CN113065494B - A system, method, device and electronic device for vortex electronic mode recognition - Google Patents

Abstract

The invention discloses a vortex electronic mode recognition system, method, device and electronic equipment. The system includes a vortex electronic generation module, a diffraction amplification module, an image receiving and acquisition module, a data processing and recognition module, and a training data preparation sub-module. , artificial intelligence model training sub-module and modal recognition module. The cyclotrons generated by the vortex electron generation module are coupled with orbital angular momentum to generate vortex electron beams carrying different modes of OAM; the diffraction amplification module diffracts and amplifies the vortex electron beams; the image receiving and acquisition module is used to receive the diffraction amplified vortices The OAM field intensity distribution image is obtained by collecting the electron beam and the data processing and identification module receives the OAM field intensity distribution image and performs OAM modal identification. The data processing and identification module not only realizes the vortex electron orbital angular momentum modal identification. , which can reduce the influence of nonlinear distortion on the recognition results, the recognition accuracy is high, and the hardware complexity and cost are very low.

Description

A system, method, device and electronic device for vortex electronic mode recognition

technical field

The invention belongs to the technical field of artificial intelligence and Orbital Angular Momentum (OAM) wireless communication of electromagnetic waves, and particularly relates to a vortex electronic mode identification system, method, device and electronic equipment.

Background technique

Compared with conventional electromagnetic wave communication, the introduction of OAM can effectively increase the transmission capacity and spectrum efficiency, alleviate the shortage of spectrum resources, and realize high-capacity and high-speed transmission. Optical OAM communication systems in free space face the characteristics of low integration of OAM beam generators, and light waves are greatly affected by environmental factors such as atmospheric turbulence and haze. The OAM beam of radio frequency electromagnetic waves (below 300 GHz) has a large divergence angle, making it difficult to achieve long-distance transmission, and the OAM antenna at the receiving end has a complex structure, is difficult to integrate, and has a high cost. Different from the above statistical state OAM beam, the use of quantum state OAM for communication can well solve the above problems.

Quantum state OAM realizes information transmission through orbital angular momentum carried by vortex electrons. Reference (Zhang, C., Xu, P., & Jiang, X. (2020). Vortex electron generated by microwave photon with orbitalangular momentum in a magnetic field .AIPAdvances, 10, 105230.) proposed that, based on the quantum electrodynamics (Quantum Electrodynamics, QED) theory, the use of relativistic gyrotrons or multi-electrons at the transmitting end can directly radiate electromagnetic waves carrying OAM, and then use relativistic electrons at the receiving end to absorb OAM The quantum state electromagnetic waves become vortex electrons, and information recovery is achieved by identifying the OAM modes of the vortex electrons. The literature (Zhang, C., Xu, P. & Jiang, X. Detecting superposed orbital angular momentum states in the magnetic field by the crystal diffraction. Eur. Phys. J. Plus 136, 60 (2021).) gives a corresponding Vortex electron OAM mode identification method: the vortex electrons are diffracted by gold foil to amplify the size of the beam, and then the different modal OAM electrons are separated into different positions by the mapper and sorting device, and then the information is recovered by the method of position judgment. . In practical applications, the magnetic field generated by the gyrotron at the receiving end cannot be cut off directly, but gradually attenuates between the diffractive gold foil and the phosphor screen. Therefore, the vortex electrons will be affected by the residual magnetic field and will be offset and severely nonlinearly distorted. The device-based method based on the ideal magnetic field modeling is not ideal for OAM mode separation. In addition, an image acquisition device, such as a CCD (Charge Coupled Device, Charge Coupled Device) camera, can also be used to collect the diffraction image of the electronic OAM, and then the OAM mode is recognized and the information is recovered by image recognition, but the conventional image recognition method only It can extract the surface features such as the edge of the image, and cannot deal with the nonlinear distortion caused by the residual magnetic field. Table 1 compares the above-mentioned common vortex electron OAM mode identification methods, mainly two types of device-based methods and image processing methods. Both methods have their own advantages and disadvantages, but neither can resist vortex electron diffraction well. Nonlinear distortions that occur in an image. Therefore, it is urgent to propose a new OAM electronic modal recognition system that can cope with the influence of nonlinear distortion of magnetic field and ensure high recognition accuracy and low hardware complexity.

