Beam spot position correction method, device, equipment and storage medium

Abstract

The present application discloses a beam spot position correction method, device, equipment, and storage medium, relating to the technical field of accelerators. The method uses a pre-trained prediction network model to replace manual operation for beam spot position correction. The method accurately reads minute beam spot offsets and automatically calibrates the beam spot position without manual intervention, greatly improving beam modulation efficiency. The method comprises: obtaining magnetic field constraint data for beam spot position correction, the magnetic field constraint data including the coordinates of the beam spot center position, beam energy data, and the distance between the correction magnet and the fluorescent target; inputting the magnetic field constraint data into a pre-trained prediction network model, so that the prediction network model predicts the magnetic field strength of the correction magnet in different directions based on the magnetic field constraint data; and correcting the beam spot position based on the magnetic field strength of the correction magnet in different directions.

Description

Technical Field

[0001] The present application relates to the technical field of accelerators, and in particular to a beam spot position correction method, device, equipment, and storage medium. Background Art

[0002] With the continuous development of accelerator technology, accelerators are able to accelerate charged particles to higher energies and stronger beam intensities, while also improving beam quality and stability. Ensuring beam transmission efficiency, reducing beam loss and scattering, and maintaining the beam at the center of the vacuum orbit are crucial tasks in accelerator beam tuning.

[0003] When the beam is transmitted along the vacuum track, the beam spot position can be used to reflect the actual position of the beam in the vacuum track, that is, the beam can be kept near the center of the track by correcting the beam spot position. In related technologies, if the beam spot deviates from the center of the vacuum track, it means that the beam has also deviated from the center, and the beam spot position needs to be corrected. At this time, the beam adjuster can correct the position of the beam spot by observing the fluorescent target and changing the magnetic field of the correction iron, so that the beam spot returns to the center of the vacuum track, thereby keeping the beam at the center of the vacuum track. However, the correction process of the beam spot position needs to rely on manual operation. It is difficult to accurately judge the slight offset of the beam spot and accurately adjust it. Repeated attempts are required to reach the ideal position, resulting in low accuracy of the beam spot position correction. Summary of the Invention

[0004] In view of this, the present application provides a method, device, equipment and storage medium for correcting the beam spot position. The main purpose is to solve the problem that the existing method of correcting the beam spot position through manual operation is difficult to accurately judge the slight offset of the beam spot and accurately adjust it, and repeated attempts are required to reach the ideal position, resulting in low accuracy of beam spot position correction.

[0005] According to a first aspect of the present application, a method for correcting a beam spot position is provided, comprising:

[0006] Acquiring magnetic field constraint data for beam spot position correction, the magnetic field constraint data including beam spot center position coordinates, beam energy data, and the distance between the correction magnet and the fluorescent target;

[0007] Inputting the magnetic field constraint data into a pre-trained prediction network model, so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model;

[0008] The beam spot position is corrected according to the magnetic field strength of the correction magnet in different directions.

[0009] Furthermore, before inputting the magnetic field constraint data into a pre-trained prediction network model so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model, the method further includes:

[0010] Acquiring training data for beam spot position correction, the training data including input data for magnetic field intensity constraint and output data for magnetic field intensity prediction;

[0011] Training a prediction network model according to the training data, and updating a mapping relationship between the input data and the output data during the training process, wherein the mapping relationship is used to predict the magnetic field intensity of the output data based on the input data as a magnetic field constraint;

[0012] When the training meets the iteration stopping condition, a prediction network model is generated according to the updated mapping relationship.

[0013] Furthermore, the obtaining of training data for beam spot position correction includes:

[0014] Pre-establishing a simulation system for particle acceleration simulation to simulate variable data for beam spot position correction and particle distribution at a fluorescent target through the simulation system;

[0015] Changing the variable data during the simulation process to correspondingly change the particle distribution at the fluorescent target and obtain the beam spot position distribution accompanying the change of the variable data;

[0016] Training data for beam spot position correction is constructed based on the beam spot position distribution accompanied by the change in variable data.

[0017] Furthermore, the variable data includes the magnetic field strength of the correction magnet in different directions, beam energy data, and the distance between the correction magnet and the fluorescent target. Changing the variable data during the simulation process to obtain a corresponding change in the particle distribution at the fluorescent target and obtain training data for beam spot position correction includes:

[0018] During the simulation process, the magnetic field strength of the correction magnet in different directions is changed to obtain a corresponding change in the particle distribution at the fluorescent target to obtain first training data for beam spot position correction; and / or

[0019] Changing the beam energy data during the simulation process to correspondingly change the particle distribution at the fluorescent target and obtain second training data for beam spot position correction; and / or

[0020] During the simulation process, the distance between the correction magnet and the fluorescent target is changed to obtain the corresponding change in particle distribution at the fluorescent target, thereby obtaining third training data for beam spot position correction.

