Particle accelerator beam commissioning method, apparatus, device, and medium

Abstract

The application discloses a particle accelerator beam commissioning method, device, equipment and medium, and relates to the accelerator commissioning technical field, which can make each iteration process have clear data support by splitting the accelerator beam commissioning process into a stage solidified reproducible process, and guarantee the accuracy of the beam commissioning result. The method comprises the following steps: creating process configuration information of the particle accelerator according to a beam commissioning task; on the basis of setting a commissioning target, using the process configuration information to perform iterative commissioning on the particle accelerator, so as to generate setting information of a control variable according to an observation variable obtained in each iteration commissioning; applying the setting information of the control variable to an actual particle accelerator and / or a virtual particle accelerator, so as to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator; and if the iteration commissioning process meets an iteration termination condition, generating a first beam commissioning result according to the current beam state.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of accelerator debugging, and in particular to a particle accelerator beam debugging method, device, equipment and medium. BACKGROUND

[0002] The particle accelerator is a large device for basic physical research by accelerating charged particles through an electric field to obtain extremely high energy and speed. The performance of the particle accelerator is highly dependent on the accuracy of beam debugging, that is, the charged particles in the accelerator do not automatically form high-quality beams, and the key parameters need to be adjusted through beam debugging to ensure the quality of the beam in the accelerator.

[0003] In the actual beam debugging process, the traditional beam debugging method mainly relies on experience operation, which highly depends on expert experience. However, as a complex, nonlinear and strongly coupled system, the running state of the accelerator will drift with time, temperature and other factors, which makes the results of expert experience debugging highly dependent on the equipment state and personal judgment, and it is difficult to reproduce the same beam debugging results, and it is difficult to guarantee the accuracy of the beam debugging results. SUMMARY

[0004] Therefore, the present application provides a particle accelerator beam debugging method, device, equipment and medium, which mainly aims to solve the problem that the results of expert experience debugging are highly dependent on the equipment state and personal judgment, it is difficult to reproduce the same beam debugging results, and it is difficult to guarantee the accuracy of the beam debugging results.

[0005] According to a first aspect of the present application, a particle accelerator beam debugging method is provided, comprising:

[0006] According to the beam debugging task, the process configuration information of the particle accelerator is created, and the process configuration information at least includes variable information and algorithm information for process configuration of the particle accelerator, and the variable information at least includes observation variables and control variables;

[0007] On the basis of setting the debugging target, the process configuration information is used to iteratively debug the particle accelerator, so as to generate setting information of the control variable according to the observation variable obtained by each iteration debugging;

[0008] apply the setting information of the control variable to the actual particle accelerator and / or the virtual particle accelerator to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator, the current beam state being obtained by using the actual particle accelerator to obtain real beam data to obtain standardized actual observation variables, changing the setting information of the control variable according to the standardized actual observation variables, applying the changed setting information of the control variable to the virtual particle accelerator, and iteratively debugging the virtual particle accelerator according to the changed setting information of the control variable to obtain the current beam state, the virtual particle accelerator being a digital twin model of a simulated actual accelerator operating state obtained according to deviation data of the beam debugging process;

[0009] If the iterative debugging process meets an iteration termination condition, a first beam debugging result is generated according to the current beam state, otherwise, the iterative debugging process is repeatedly executed until the iterative debugging process meets the iteration termination condition, the iteration termination condition being that the current beam state meets the set debugging target and / or the number of iterations reaches an iteration threshold.

[0010] Further, before the process configuration information of the particle accelerator is created according to the beam debugging task, the method further comprises:

[0011] constructing basic configuration information of the particle accelerator, the basic configuration information at least including element information and structure information used for basic configuration of the particle accelerator;

[0012] Correspondingly, the process configuration information of the particle accelerator is created according to the beam debugging task, comprising:

[0013] In the accelerator structure constructed using the basic configuration information, the process configuration information of the particle accelerator is created according to the beam debugging task.

[0014] Further, the process configuration information of the particle accelerator is created according to the beam debugging task in the accelerator structure constructed using the basic configuration information, comprising:

[0015] In the accelerator structure constructed using the basic configuration information, element information of the particle accelerator is determined;

[0016] According to the beam debugging task, role configuration is performed on the element information of the particle accelerator to obtain variable information of the particle accelerator in the debugging process;

[0017] According to the beam debugging task, algorithm configuration is performed on the variable information of the particle accelerator in the debugging process to obtain algorithm information of the particle accelerator in the debugging process.