Table 1: Comparison of vortex electron OAM mode identification methods

SUMMARY OF THE INVENTION

In order to overcome the above-mentioned deficiencies of the prior art, the present invention provides a vortex electronic mode recognition system based on artificial intelligence. The network performs deep feature extraction. The system can realize vortex electron modal detection under the premise of low cost and low hardware complexity, and can still ensure high recognition accuracy and high recognition rate when the image is severely distorted. It can be used in OAM quantum state wireless communication system. Reduce the modal crosstalk caused by the defect of the receiver's judgment method.

A vortex electronic mode identification system, comprising:

The vortex electron generation module is used to generate swirling electrons and couple with the orbital angular momentum of the received OAM electromagnetic waves to generate vortex electron beams carrying different modes of OAM;

a diffractive amplifying module for diffractively amplifying the size of the vortex electron beam;

The image receiving and collecting module is used to receive the diffraction-amplified vortex electron beam and collect the OAM field intensity distribution image;

A data processing and identification module, after being connected to the image receiving and collecting module, is used to receive the OAM field intensity distribution image and perform data processing and OAM modal identification, and the data processing and identification module includes:

The training data preparation sub-module is used to generate the OAM field strength distribution image containing nonlinear distortion affected by the residual magnetic field as the training data of the artificial intelligence model training sub-module, wherein the residual magnetic field refers to the generated magnetic field generated by the vortex electron generation module. The magnetic field that gradually decays between the diffraction amplifying module and the image receiving and acquisition module;

The artificial intelligence model training sub-module is used to construct an artificial intelligence model, and use the training data to train the artificial intelligence model, so as to obtain the artificial intelligence model with the highest verification and recognition accuracy;

The modal recognition sub-module uses the artificial intelligence model with the highest verification and recognition accuracy to perform real-time OAM modal recognition on the OAM field intensity distribution image collected by the image receiving and acquisition module.

Optionally, the Laguerre-Gaussian vortex electron beam is used to simulate the vortex electron beam carrying different modes of OAM along the z-axis, and its expression is:

是携带OAM的涡旋电子波束的波函数，

是由径向量子数n和OAM模态数l决定的常数，

是广义拉盖尔多项式，

是束腰半径，其中

是约化普朗克常数，e是元电荷，B是电子波束外部环境的磁感应强度，i是虚数单位，

是柱坐标系的三个坐标变量，k是波数。in

is the wave function of the vortex electron beam carrying the OAM,

is a constant determined by the radial quantum number n and the OAM mode number l,

is the generalized Laguerre polynomial,

is the waist radius, where

is the reduced Planck constant, e is the elementary charge, B is the magnetic induction intensity of the external environment of the electron beam, i is the imaginary unit,

are the three coordinate variables of the cylindrical coordinate system, and k is the wave number.

Optionally, the vortex electron beam generated by the vortex electron generation module includes a vortex electron beam carrying pure OAM, and/or a plurality of vortex electron beams of mixed state OAM, wherein any mixed state OAM is A combination of multiple pure OAMs.

Optionally, the vortex electron generating module includes a power source, a gyrotron, and a superconducting magnet, wherein the gyrotron includes an electron gun and an electron cyclotron generating module, the superconducting magnet is located outside the electron cyclotron generating module, the power source supplies power to the electron gun, and the electron gun emits and emits electricity. The accelerated electrons enter the electron cyclotron generation module and perform cyclotron motion under the action of the magnetic field of the superconducting magnet. The OAM electromagnetic wave couples its orbital angular momentum to the cyclotron electrons, thereby generating vortex electron beams carrying different modes of OAM;

The diffractive amplifying module includes a polygold crystal film that diffracts and amplifies the vortex electron beam;

The image receiving and collecting module includes a phosphor screen and a camera, the phosphor screen is used for receiving the vortex electron beams amplified by diffraction, and the camera is used for collecting an image on the phosphor screen as an OAM field intensity distribution image.

Optionally, the training data includes OAM field strength distribution images affected by different degrees of residual magnetic field, wherein the nonlinear distortion caused by the influence of the residual magnetic field refers to the nonlinear divergence and nonlinearity of the OAM field strength distribution image. One or more of the cases in convergence.