[0021] Furthermore, constructing training data for beam spot position correction according to the beam spot position distribution that is changed by the accompanying variable data includes:

[0022] Determining a plurality of sets of associated simulation data based on the beam spot position distribution associated with the change in the variable data, the simulation data including the magnetic field strength of the correction magnet in different directions, beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the center position of the beam spot;

[0023] Select beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the beam spot center position to construct input data for magnetic field intensity constraint;

[0024] The magnetic field strength of the correction magnet in different directions is selected to construct output data for magnetic field strength prediction.

[0025] Furthermore, before training the prediction network model according to the training data and updating the mapping relationship between the input data and the output data during the training process, the method further includes:

[0026] Set the loss function to measure the training effect of the loss prediction network model based on the mean square error function;

[0027] Correspondingly, during the training process, the loss value between the predicted output data of the prediction network model and the output data for magnetic field strength prediction is calculated according to the loss function, so as to update the mapping relationship between the input data and the output data through the loss value.

[0028] Furthermore, the correction of the beam spot position according to the magnetic field strength of the correction magnet in different directions includes:

[0029] According to the magnetic field strength of the correction magnet in different directions, using the opposite value of the magnetic field strength as a position correction parameter;

[0030] The position correction parameter is sent to the correction magnet through a control instruction, so that the correction magnet corrects the beam spot position according to the position correction parameter.

[0031] According to a second aspect of the present application, a device for correcting a beam spot position is provided, comprising:

[0032] a first acquisition unit, configured to acquire magnetic field constraint data for beam spot position correction, wherein the magnetic field constraint data includes beam spot center position coordinates, beam energy data, and a distance between the correction magnet and the fluorescent target;

[0033] A prediction unit, configured to input the magnetic field constraint data into a pre-trained prediction network model, so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model;

[0034] The correction unit is used to correct the beam spot position according to the magnetic field strength of the correction magnet in different directions.

[0035] Furthermore, the device further comprises:

[0036] a second acquisition unit, configured to acquire training data for beam spot position correction before inputting the magnetic field constraint data into a pre-trained prediction network model so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model, wherein the training data includes input data for magnetic field strength constraint and output data for magnetic field strength prediction;

[0037] a training unit, configured to train a prediction network model according to the training data, and update a mapping relationship between the input data and the output data during the training process, wherein the mapping relationship is used to predict the magnetic field intensity of the output data based on the input data as a magnetic field constraint;

[0038] The generation unit is used to generate a prediction network model according to the updated mapping relationship when the training meets the iteration stopping condition.

[0039] Furthermore, the second acquiring unit includes:

[0040] An establishment module is used to pre-establish a simulation system for particle acceleration simulation, so as to simulate variable data for beam spot position correction and particle distribution at a fluorescent target through the simulation system;

[0041] An acquisition module, configured to change the variable data during the simulation process to correspondingly change the particle distribution at the fluorescent target and obtain a beam spot position distribution accompanying the change of the variable data;

[0042] A construction module is used to construct training data for beam spot position correction according to the beam spot position distribution that changes with the variable data.

[0043] Furthermore, the variable data includes the magnetic field strength of the correction magnet in different directions, beam energy data, and the distance between the correction magnet and the fluorescent target. The acquisition module is specifically used to:

[0044] During the simulation process, the magnetic field strength of the correction magnet in different directions is changed to obtain a corresponding change in the particle distribution at the fluorescent target to obtain first training data for beam spot position correction; and / or

[0045] Changing the beam energy data during the simulation process to correspondingly change the particle distribution at the fluorescent target and obtain second training data for beam spot position correction; and / or

[0046] During the simulation process, the distance between the correction magnet and the fluorescent target is changed to obtain the corresponding change in particle distribution at the fluorescent target, thereby obtaining third training data for beam spot position correction.

[0047] Furthermore, the construction module is specifically used to:

[0048] Determining a plurality of sets of associated simulation data based on the beam spot position distribution associated with the change in the variable data, the simulation data including the magnetic field strength of the correction magnet in different directions, beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the center position of the beam spot;

[0049] Select beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the beam spot center position to construct input data for magnetic field intensity constraint;

[0050] The magnetic field strength of the correction magnet in different directions is selected to construct output data for magnetic field strength prediction.

[0051] Furthermore, the training unit is further configured to set a loss function for measuring the loss prediction network model training effect according to a mean square error function before updating the mapping relationship between the input data and the output data during the training process when training the prediction network model according to the training data;

[0052] Correspondingly, during the training process, the loss value between the predicted output data of the prediction network model and the output data for magnetic field strength prediction is calculated according to the loss function, so as to update the mapping relationship between the input data and the output data through the loss value.

[0053] Furthermore, the correction unit is specifically configured to:

[0054] According to the magnetic field strength of the correction magnet in different directions, using the opposite value of the magnetic field strength as a position correction parameter;

[0055] The position correction parameter is sent to the correction magnet through a control instruction, so that the correction magnet corrects the beam spot position according to the position correction parameter.

[0056] According to a third aspect of the present application, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method described in the first aspect when executing the computer program.