[0018] Further, on the basis of setting the debugging target, the particle accelerator is iteratively debugged using the process configuration information to generate setting information of the control variable according to the observation variable obtained in each iteration debugging, including:

[0019] On the basis of setting the debugging target, the particle accelerator injection beam parameters and the objective function of the optimization algorithm are determined;

[0020] The particle accelerator is iteratively debugged using the process configuration information to obtain the state data of the observation variable of the particle beam parameter in the acceleration process;

[0021] According to the quantitative deviation formed by the state data of the observation variable and the objective function of the optimization algorithm, the setting information of the control variable is generated.

[0022] Further, before the setting information of the control variable is applied to the actual particle accelerator and / or virtual particle accelerator to obtain the current beam state output by the actual particle accelerator and / or virtual particle accelerator, the method further comprises:

[0023] According to the deviation data of the beam debugging process, a model is trained to obtain a virtual particle accelerator, which is used to simulate the running state of the actual particle accelerator in a virtual environment.

[0024] Further, the model training according to the deviation data of the beam debugging process to obtain a virtual particle accelerator comprises:

[0025] A digital twin model is pre-constructed in a virtual environment, which is constructed based on the basic configuration information of the actual particle accelerator;

[0026] The setting information of the control variable is applied to the actual particle accelerator and the digital twin model respectively to obtain the deviation data of the beam debugging process through the current beam state output by the actual particle accelerator and the current beam state output by the digital twin model;

[0027] According to the deviation data of the beam debugging process, the digital twin model is trained to obtain a virtual particle accelerator.

[0028] Further, on the basis of setting the debugging target, the particle accelerator is iteratively debugged using the process configuration information to generate setting information of the control variable according to the observation variable obtained in each iteration debugging, the method further comprises:

[0029] The setting information of the control variable is applied to the actual particle accelerator to obtain the current beam state output by the actual particle accelerator;

[0030] If the intermediate beam state output by the actual particle accelerator does not satisfy the set commissioning target, the setting information of the observation variable changing the control variable is obtained through iterative commissioning;

[0031] The setting information of the changed control variable is applied to the virtual particle accelerator, so that the virtual particle accelerator performs iterative commissioning on the particle accelerator according to the setting information of the changed control variable, until the iterative commissioning process satisfies the iterative termination condition, and a second beam commissioning result is generated according to the current beam state output by the virtual particle accelerator;

[0032] According to the current beam state output by the virtual particle accelerator, the setting information of the corresponding control variable in the virtual particle accelerator is applied to the actual particle accelerator, so as to verify the second beam commissioning result through the current beam state output by the actual particle accelerator.

[0033] According to a second aspect of the present application, a beam commissioning device of a particle accelerator is provided, comprising:

[0034] A creating unit is configured to create flow configuration information of the particle accelerator according to a beam commissioning task, wherein the flow configuration information at least includes variable information and algorithm information used for flow configuration of the particle accelerator, and the variable information at least includes observation variables and control variables;

[0035] A commissioning unit is configured to perform iterative commissioning on the particle accelerator using the flow configuration information on the basis of a set commissioning target, so as to generate setting information of the control variable according to the observation variable obtained through each iterative commissioning;

[0036] A first obtaining unit is configured to apply the setting information of the control variable to the actual particle accelerator and / or the virtual particle accelerator, so as to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator, wherein the current beam state is realized by the following manner: first, using the actual particle accelerator to obtain real beam data to obtain standardized actual observation variables, changing the setting information of the control variable according to the standardized actual observation variables, applying the setting information of the changed control variable to the virtual particle accelerator, and performing iterative commissioning on the virtual particle accelerator according to the setting information of the changed control variable to obtain the current beam state, wherein the virtual particle accelerator is a digital twin model of a simulated actual accelerator operating state obtained according to deviation data of a beam commissioning process;

[0037] The first generating unit is configured to generate a first beam debugging result according to the current beam state if the iterative debugging process meets an iteration termination condition, or repeat the execution of the iterative debugging process until the iterative debugging process meets the iteration termination condition, wherein the iteration termination condition is that the current beam state meets the set debugging target and / or the number of iterations reaches an iteration threshold.