Optionally, the nonlinear divergence and nonlinear convergence refer to that the distances between two points before and after the electrons are distorted by the residual magnetic field and the electron cyclotron center O satisfy the following relationship:

where r and r' are the distance of the electron before and after the distortion from the electron cyclotron center O in cylindrical coordinates, R is the radius of the electron cyclotron motion, β is the central angle of the electron cyclotron motion, and the position of the electron before and after the distortion is relative to the electron cyclotron The deflection angle of the center O is given by:

α=arcsin(2R[sin(β/2)] ² /r′ ² ).

Optionally, the artificial intelligence model includes a combination of one or more of convolutional neural network, decision tree, multi-class support vector machine (SVM) and K-nearest neighbor algorithm.

The present invention also provides a vortex electronic mode identification method, comprising the following steps:

After receiving the OAM electromagnetic wave propagating in free space, the vortex electron generating module couples the orbital angular momentum of the OAM electromagnetic wave to the cyclotron to generate vortex electron beams carrying different modes of OAM;

The diffraction amplification module performs diffraction amplification on the vortex electron beam;

The image receiving and collecting module includes a phosphor screen and a camera, the phosphor screen receives the diffracted and amplified vortex electron beam to obtain an OAM field intensity distribution image, and the camera collects the OAM field intensity distribution image in real time and sends it to the data processing and identification module;

The data processing and identification module uses the artificial intelligence model with the highest accuracy to identify the received OAM field intensity distribution image in real time, and complete the OAM modal identification.

The training data preparation sub-module of the data processing and identification module performs binarization data processing on multiple OAM field strength distribution images that are affected by the residual magnetic field and contains nonlinear distortion, and then sends them to artificial intelligence as part of the model training data. The model training sub-module, the training data preparation sub-module also generates a part of the distorted and/or non-distorted simulated OAM field strength distribution images through simulation as another part of the model training data, the part of the model training data and/or the other part of the model training data The data together form the training data,

The artificial intelligence model training sub-module uses the training data to train the artificial intelligence model, thereby obtaining the artificial intelligence model with the highest verification and recognition accuracy.

The present invention also provides an OAM modal identification device, comprising:

The training data preparation sub-module is used to generate the OAM field strength distribution image containing nonlinear distortion affected by the residual magnetic field as the training data of the artificial intelligence model training sub-module, wherein the residual magnetic field refers to the generated magnetic field generated by the vortex electron generation module. The magnetic field that gradually decays between the diffraction amplifying module and the image receiving and acquisition module;

The artificial intelligence model training sub-module is used to construct an artificial intelligence model, and use the training data to train the artificial intelligence model, so as to obtain the artificial intelligence model with the highest verification and recognition accuracy;

The modal recognition sub-module uses the artificial intelligence model with the highest verification and recognition accuracy to perform real-time OAM modal recognition on the OAM field intensity distribution image collected by the image receiving and acquisition module.

The present invention also provides an electronic device, the electronic device comprising:

at least one processor; and,

a memory communicatively coupled to the at least one processor; wherein,

The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of:

The multiple OAM field strength distribution images that are affected by the residual magnetic field and contain nonlinear distortion received in advance are processed into binarized data and sent to the artificial intelligence model training sub-module as part of the model training data. The training data preparation sub-module is also simulated by simulation. generating a part of the distorted and/or non-distorted simulated OAM field strength distribution images as another part of the model training data, and the part of the model training data and/or the other part of the model training data together constitute the training data;

Use the training data to perform artificial intelligence model training, thereby obtaining the artificial intelligence model with the highest verification and recognition accuracy;

Use the artificial intelligence model with the highest accuracy to identify the received OAM field intensity distribution images in real time, and complete the OAM modal identification.

Compared with the prior art, the beneficial effects of the present invention are: the vortex electron OAM modal identification system of the present invention does not require too many devices, the hardware complexity and cost are very low, and the residual magnetic field and other causes can be reduced. The influence of nonlinear distortion on the recognition results, the recognition accuracy is high, and it can be used in many application scenarios such as quantum state OAM wireless communication.