[0057] According to a fourth aspect of the present application, a readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0058] By leveraging the above-mentioned technical solution, the present application provides a beam spot position correction method, device, equipment, and storage medium. Compared to the existing technology that relies on manual operation to achieve beam spot position correction, the present application obtains magnetic field constraint data for beam spot position correction, including the beam spot center position coordinates, beam energy data, and the distance between the correction magnet and the fluorescent target; inputs the magnetic field constraint data into a pre-trained prediction network model, which uses the prediction network model to predict the magnetic field strength of the correction magnet in different directions based on the magnetic field constraint data; and corrects the beam spot position based on the magnetic field strength of the correction magnet in different directions. The entire process uses the pre-trained prediction network model to replace manual operation for beam spot position correction. The prediction network model can accurately read slight beam spot offsets based on beam currents of different energies and beam spot positions, automatically calibrating the beam spot position without manual intervention, greatly improving beam adjustment efficiency.

[0059] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0061] Figure 1 1 is a flow chart of a method for correcting the beam spot position in one embodiment of the present application;

[0062] Figure 2 is a flow chart of a method for correcting the beam spot position in another embodiment of the present application;

[0063] Figure 3 yes Figure 2 A schematic flow chart of a specific implementation of step 201;

[0064] Figure 4 This is a schematic diagram of the structure of variable data in a simulation system in one embodiment of the present application;

[0065] Figure 5 yes Figure 1 A schematic flow chart of a specific implementation of step 103;

[0066] Figure 6 1 is a schematic structural diagram of a beam spot position correction device in one embodiment of the present application;

[0067] Figure 7 The figure is a schematic diagram of the device structure of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0068] The present invention will now be discussed with reference to several exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the present invention, rather than to imply any limitation on the scope of the present invention.

[0069] As used herein, the term "including" and variations thereof are to be interpreted as open-ended terms meaning "including, but not limited to." The term "based on" is to be interpreted as "based, at least in part, on." The terms "one embodiment" and "an embodiment" are to be interpreted as meaning "at least one embodiment." The term "another embodiment" is to be interpreted as meaning "at least one other embodiment."

[0070] In related technologies, if the beam spot deviates from the center of the vacuum track, it means that the beam current has also deviated from the center, and the beam spot position needs to be corrected. At this time, the beam adjuster can correct the position of the beam spot by observing the fluorescent target and changing the magnetic field of the correction iron, so that the beam spot returns to the center of the vacuum track, thereby keeping the beam current at the center of the vacuum track. However, the correction process of the beam spot position needs to rely on manual operation. It is difficult to accurately judge the slight deviation of the beam spot and accurately adjust it. Repeated attempts are required to reach the ideal position, resulting in low accuracy of the beam spot position correction.

[0071] In order to solve this problem, this embodiment provides a method for correcting the beam spot position, such as Figure 1 As shown, the method is applied to the control end of the beam adjustment system and includes the following steps:

[0072] 101. Obtain magnetic field constraint data for beam spot position correction.

[0073] The beam spot position refers to the location of the beam spot formed on a specific plane or target object in a particle accelerator. It is typically accurately expressed using three-dimensional spatial coordinates (x, y, z). The x and y coordinates represent the position of the beam spot on a plane perpendicular to the direction of the particle beam's motion, while the z coordinate represents its position along the direction of the beam's motion. For example, in a linear accelerator, the z-axis generally aligns with the accelerator's axis, and particles are accelerated along the z-axis. The position of the beam spot in the xy plane, perpendicular to the axis, determines the lateral distribution of the particle beam. Here, the beam spot position primarily refers to the position of the particle beam in the xy plane.

[0074] It is understandable that in a particle accelerator, the particle beam needs to be efficiently transmitted between complex pipes and numerous components. If the beam spot position is inaccurate, the transmission efficiency will be seriously affected. When the beam spot position deviates, the particle beam may deviate from the designed optimal transmission trajectory, which will cause the particles to collide with the accelerator pipe wall, resulting in a large number of particles being lost and the beam intensity being reduced. For example, in some large accelerators, even if the beam spot position is slightly offset, after long-distance transmission, a large number of particles will hit the pipe wall, significantly reducing the number of particles that ultimately reach the target position. By correcting the beam spot position, the particle beam can be transmitted strictly along the designed center track, minimizing the collision of particles with the pipe wall, improving the particle transmission efficiency, and ensuring the beam intensity and uniformity, thereby providing a stable, high-quality particle beam for subsequent experiments and applications.

[0075] In this embodiment, magnetic field constraint data includes the coordinates of the beam spot center, beam energy data, and the distance between the calibration magnet and the fluorescent target. The beam spot center coordinates refer to the coordinates of the center of the spot formed by the particle beam on a specific observation plane of the particle accelerator. The beam spot center coordinates can be used to determine the actual position of the particles during transmission and to determine whether they have deviated from the designed trajectory. Beam energy data describes the energy carried by particles in the particle beam. In the accelerator, particles gain energy through methods such as magnetic field acceleration. Beam energy data reflects the overall energy state of these particles. Different experiments or applications require particle beams with specific energies. The calibration magnet is a device used to generate a magnetic field to change the trajectory of the particle beam, and the fluorescent target is a device used to observe the position and morphology of the particle beam. The distance between the calibration magnet and the fluorescent target refers to the distance between the calibration magnet and the fluorescent target. This distance directly affects the effect of the magnetic field on beam spot position correction. By accurately measuring and controlling this distance, the effect of the magnetic field on the particle beam trajectory can be more accurately analyzed. The calibration magnet can then be precisely adjusted based on the beam spot observed on the fluorescent target, achieving effective correction of the beam spot position.

