment according to claim 1, characterized in that, The optical emission guiding unit comprises, in sequence, a laser, a collimating and expanding mirror, a polarizer, a first beam-splitting prism and a mirror; The laser emitted by the laser first passes through the collimating and expanding mirror for expansion and collimation, and then transmits through the polarizer to obtain polarized light; the polarized light is split into two signal lights by the first beam-splitting prism; one of the signal lights is incident into the vortex light generating device; the other signal light is incident into the interference detection unit through the mirror.

3. The device for measuring the quality of vortex light transmission in a sea-air environment according to claim 2, characterized in that The vortex light generating device comprises: A spatial light modulation controller connected with the spatial light modulator and the data processing unit respectively, used for receiving the control instruction of the data processing unit and controlling the spiral phase pattern of the spatial light modulator; The spatial light modulator is used for converting the signal light into a vortex light beam through the spiral phase pattern.

4. The device for measuring the quality of vortex light transmission in a sea-air environment according to claim 1, characterized in that, The channel simulation unit comprises a sea fog box and a turbulence pool; the sea fog box is used for generating simulated sea fog in a sea-air environment; and the turbulence pool is used for generating turbulence in the sea-air environment.

5. A method for measuring the quality of vortex light transmission in a sea-air environment, characterized in that, The measurement method of vortex light transmission quality in a sea-air environment is applied to the measurement device of vortex light transmission quality in a sea-air environment in claims 1-4, and comprises: acquiring signal light and converting the signal light into a vortex light beam, wherein the vortex light beam generates a sea-air vortex light beam by passing through the sea-air environment; constructing a comprehensive performance index based on the sea-air vortex light beam and the vortex light beam, and generating a control vector by using a modified SPGD algorithm based on the comprehensive performance index to perform phase compensation on the sea-air vortex light beam and generate a compensated vortex light beam; performing interference fusion on the compensated vortex light and the signal light to generate an interference image; The interference image is input into an improved EfficientNet-B0 model to obtain transmission quality data of the vortex light; the transmission quality data includes a topological charge number and a degradation degree of an OAM mode; the improved EfficientNet-B0 model adds a lightweight attention module and an early classification outlet to the EfficientNet-B0 model.

6. The method according to claim 5, wherein Based on the sea-air vortex light beam and the vortex light beam, a comprehensive performance index is dynamically constructed, and based on the comprehensive performance index, a control vector is generated by using a modified SPGD algorithm to perform phase compensation on the sea-air vortex light beam to generate a compensated vortex light beam, specifically including: Based on the sea-air vortex light beam and the vortex light beam, a comprehensive performance index is dynamically constructed; An initial control vector is obtained; Based on the vortex light field of the sea-air vortex light beam, the initial control vector is iterated multiple times by using a modified SPGD algorithm until the comprehensive performance index reaches a threshold value or a maximum iteration round is stopped to generate a control vector; Based on the control vector, phase compensation is performed on the sea-air vortex light beam to obtain a compensated vortex light beam.

7. The method according to claim 6, wherein The calculation formula of the reference weight is: ； in, As the benchmark weight for the model purity index, As the initial weights, The first turbulence sensitivity coefficient, For the first Turbulence intensity estimation at the next iteration The first sea fog sensitivity coefficient, For the first Sea fog intensity estimation at the next iteration The benchmark weight for the performance index of phase topology integrity, The initial weights for the performance index of phase topology integrity, The second turbulence sensitivity coefficient, The second sea fog sensitivity coefficient, The benchmark weights for performance indicators representing the spatial distribution characteristics of beam intensity. The initial weights for the performance index of the spatial distribution characteristics of the beam intensity are given. This is the third turbulence sensitivity coefficient.

8. The method according to claim 5, wherein The improved EfficientNet-B0 model includes an initial convolution layer, a first feature extraction module, a first early classification outlet, a second feature extraction module, a second early classification outlet, a third feature extraction module, a third early classification outlet, a first MBConv module, a convolution layer, an average pooling layer, and a full connection layer connected in sequence; the first early classification outlet, the second early classification outlet, and the third early classification outlet each consist of a global average pooling layer and a full connection classifier; the first feature extraction module, the second feature extraction module, and the third feature extraction module each consist of a lightweight attention module and two MBConv modules; The interference image is input into an improved EfficientNet-B0 model by a data processing unit to obtain transmission quality data of the vortex light, specifically including: The interference image is input into the initial convolution layer for feature extraction to obtain an initial feature map; The initial feature map is input into the first feature extraction module for weighted feature extraction to obtain a first feature map; The first feature map is input into the first early classification outlet to obtain a first confidence and first transmission quality data of the vortex light; if the first confidence is greater than a confidence threshold, the first transmission quality data of the vortex light is determined as the transmission quality data of the vortex light; if the first confidence is less than or equal to the confidence threshold, the first feature map is input into the second feature extraction module for weighted feature extraction to obtain a second feature map; The interference image is input into the initial convolution layer for feature extraction to obtain an initial feature map; inputting the second feature map into the second early classification outlet to obtain a second confidence and transmission quality data of the second vortex light; if the second confidence is greater than a confidence threshold, determining that the transmission quality data of the second vortex light is the transmission quality data of the vortex light; if the second confidence is less than or equal to the confidence threshold, inputting the second feature map into the third feature extraction module to perform weighted feature extraction to obtain a third feature map; inputting the third feature map into the third early classification outlet to obtain a third confidence and transmission quality data of the third vortex light; if the third confidence is greater than the confidence threshold, determining that the transmission quality data of the third vortex light is the transmission quality data of the vortex light; if the third confidence is less than or equal to the confidence threshold, inputting the third feature map into the first MBConv module to perform weighted feature extraction to obtain a fourth feature map; inputting the fourth feature map through the convolution layer, the average pooling layer and the full connection layer in sequence to perform feature extraction, dimension reduction and classification processing to obtain the transmission quality data of the vortex light.

Images