Description of drawings

The above and/or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of embodiments taken in conjunction with the accompanying drawings, wherein:

1 is a schematic structural diagram of a vortex electron mode identification system according to an embodiment of the present invention;

2 is a schematic diagram of a system hardware structure according to a specific embodiment of the present invention;

3 is a geometrical relationship diagram of the position before and after electronic distortion according to a specific embodiment of the present invention;

Figure 4(a) is the OAM field strength distribution of pure OAM (l=1) before being distorted by the residual magnetic field;

Figure 4(b) is the OAM field strength distribution diagram of the convergent pure state OAM (l=1) under the influence of the residual magnetic field;

Figure 4(c) is the OAM field strength distribution diagram of the pure OAM (l=1) with divergent distortion after being affected by the residual magnetic field;

Figure 5(a1)-(d1) are the electrons whose OAM modes are l={1,{1,-2},{3,-3},{1,2,3}} before being distorted by the residual magnetic field The field strength distribution of the beam;

Figure 5(a2)-(d2) are the electron beams with OAM modes l={1,{1,-2},{3,-3},{1,2,3}} after being affected by the residual magnetic field, respectively Field strength distribution map with convergence distortion;

Figure 5(a3)-(d3) are the electron beams with OAM modes l={1,{1,-2},{3,-3},{1,2,3}} after being affected by the residual magnetic field, respectively Field strength distribution map with divergent distortion;

Figure 6(a) is the OAM field strength distribution diagram of the modal number l={3,-3} when there is no noise (signal-to-noise ratio is +∞);

Figure 6(b)-(d) are the OAM field strength distribution diagrams of the modal number l={3,-3} when the SNR is 30, 20 and 15 after adding white Gaussian noise;

7 is a convolutional neural network structure according to a specific embodiment of the present invention;

8 is a loss function value curve of a neural network according to a specific embodiment of the present invention;

FIG. 9 is an accuracy curve of a neural network according to an embodiment of the present invention.

Detailed ways

The following describes in detail the embodiments of the present invention, examples of which are illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary, and are intended to explain the present invention and should not be construed as limiting the present invention.

As shown in FIG. 1 , a vortex electron mode identification system 10 includes a vortex electron generation module 100 , a diffraction amplification module 200 , an image receiving and collecting module 300 , and a data processing and identifying module 400 .

The vortex electron generating module 100 includes a high-voltage power supply 101, a gyrotron 102 and a superconducting magnet 103, wherein the gyrotron 102 is mainly composed of a high-speed electron gun 1021 and a cylindrical electron cyclotron generating module. The high-voltage power supply 101 supplies power to the electron gun 1021 and provides high-voltage direct current required for electron acceleration. The electrons are emitted and accelerated by the electron gun 1021; High-speed cyclotron motion, after the OAM electromagnetic wave propagating in free space is irradiated on the high-speed cyclotron, its orbital angular momentum is coupled to the cyclotron, that is, a vortex electron beam carrying different modes of OAM is generated; the diffraction amplification module 200 uses a polygold crystal film 201 diffractively amplifies the vortex electron beam; the image receiving and collecting module 300 is located behind the diffractive amplifying module, and includes a phosphor screen 301 and a high-speed camera 302 . The fluorescent screen 301 is used to receive the vortex electron beams amplified by diffraction; the high-speed camera 302 is located behind the fluorescent screen and is used to collect the OAM field intensity distribution image (ie, the OAM electric field intensity distribution image) on the fluorescent screen; the data processing and identification module 400 is: A computer built with a convolutional neural network includes the following submodules: a training data preparation submodule 401 for generating training data for the neural network; an artificial intelligence model training module 402 for constructing an artificial intelligence model, such as a convolutional neural network The modal identification module 403 inputs the collected OAM field intensity distribution image into the trained convolutional neural network for OAM modal identification. The positional relationship of each module in the system is shown in the schematic diagram of the system hardware structure in FIG. 2 .

This system can be used as the receiving end of quantum state OAM wireless communication. The transmitting end transmits information by emitting OAM electromagnetic waves of different modal numbers. The receiving end receives the OAM electromagnetic waves emitted by the transmitting end, and the received OAM modal After coupling to the vortex electrons, diffraction amplification, and data processing to realize the identification of the OAM mode, the original information sent can be recovered. The values of the relevant parameters involved in the system in this example are shown in Table 2.