[0076] The executor of this embodiment may be a device or apparatus for correcting the beam spot position, which may be configured at the control end of a beam modulation system. The beam modulation system may adjust and control the particle beam so that the light spot generated by the particles after passing through the accelerator can be accurately presented at the center of the fluorescent target.

[0077] 102. Input the magnetic field constraint data into a pre-trained prediction network model, so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model.

[0078] In this embodiment, the pre-trained prediction network model can predict the magnetic field strength acting on the beam based on the magnetic field constraint data, and the magnetic field strength includes the prediction of the magnetic field strength of the correction magnet in different directions, and accordingly output the magnetic field strength of the correction magnet in different directions.

[0079] The above-mentioned prediction network model can be trained using a neural network. Specifically, during the training process, the neural network has multiple layers. The number of neurons in the input layer can be set according to the input magnetic field constraint data, and the number of neurons in the output layer can be set according to the output magnetic field strength. The number of neurons in the intermediate layer can be customized. For example, the input data includes the x-coordinate of the beam spot center position, the y-coordinate of the beam spot center position, the beam energy data, and the distance between the correction magnet and the fluorescent target. The output data includes the magnetic field strength of the correction magnet in the x-direction and the magnetic field strength of the correction magnet in the y-direction. In this case, a four-layer neural network can be set, containing 4, 64, 128, 64, and 2 neurons, respectively. After model training, the prediction network model can learn the mapping relationship between the input data and the output data to achieve correction of the beam spot position.

[0080] Accordingly, in the application stage of the prediction network model, the magnetic field constraint data can be input into the trained prediction network model to predict the magnetic field strength and obtain the magnetic field strength of the correction magnet in different directions.

[0081] 103. Correct the beam spot position according to the magnetic field strength of the correction magnet in different directions.

[0082] Understandably, there are many reasons why the beam spot can deviate from the center of the fluorescent target, such as excessive or insufficient beam deflection in the upstream magnet. Macroscopically, this can be equated or abstracted to the collimated beam shifting under the action of the deflection magnet. The beam spot shift caused by these reasons can be understood as being caused by the equivalent magnetic field strength, that is, the magnetic field strength applied in different directions by the correction magnet output by the prediction network model.

[0083] Specifically, after the actual beam spot position is input into the prediction network model, the model outputs the magnetic field strength of the correction magnet in different directions, which is the equivalent magnetic field strength Bx,By that causes the beam to deflect. To maintain the beam at the center of the vacuum orbit, an equivalent magnetic field strength of equal magnitude and opposite direction, -Bx,-By, is used to neutralize the beam spot position, aligning it with the fluorescent target.

[0084] The beam spot position correction method provided in the embodiments of the present application, compared to the existing technology that relies on manual operation to achieve beam spot position correction, obtains magnetic field constraint data for beam spot position correction. The magnetic field constraint data includes the coordinates of the beam spot center position, beam energy data, and the distance between the correction magnet and the fluorescent target; inputs the magnetic field constraint data into a pre-trained prediction network model, which uses the prediction network model to predict the magnetic field strength of the correction magnet in different directions based on the magnetic field constraint data; and corrects the beam spot position based on the magnetic field strength of the correction magnet in different directions. The entire process uses the pre-trained prediction network model to replace manual operation for beam spot position correction. The prediction network model can accurately read slight beam spot offsets based on beam currents of different energies and beam spot positions, automatically calibrating the beam spot position without manual intervention, greatly improving beam adjustment efficiency.

[0085] Furthermore, considering the training process of the prediction network model, in the above embodiment, as Figure 2 As shown, before step 102, the method further includes the following steps:

[0086] 201. Obtain training data for beam spot position correction.

[0087] 202. Train the prediction network model according to the training data, and update the mapping relationship between the input data and the output data during the training process.

[0088] 203. When the training meets the iteration stopping condition, a prediction network model is generated according to the updated mapping relationship.

[0089] The training data includes input data for magnetic field strength constraints and output data for magnetic field strength prediction. The input data includes beam energy, the distance between the calibration magnet and the fluorescent target, and the coordinates of the beam spot center. The output data includes the magnetic field strength of the calibration magnet in different directions. This training data can be obtained through a particle acceleration simulation system, experimentally obtained using magnetic field measurement equipment, or calculated using electromagnetic theory. These are not limited and can be selected based on practical needs.

[0090] Furthermore, considering the training accuracy of the prediction network model, a loss function can be set based on the mean square error function to measure the training effect of the loss prediction network model. Accordingly, during the training process, the loss value between the predicted output data of the prediction network model and the output data used for magnetic field intensity prediction is calculated based on the loss function, so as to update the mapping relationship between the input data and the output data through the loss value. Here, the loss function can quantitatively describe the degree of difference between the model's predicted results and the actual results. By calculating the average of the squares of the differences between the predicted output data and the actual output data as the loss value, the accuracy of the prediction network model can be intuitively understood. Generally, the smaller the value, the closer the prediction network model's prediction results are to the actual results.