Table 2: List of relevant parameters

In this embodiment, the vortex electron beam carrying OAM generated by the vortex electron generating module 100 can be simulated by a Laguerre-Gaussian (LG) vortex electron beam, and the vortex electron beam along the z-axis direction (with the vortex electron beam) The LG vortex electrons with the same axial direction of the central axis of rotation) can be described as

是携带OAM的涡旋电子波束的波函数，

是由径向量子数n和OAM模态数l决定的常数，

是广义拉盖尔多项式，

是束腰半径，其中

是约化普朗克常数，e是元电荷，B是电子波束外部环境(即回旋管外的超导磁体)的磁感应强度，i是虚数单位，

是柱坐标系的三个坐标变量，k是波数。in

is the wave function of the vortex electron beam carrying the OAM,

is a constant determined by the radial quantum number n and the OAM mode number l,

is the generalized Laguerre polynomial,

is the waist radius, where

is the reduced Planck constant, e is the elementary charge, B is the magnetic induction of the external environment of the electron beam (ie, the superconducting magnet outside the gyrotron), i is the imaginary unit,

are the three coordinate variables of the cylindrical coordinate system, and k is the wave number.

In this embodiment, the residual magnetic field is along the z-axis direction, and the magnitude is B _z decays along the z-axis, which is a constant magnetic field. Affected by the Lorentz force, the electrons move in a spiral, that is, in addition to moving along the z-axis, they will also make a counterclockwise circular motion along a plane perpendicular to the z-axis. Since the diffracted exit angles of vortex electrons in different modes are different, the radius of gyration in the residual magnetic field is also different, and the radius of gyration R can be calculated by the following formula:

where γ is the Lorentz factor of the electron, me is the mass of the electron, _e is the electron elementary charge, v is the velocity of the electron before diffraction, and θ is the exit angle of the electron after diffraction. Since the exit angle θ is very small, the residual magnetic field is relatively large at the beginning compared with before diffraction, so the gyration radius of the electrons is small. Since the residual magnetic field decays along the z-axis direction, the electron gyration radius will gradually increase.

Under the action of the residual magnetic field, the cyclotron motion of the electrons causes the OAM field intensity distribution image collected by the image receiving and acquisition module 300 to be distorted under the action of the residual magnetic field, that is, the electrons diverge or converge along the radial direction under the cylindrical coordinates, and the electrons are distorted. The geometric relationship between the front and rear positions is shown in Figure 3, where point A and point A' are the position after electron diffraction and the position on the phosphor screen after being distorted by the magnetic field. :

Among them, r and r' are the distances of the electrons before and after the distortion from the electron cyclotron center in cylindrical coordinates, R is the radius of the electron cyclotron motion, and β is the central angle of the electron cyclotron motion. In addition, after the distortion occurs, the relative position falling on the phosphor screen will be deflected by an angle in the counterclockwise direction, and the deflection angle of the electron position before and after the distortion relative to the central axis O of the electron gyration can be obtained by the following formula

α=arcsin(2R[sin(β/2)] ² /r' ² ) (4)

For the pure OAM of a single mode, the field strength distribution diagram of the OAM can be used to represent the influence of the residual magnetic field on the pure OAM vortex electrons (taking l=1 as an example), in which Fig. 4(a) is distorted by the residual magnetic field. Figure 4(b) is the OAM field strength distribution with convergence distortion after being affected by the residual magnetic field, and Figure 4(c) is the OAM field strength distribution with divergent distortion after being affected by the residual magnetic field. The central hole of the distorted image becomes smaller or larger, which causes modal blurring between different pure OAM modes, causing the receiver to fail to recognize it. The vortex electrons generated by the vortex electron generating module in the present invention may be vortex electrons carrying any pure state OAM, and/or vortex electrons carrying mixed state OAMs, where the mixed state OAM refers to a combination of multiple pure state OAMs . For example, the following is the field strength distribution diagram of 1 pure state OAM and 3 mixed state OAM, the OAM mode number l={1,{1,-2},{3,-3},{1,2, 3}}, where l=1 is pure OAM, and l={1,-2}, {3,-3}, {1,2,3} are three mixed OAMs, respectively.