[0091] Specifically, if Figure 3 As shown, step 201 includes the following steps:

[0092] 301. Pre-establish a simulation system for particle acceleration simulation to simulate variable data for beam spot position correction and particle distribution at the fluorescent target through the simulation system.

[0093] 302. During the simulation process, the variable data is changed to obtain a corresponding change in the particle distribution at the fluorescent target, and the beam spot position distribution accompanying the change in the variable data is obtained.

[0094] 303. Construct training data for beam spot position correction based on the beam spot position distribution associated with the change in variable data.

[0095] In this embodiment, a simulation system for particle acceleration simulation is used to simulate the interaction of multiple physical fields, such as electromagnetic fields and particle motion. It can simulate the particle distribution at the correction magnet and fluorescent target based on beam optics and the Montlock method. Specifically, a three-dimensional geometric model containing components such as the acceleration cavity, correction magnet, and fluorescent target can be created based on the actual accelerator structure. The simulation system must ensure the relative accuracy of the size, shape, and position of each component, especially the precise setting of the distance between the correction magnet and the fluorescent target. For the correction magnet, the magnetic field intensity distribution in different directions can be set based on its physical properties and design requirements. By continuously changing the magnetic field value of the correction magnet within the simulation system, the beam direction can be changed, thereby changing the beam spot position on the fluorescent target. At this point, the magnetic field value of the correction magnet and the center of the bunch distribution at the fluorescent target are recorded separately. This allows the variation of the particle distribution at the fluorescent target with beam energy and magnet position to be understood, thereby obtaining multiple sets of training data relevant to beam spot position correction.

[0096] The aforementioned variable data includes the calibration magnet's magnetic field strength in different directions, beam energy data, and the distance between the calibration magnet and the fluorescent target. To better understand how the particle distribution at the fluorescent target changes with beam energy and magnet position, any parameter in the variable data can be adjusted as a simulation variable. During the simulation, by changing any parameter in the variable data, the beam spot position distribution under different variable data combinations can be obtained.

[0097] Specifically, after variable data is input into the simulation system, the magnetic field strength of the correction magnet in different directions is changed during the simulation process to obtain the corresponding change in the particle distribution at the fluorescent target, thereby obtaining the first training data for beam spot position correction; and / or the beam energy data is changed during the simulation process to obtain the corresponding change in the particle distribution at the fluorescent target, thereby obtaining the second training data for beam spot position correction; and / or the distance between the correction magnet and the fluorescent target is changed during the simulation process to obtain the corresponding change in the particle distribution at the fluorescent target, thereby obtaining the third training data for beam spot position correction.

[0098] The structure of variable data in the above simulation system can be referred to Figure 4 As shown, by changing the magnetic field strength Bx,By of the correction magnet in different directions, the distance length between the correction magnet and the fluorescent target, and any variable data in the beam energy data energy, the center position coordinates centX,centY of the beam at the center of the fluorescent target can be changed accordingly.

[0099] It is understandable that in addition to the variable data used as simulation variables, simulation parameters such as the simulation time step and the total simulation time also need to be set. In order to ensure the calculation accuracy of the particle motion trajectory, the time step should be small enough, but it should not be too small to cause excessive calculation. The total simulation time needs to be determined based on the expected movement time of the particle from the initial position to the fluorescent target. During the iterative calculation process, the simulation system is run for each value of the simulation variable. When the simulation variable is the magnetic field intensity, the force on the particle will change with the change of the magnetic field intensity, thereby affecting the particle's motion trajectory. Accordingly, after each simulation calculation is completed, the beam spot position distribution that accompanies the change in variable data is recorded, such as the coordinates and number of particles, as well as parameters related to the beam spot position, such as the coordinates of the beam spot center position, the beam spot size, etc.

[0100] Correspondingly, after obtaining the beam spot position distribution that changes with the accompanying variable data, multiple groups of associated simulation data can be determined based on the beam spot position distribution that changes with the accompanying variable data. The simulation data include the magnetic field strength of the correction magnet in different directions, beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the center position of the beam spot; the beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the center position of the beam spot are selected to construct input data for magnetic field strength constraint; the magnetic field strength of the correction magnet in different directions is selected to construct output data for magnetic field strength prediction.

[0101] In practical applications, the training data constructed for the prediction network model training process can be expressed as [energy, Bx, By, centX, centY, length], where energy is the beam energy data, Bx, By is the magnetic field strength of the calibration magnet in different directions, centX, centY are the coordinates of the beam spot center, and length is the distance between the calibration magnet and the fluorescent target. Accordingly, [energy, centX, centY, length] is the input data for magnetic field strength constraint, and [Bx, By] is the output data for magnetic field strength prediction. Here, magnetic field strength is equivalent to the label of the supervised training process, and the output data of the prediction network model must be aligned with the label. This allows the prediction network model to accurately output [Bx, By] based on the input data [energy, centX, centY, length].