Figure 5(a1)-(d1) is the field strength distribution of different electron beams before being distorted by the residual magnetic field, and Figure 5(a2)-(d2) is the field strength distribution of different electron beams with convergence distortion after being affected by the residual magnetic field , Figures 5(a3)-(d3) are the field strength distribution diagrams of divergent distortion of different electron beams under the influence of the residual magnetic field. In this example, as the receiving end of quantum state OAM wireless communication, in order to ensure a low bit error rate, the transmitting end uses a pure state and a mixed state OAM with a large difference between the three types of field intensity distribution diagrams for data transmission. , its LG modal number l={1,{1,-2},{3,-3},{1,2,3}}, which respectively represent quaternary numbers 0-3, that is, the communication of fourth-order modulation system.

In this embodiment, the training data preparation sub-module generates distortion data of various residual magnetic field levels as training data, and introduces positional Gaussian noise of different signal-to-noise ratio levels, and the proportion of noisy data sets at each level is the same, as shown in the figure Shown in 6 are the OAM field strength distributions for ±3 modes with no noise and signal-to-noise ratio (SNR) of 30, 20, and 15, respectively. In order to verify the ubiquity of the training model, the data in the validation set are distorted data of different degrees that are not in the training set.

In this embodiment, the AI (artificial intelligence) model specifically adopts a convolutional neural network, the input data is an OAM field intensity distribution image with a size of 128×128, and the corresponding modal category is output. The 4 modalities in this example correspond to 4 categories. Preferably, one-hot encoding is used for the modality categories.

The structure of the convolutional neural network used in this embodiment is shown in Figure 7, which includes an input layer, 5 convolutional pooling layers, 1 fully connected layer, and an output layer connected in sequence, wherein each convolutional pooling layer includes The convolutional layer and the pooling layer are connected in turn. The convolutional pooling layer performs feature learning and extraction on the OAM field intensity distribution image. In addition to the features such as shape, texture, color, and edge, it can also be extracted after a large amount of data training. Features of nonlinear distortion introduced by residual magnetic fields, i.e. deeper, dataset-specific feature representations. These hidden features are more efficient and accurate in expressing the dataset, and the extracted abstract features are more robust, so that higher recognition accuracy can be obtained. As shown in Figure 7, each convolutional layer contains multiple 5×5 convolution kernels, for example, the first convolutional layer contains 20 5×5 convolution kernels. The max pooling layer contains a 2×2 pooling window.

The cross entropy is used to represent the degree of unfitness of the currently trained convolutional neural network to the training data, that is, the cross entropy loss function, whose expression is

Where X is the input data of the neural network, that is, the image data of the OAM field intensity distribution, n is the number of input data samples, f( ) represents the OAM modal category output by the neural network, and Y represents the real OAM modal category.

The change curves of the loss function value and accuracy rate of the convolutional neural network in this embodiment with the number of training times are shown in Figure 8 and Figure 9. 1000 OAM field strength distribution images are trained and verified, and the network parameters are updated to make the loss function as much as possible. get the minimum value.

The adaptive moment estimation (Adam) optimizer is used to obtain the gradient of the loss function during the training process, so that the network parameter value is updated through the gradient, so that the loss function continues to iterate toward the minimum value. The batch size of the optimizer is 32, and the accuracy of the artificial intelligence model after training 30 times is 99.46% on the training set and 99.99% on the validation set. For the conventional method, that is, the general convolutional neural network, the influence of the residual magnetic field on the OAM field strength distribution is not considered, and the training parameters such as the same network structure, loss function and optimization method as in this example are used, but the training data does not contain Distorted samples, the AI model obtained from training and the AI model proposed in this example are tested on the same test data set. The recognition accuracy of the method of the example of the present invention and the conventional method are 95.75% and 69.75% respectively, so The identification method of the example described in the present invention has obvious advantages in identification accuracy.

The above uses convolutional neural network as an example of AI model. In fact, the AI model can also be one of convolutional neural network, decision tree, multi-class support vector machine (SVM) and K-nearest neighbor algorithm. various combinations. Specifically, the ensemble learning technology is used to integrate these models for image recognition and classification, that is, to train different weak classifiers for the same training set, and then combine these weak classifiers to form a stronger strong classifier as a The final classifier, the commonly used combination methods are boosting and adaboost.

Although a specific embodiment of the present invention has been given above, it should be understood that the above-mentioned embodiment is exemplary and cannot be used as a limitation of the present invention. Variations, modifications, substitutions and alterations are made to the above-described embodiments.

Claims (8)

1. a vortex electronic mode identification system, is characterized in that, comprises: The vortex electron generation module is used to generate swirling electrons and couple with the orbital angular momentum of the received OAM electromagnetic waves to generate vortex electron beams carrying different modes of OAM; a diffractive amplifying module for diffractively amplifying the size of the vortex electron beam; The image receiving and collecting module is used to receive the diffraction-amplified vortex electron beam and collect the OAM field intensity distribution image; A data processing and identification module, after being connected to the image receiving and collecting module, is used to receive the OAM field intensity distribution image and perform data processing and OAM modal identification, and the data processing and identification module includes: The training data preparation sub-module is used to generate the OAM field strength distribution image containing nonlinear distortion affected by the residual magnetic field as the training data of the artificial intelligence model training sub-module, wherein the residual magnetic field refers to the generated magnetic field generated by the vortex electron generation module. The magnetic field that gradually decays between the diffraction amplifying module and the image receiving and acquisition module; The artificial intelligence model training sub-module is used to construct an artificial intelligence model, and use the training data to train the artificial intelligence model, so as to obtain the artificial intelligence model with the highest verification and recognition accuracy; The modal recognition sub-module uses the artificial intelligence model with the highest verification and recognition accuracy to carry out real-time OAM modal recognition on the OAM field strength distribution images collected by the image receiving and acquisition module, and the training data includes residual magnetic fields affected by different degrees. The OAM field strength distribution image, wherein the nonlinear distortion caused by the influence of the residual magnetic field refers to one or more of the nonlinear divergence and nonlinear convergence of the OAM field strength distribution image, The nonlinear divergence and nonlinear convergence refer to that the distances between the two points before and after the electron is distorted by the residual magnetic field and the electron gyration center O satisfy the following relationship:

where r and r' are the distance of the electron before and after the distortion from the electron cyclotron center O in cylindrical coordinates, R is the radius of the electron cyclotron motion, β is the central angle of the electron cyclotron motion, and the position of the electron before and after the distortion is relative to the electron cyclotron The deflection angle of the center O is given by: α=arcsin(2R[sin(β/2)] ² /r′ ² ). 2. a kind of vortex electron mode identification system according to claim 1, it is characterized in that, use Laguerre-Gaussian vortex electron beam to simulate the vortex electron beam that carries different modal OAM along z-axis direction, its The expression is:

in

is the wave function of the vortex electron beam carrying the OAM,

is a constant determined by the radial quantum number n and the OAM mode number l,

is the generalized Laguerre polynomial,

is the waist radius, where

is the reduced Planck constant, e is the elementary charge, B is the magnetic induction intensity of the external environment of the electron beam, i is the imaginary unit,