[0102] In actual application scenarios, the coordinates of the center position of the beam spot are used as the current beam spot position, and the center position of the fluorescent target is used as the expected beam spot position. The process of correcting the beam spot position is to adjust the current beam spot position to the expected beam spot position. Specifically, Figure 5 As shown, step 103 includes the following steps:

[0103] 401. Based on the magnetic field strength of the correction magnet in different directions, use the opposite value of the magnetic field strength as a position correction parameter.

[0104] 402. Send the position correction parameter to the correction magnet via a control instruction, so that the correction magnet corrects the beam spot position according to the position correction parameter.

[0105] In this embodiment, the magnetic field strength of the correction magnet in different directions has a quantitative relationship with the change in beam spot position. For example, increasing the magnetic field strength of the correction magnet in the x-direction will cause the beam spot position to deviate to the left in the y-direction. Correspondingly, the magnetic field strength of the correction magnet in the x-direction needs to be controlled to decrease in order to move the beam spot position closer to the center in the y-direction. In other words, the beam spot position correction process is based on the reaction of the magnetic field strength of the correction magnet in different directions. By negating the magnetic field strength, a force opposite to the current beam spot deviation direction is generated, pushing the beam spot toward the center of the fluorescent target.

[0106] Specifically, the opposite values of the magnetic field strength of the correction magnet in different directions can be used as position correction parameters for beam spot position correction. This position correction parameter allows the correction magnet to generate a force on the beam that is opposite to the current beam spot offset direction. After such an opposite force, the beam direction will move in the direction opposite to the current beam spot offset, so that the beam spot position is located at the center of the fluorescent target.

[0107] Furthermore, in order to precisely control the correction magnet, the position correction parameters can be encoded and packaged and converted into control instructions in a specific format. The control instructions contain detailed correction parameter information and the requirements for executing the operation, which can be sent to the control system of the correction magnet through a set communication link. Accordingly, after receiving the control instruction, the control system of the correction magnet will parse and verify the control instruction to ensure the accuracy and completeness of the control instruction. Once the verification is passed, the control system will adjust the excitation current of the correction magnet according to the position correction parameters in the control instruction. Since the magnetic field strength of the correction magnet is linearly related to the excitation current, by changing the magnitude and direction of the excitation current, the magnetic field strength of the correction magnet in different directions can be precisely adjusted, so that it generates a force opposite to the original magnetic field strength. In this way, after the beam is subjected to this force, its motion trajectory will change, thereby causing the beam spot position to produce a corresponding displacement, thereby achieving correction of the beam spot position.

[0108] Furthermore, as a specific implementation of the above method, the embodiment of the present application provides a device for correcting the beam spot position, such as Figure 6 As shown, the device includes: a first acquisition unit 51, a prediction unit 52 and a correction unit 53.

[0109] A first acquisition unit 51 is used to acquire magnetic field constraint data for beam spot position correction, wherein the magnetic field constraint data includes the coordinates of the beam spot center position, beam energy data, and the distance between the correction magnet and the fluorescent target;

[0110] The prediction unit 52 is configured to input the magnetic field constraint data into a pre-trained prediction network model, so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model;

[0111] The correction unit 53 is used to correct the beam spot position according to the magnetic field strength of the correction magnet in different directions.

[0112] Compared to the existing techniques that rely on manual operation to achieve beam spot position correction, the beam spot position correction device provided in the present invention obtains magnetic field constraint data for beam spot position correction. The magnetic field constraint data includes the coordinates of the beam spot center, beam energy data, and the distance between the correction magnet and the fluorescent target. The magnetic field constraint data is input into a pre-trained prediction network model, which uses the prediction network model to predict the magnetic field strength of the correction magnet in different directions based on the magnetic field constraint data. The beam spot position is then corrected based on the magnetic field strength of the correction magnet in different directions. The entire process uses the pre-trained prediction network model to replace manual operation for beam spot position correction. The prediction network model can accurately read slight beam spot offsets based on beam currents of different energies and beam spot positions, automatically calibrating the beam spot position without manual intervention, greatly improving beam adjustment efficiency.

[0113] In a specific application scenario, the device further includes:

[0114] a second acquisition unit, configured to acquire training data for beam spot position correction before inputting the magnetic field constraint data into a pre-trained prediction network model so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model, wherein the training data includes input data for magnetic field strength constraint and output data for magnetic field strength prediction;

[0115] a training unit, configured to train a prediction network model according to the training data, and update a mapping relationship between the input data and the output data during the training process, wherein the mapping relationship is used to predict the magnetic field intensity of the output data based on the input data as a magnetic field constraint;

[0116] The generation unit is used to generate a prediction network model according to the updated mapping relationship when the training meets the iteration stopping condition.