are the three coordinate variables of the cylindrical coordinate system, and k is the wave number. 3. A vortex electron mode identification system according to claim 1, wherein the vortex electron beam generated by the vortex electron generating module comprises a vortex electron beam carrying pure OAM, and/or Vortex electron beams of multiple mixed-state OAMs, wherein any mixed-state OAM is composed of multiple pure-state OAMs. 4. a kind of vortex electron mode identification system according to claim 1, is characterized in that, The vortex electron generating module includes a power supply, a gyrotron, and a superconducting magnet, wherein the gyrotron includes an electron gun and an electron cyclotron generating module, the superconducting magnet is located outside the electron cyclotron generating module, the power supply supplies power to the electron gun, and the electrons emitted and accelerated by the electron gun enter The electron cyclotron generation module performs cyclotron motion under the action of the magnetic field of the superconducting magnet, and the OAM electromagnetic wave couples its orbital angular momentum to the cyclotron electrons, thereby generating vortex electron beams carrying different modes of OAM; The diffractive amplifying module includes a polygold crystal film that diffracts and amplifies the vortex electron beam; The image receiving and collecting module includes a phosphor screen and a camera, the phosphor screen is used for receiving the vortex electron beams amplified by diffraction, and the camera is used for collecting an image on the phosphor screen as an OAM field intensity distribution image. 5. a kind of vortex electronic modal identification system according to claim 1, is characterized in that, described artificial intelligence model comprises one or in convolutional neural network, decision tree, multi-classification support vector machine and K nearest neighbor algorithm. various combinations. 6. a vortex electron mode identification method, is characterized in that, comprises the steps: After receiving the OAM electromagnetic wave propagating in free space, the vortex electron generating module couples the orbital angular momentum of the OAM electromagnetic wave to the cyclotron to generate vortex electron beams carrying different modes of OAM; The diffraction amplification module performs diffraction amplification on the vortex electron beam; The image receiving and collecting module includes a phosphor screen and a camera, the phosphor screen receives the diffracted and amplified vortex electron beam to obtain an OAM field intensity distribution image, and the camera collects the OAM field intensity distribution image in real time and sends it to the data processing and identification module; The data processing and identification module uses the artificial intelligence model with the highest accuracy to identify the received OAM field intensity distribution image in real time, and complete the OAM modal identification. The training data preparation sub-module of the data processing and identification module performs binarization data processing on multiple OAM field strength distribution images that are affected by the residual magnetic field and contains nonlinear distortion, and then sends them to artificial intelligence as part of the model training data. The model training sub-module, the training data preparation sub-module also generates a part of the distorted and/or non-distorted simulated OAM field strength distribution images through simulation as another part of the model training data, the part of the model training data and/or the other part of the model training data The data together form the training data, The artificial intelligence model training sub-module uses the training data to perform artificial intelligence model training, so as to obtain the artificial intelligence model with the highest verification and recognition accuracy, The training data includes OAM field strength distribution images affected by different degrees of residual magnetic field, wherein the nonlinear distortion caused by the influence of the residual magnetic field refers to one of nonlinear divergence and nonlinear convergence of the OAM field strength distribution image. one or more circumstances, The nonlinear divergence and nonlinear convergence refer to that the distances between the two points before and after the electron is distorted by the residual magnetic field and the electron gyration center O satisfy the following relationship:

where r and r' are the distance of the electron before and after the distortion from the electron cyclotron center O in cylindrical coordinates, R is the radius of the electron cyclotron motion, β is the central angle of the electron cyclotron motion, and the position of the electron before and after the distortion is relative to the electron cyclotron The deflection angle of the center O is given by: α=arcsin(2R[sin(β/2)] ² /r′ ² ). 7. An OAM modal identification device, characterized in that, comprising: The training data preparation sub-module is used to generate the OAM field strength distribution image containing nonlinear distortion affected by the residual magnetic field as the training data of the artificial intelligence model training sub-module, wherein the residual magnetic field refers to the generated magnetic field generated by the vortex electron generation module. The magnetic field that gradually decays between the diffraction amplifying module and the image receiving and acquisition module; The artificial intelligence model training sub-module is used to construct an artificial intelligence model, and use the training data to train the artificial intelligence model, so as to obtain the artificial intelligence model with the highest verification and recognition accuracy; The modal recognition sub-module uses the artificial intelligence model with the highest verification and recognition accuracy to perform real-time OAM modal recognition on the OAM field intensity distribution image collected by the image receiving and acquisition module, The training data includes OAM field strength distribution images affected by different degrees of residual magnetic field, wherein the nonlinear distortion caused by the influence of the residual magnetic field refers to one of nonlinear divergence and nonlinear convergence of the OAM field strength distribution image. one or more circumstances, The nonlinear divergence and nonlinear convergence refer to that the distances between the two points before and after the electron is distorted by the residual magnetic field and the electron gyration center O satisfy the following relationship:

where r and r' are the distance of the electron before and after the distortion from the electron cyclotron center O in cylindrical coordinates, R is the radius of the electron cyclotron motion, β is the central angle of the electron cyclotron motion, and the position of the electron before and after the distortion is relative to the electron cyclotron The deflection angle of the center O is given by: α=arcsin(2R[sin(β/2)] ² /r′ ² ). 8. An electronic device, characterized in that the electronic device comprises: at least one processor; and, a memory communicatively coupled to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the vortex electronic modality identification method of claim 6 .

Images