[0117] In a specific application scenario, the second obtaining unit includes:

[0118] An establishment module is used to pre-establish a simulation system for particle acceleration simulation, so as to simulate variable data for beam spot position correction and particle distribution at a fluorescent target through the simulation system;

[0119] An acquisition module, configured to change the variable data during the simulation process to correspondingly change the particle distribution at the fluorescent target and obtain a beam spot position distribution accompanying the change of the variable data;

[0120] A construction module is used to construct training data for beam spot position correction according to the beam spot position distribution that changes with the variable data.

[0121] In a specific application scenario, the variable data includes the magnetic field strength of the correction magnet in different directions, beam energy data, and the distance between the correction magnet and the fluorescent target. The acquisition module is specifically used to:

[0122] During the simulation process, the magnetic field strength of the correction magnet in different directions is changed to obtain a corresponding change in the particle distribution at the fluorescent target to obtain first training data for beam spot position correction; and / or

[0123] Changing the beam energy data during the simulation process to correspondingly change the particle distribution at the fluorescent target and obtain second training data for beam spot position correction; and / or

[0124] During the simulation process, the distance between the correction magnet and the fluorescent target is changed to obtain the corresponding change in particle distribution at the fluorescent target, thereby obtaining third training data for beam spot position correction.

[0125] In a specific application scenario, the construction module is specifically used to:

[0126] Determining a plurality of sets of associated simulation data based on the beam spot position distribution associated with the change in the variable data, the simulation data including the magnetic field strength of the correction magnet in different directions, beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the center position of the beam spot;

[0127] Select beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the beam spot center position to construct input data for magnetic field intensity constraint;

[0128] The magnetic field strength of the correction magnet in different directions is selected to construct output data for magnetic field strength prediction.

[0129] In a specific application scenario, the training unit is further configured to set a loss function for measuring the loss prediction network model training effect according to a mean square error function before the prediction network model is trained according to the training data and the mapping relationship between the input data and the output data is updated during the training process;

[0130] Correspondingly, during the training process, the loss value between the predicted output data of the prediction network model and the output data for magnetic field strength prediction is calculated according to the loss function, so as to update the mapping relationship between the input data and the output data through the loss value.

[0131] In a specific application scenario, the correction unit is specifically used to:

[0132] According to the magnetic field strength of the correction magnet in different directions, using the opposite value of the magnetic field strength as a position correction parameter;

[0133] The position correction parameter is sent to the correction magnet through a control instruction, so that the correction magnet corrects the beam spot position according to the position correction parameter.

[0134] It should be noted that, for other corresponding descriptions of the functional units involved in the beam spot position correction device provided in this embodiment, reference may be made to the corresponding descriptions of the beam spot position correction method described above, and will not be repeated here.

[0135] Based on the above-mentioned beam spot position correction method, accordingly, an embodiment of the present application further provides a storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned beam spot position correction method is implemented.

[0136] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0137] Based on the above-mentioned beam spot position correction method and beam spot position correction device embodiment, in order to achieve the above-mentioned purpose, the embodiment of the present application also provides a physical device for correcting the beam spot position, which can be specifically a computer, a smart phone, a tablet computer, a smart watch, a server, or a network device, etc. The physical device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to implement the above-mentioned beam spot position correction method.

[0138] Optionally, the physical device may also include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a Wi-Fi module, etc. The user interface may include a display, an input unit such as a keyboard, etc. Optional user interfaces may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a Wi-Fi interface), etc.

[0139] In an exemplary embodiment, see Figure 7The physical device includes a communication bus, a processor, a memory, and a communication interface. It may also include an input / output interface and a display device. The various functional units can communicate with each other via the bus. The memory stores a computer program, and the processor is configured to execute the program stored in the memory and perform the beam spot position correction method described in the above embodiment.

[0140] Those skilled in the art will understand that the physical device structure for correcting the beam spot position provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0141] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the physical device for beam spot position correction, supporting the execution of information processing programs and other software and / or programs. The network communication module is used to enable communication between components within the storage medium and with other hardware and software in the physical information processing device.

[0142] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus the necessary general hardware platform, or by means of hardware. By applying the technical solution of the present application, compared with the current existing methods, the present application uses a pre-trained prediction network model to replace manual operation to calibrate the beam spot position. The prediction network model can accurately read the slight offset of the beam spot based on the beam current of different energies and the beam spot position, and can automatically calibrate the beam spot position without manual intervention, greatly improving the beam adjustment efficiency.

[0143] Those skilled in the art will understand that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required to implement the present application. Those skilled in the art will understand that the modules in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more devices different from the implementation scenario. The modules of the above-mentioned implementation scenario can be combined into one module, or can be further split into multiple sub-modules.

[0144] The serial numbers of the above application are for descriptive purposes only and do not represent the advantages or disadvantages of the implementation scenarios. The above disclosure only discloses several specific implementation scenarios of the present application, but the present application is not limited thereto. Any changes that can be conceived by those skilled in the art should fall within the scope of protection of the present application.

Claims

1. A method for correcting a beam spot position, characterized in that: include: Acquiring magnetic field constraint data for beam spot position correction, the magnetic field constraint data including beam spot center position coordinates, beam energy data, and the distance between the correction magnet and the fluorescent target; Inputting the magnetic field constraint data into a pre-trained prediction network model, so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model; Correcting the beam spot position according to the magnetic field strength of the correction magnet in different directions; Before inputting the magnetic field constraint data into a pre-trained prediction network model so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model, training data for beam spot position correction is acquired, wherein the training data includes input data for magnetic field strength constraint and output data for magnetic field strength prediction; Training a prediction network model according to the training data, and updating a mapping relationship between the input data and the output data during the training process, wherein the mapping relationship is used to predict the magnetic field intensity of the output data based on the input data as a magnetic field constraint; When the training meets the iteration stop condition, a prediction network model is generated according to the updated mapping relationship; The method of obtaining training data for beam spot position correction includes: pre-establishing a simulation system for particle acceleration simulation to simulate variable data used for beam spot position correction and particle distribution at a fluorescent target through the simulation system; changing the variable data during the simulation process to obtain a corresponding change in the particle distribution at the fluorescent target, thereby obtaining a beam spot position distribution that is accompanied by the change in the variable data; and constructing training data for beam spot position correction based on the beam spot position distribution that is accompanied by the change in the variable data.

2. The beam spot position correction method according to claim 1, characterized in that: The variable data includes the magnetic field strength of the correction magnet in different directions, beam energy data, and the distance between the correction magnet and the fluorescent target. The variable data is changed during the simulation process to obtain the particle distribution at the fluorescent target accordingly to obtain training data for beam spot position correction, including: During the simulation process, the magnetic field strength of the correction magnet in different directions is changed to obtain a corresponding change in the particle distribution at the fluorescent target to obtain first training data for beam spot position correction; and / or Changing the beam energy data during the simulation process to correspondingly change the particle distribution at the fluorescent target and obtain second training data for beam spot position correction; and / or During the simulation process, the distance between the correction magnet and the fluorescent target is changed to obtain the corresponding change in particle distribution at the fluorescent target, thereby obtaining third training data for beam spot position correction.

3. The beam spot position correction method according to claim 1, characterized in that: The step of constructing training data for beam spot position correction based on the beam spot position distribution that is changed by the accompanying variable data comprises: Determining a plurality of sets of associated simulation data based on the beam spot position distribution associated with the change in the variable data, the simulation data including the magnetic field strength of the correction magnet in different directions, beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the center position of the beam spot; Select beam energy data, the distance between the correction magnet and the fluorescent target, and the coordinates of the beam spot center position to construct input data for magnetic field intensity constraint; The magnetic field strength of the correction magnet in different directions is selected to construct output data for magnetic field strength prediction.

4. The beam spot position correction method according to claim 1, wherein: Before training the prediction network model according to the training data and updating the mapping relationship between the input data and the output data during the training process, the method further includes: Set the loss function to measure the training effect of the loss prediction network model based on the mean square error function; Correspondingly, during the training process, the loss value between the predicted output data of the prediction network model and the output data for magnetic field strength prediction is calculated according to the loss function, so as to update the mapping relationship between the input data and the output data through the loss value.

5. The method for correcting the beam spot position according to any one of claims 1 to 4, characterized in that: The step of correcting the beam spot position according to the magnetic field strength of the correction magnet in different directions includes: According to the magnetic field strength of the correction magnet in different directions, using the opposite value of the magnetic field strength as a position correction parameter; The position correction parameter is sent to the correction magnet through a control instruction, so that the correction magnet corrects the beam spot position according to the position correction parameter.

6. A beam spot position correction device, characterized in that: include: a first acquisition unit, configured to acquire magnetic field constraint data for beam spot position correction, wherein the magnetic field constraint data includes beam spot center position coordinates, beam energy data, and a distance between the correction magnet and the fluorescent target; A prediction unit, configured to input the magnetic field constraint data into a pre-trained prediction network model, so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model; a correction unit, configured to correct the beam spot position according to the magnetic field strength of the correction magnet in different directions; The device further includes: a second acquisition unit, configured to acquire training data for beam spot position correction before inputting the magnetic field constraint data into a pre-trained prediction network model so as to predict the magnetic field strength of the correction magnet in different directions according to the magnetic field constraint data through the prediction network model, wherein the training data includes input data for magnetic field strength constraint and output data for magnetic field strength prediction; a training unit, configured to train the prediction network model according to the training data, and update a mapping relationship between the input data and the output data during the training process, wherein the mapping relationship is used to predict the magnetic field strength of the output data based on the input data as the magnetic field constraint; and a generation unit, configured to generate the prediction network model according to the updated mapping relationship when the training satisfies an iteration stop condition; The second acquisition unit includes: an establishment module for pre-establishing a simulation system for particle acceleration simulation, so as to simulate the variable data used for beam spot position correction and the particle distribution at the fluorescent target through the simulation system; an acquisition module for changing the variable data during the simulation process to correspondingly obtain the changed particle distribution at the fluorescent target and obtain the beam spot position distribution accompanying the change of the variable data; and a construction module for constructing training data for beam spot position correction based on the beam spot position distribution accompanying the change of the variable data.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the beam spot position correction method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the beam spot position correction method according to any one of claims 1 to 5 are implemented.

Images