[0038] Further, the apparatus further comprises:

[0039] The constructing unit is configured to construct basic configuration information of the particle accelerator before the process configuration information of the particle accelerator is created according to the beam debugging task, wherein the basic configuration information at least includes element information and structure information used for basic configuration of the particle accelerator;

[0040] Correspondingly, the creating unit is specifically configured to create the process configuration information of the particle accelerator according to the beam debugging task in the accelerator structure constructed using the basic configuration information.

[0041] Further, the creating unit is specifically further configured to:

[0042] Determine element information of the particle accelerator in the accelerator structure constructed using the basic configuration information;

[0043] Configure roles of the element information of the particle accelerator according to the beam debugging task to obtain variable information of the particle accelerator in the debugging process;

[0044] Configure algorithms of the variable information of the particle accelerator in the debugging process according to the beam debugging task to obtain algorithm information of the particle accelerator in the debugging process.

[0045] Further, the debugging unit is specifically configured to:

[0046] Determine particle beam parameters injected by the accelerator and an objective function of the optimization algorithm on the basis of the set debugging target;

[0047] Iteratively debug the particle accelerator using the process configuration information to obtain state data of an observed variable of the particle beam parameters in the acceleration process;

[0048] Generate setting information of a control variable according to a quantitative deviation formed by the state data of the observed variable and the objective function of the optimization algorithm.

[0049] Further, the apparatus further comprises:

[0050] a model training unit configured to perform model training according to the deviation data of the beam commissioning process before the setting information of the control variable is applied to the actual particle accelerator and / or the virtual particle accelerator to obtain the current beam state output by the actual particle accelerator and / or the virtual particle accelerator, so as to obtain a virtual particle accelerator used to simulate the running state of the actual particle accelerator in a virtual environment.

[0051] Further, the model training unit is specifically configured to:

[0052] pre-construct a digital twin model in a virtual environment, the digital twin model being constructed based on the basic configuration information of the actual particle accelerator;

[0053] apply the setting information of the control variable to the actual particle accelerator and the digital twin model respectively, so as to obtain the deviation data of the beam commissioning process by the current beam state output by the actual particle accelerator and the current beam state output by the digital twin model;

[0054] train the digital twin model according to the deviation data of the beam commissioning process, so as to obtain a virtual particle accelerator.

[0055] Further, the apparatus further comprises:

[0056] a second obtaining unit configured to, on the basis of the set commissioning target, perform iterative commissioning on the particle accelerator using the process configuration information, so as to generate the setting information of the control variable according to the observation variable obtained by each iteration of the commissioning, and then apply the setting information of the control variable to the actual particle accelerator to obtain the current beam state output by the actual particle accelerator;

[0057] a changing unit configured to, if the intermediate beam state output by the actual particle accelerator does not meet the set commissioning target, change the setting information of the control variable by the observation variable obtained by the iterative commissioning;

[0058] a second generating unit configured to apply the changed setting information of the control variable to the virtual particle accelerator, so that the virtual particle accelerator performs iterative commissioning on the particle accelerator according to the changed setting information of the control variable, until the iterative commissioning process meets an iteration termination condition, and then generate a second beam commissioning result according to the current beam state output by the virtual particle accelerator;

[0059] a verifying unit configured to, according to the current beam state output by the virtual particle accelerator, apply the setting information of the corresponding control variable in the virtual particle accelerator to the actual particle accelerator, so as to verify the second beam commissioning result by the current beam state output by the actual particle accelerator.

[0060] According to a third aspect of the present application, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps of the method of the first aspect when executing the computer program.

[0061] According to a fourth aspect of the present application, a readable storage medium is provided, storing a computer program, and the computer program implementing the steps of the method of the first aspect when executed by a processor.

[0062] By means of the above technical solution, the particle accelerator beam commissioning method, device, equipment and medium provided by the present application are compared with the existing technology which relies on experience to commission the accelerator. According to the present application, the process configuration information of the particle accelerator is created according to the beam commissioning task, and the process configuration information at least includes variable information and algorithm information used for process configuration of the particle accelerator. The variable information at least includes observation variables and control variables. On the basis of setting the commissioning target, the particle accelerator is iteratively commissioned using the process configuration information to generate setting information of the control variables according to the observation variables obtained in each iteration commissioning. The setting information of the control variables is applied to the actual particle accelerator and / or the virtual particle accelerator to obtain the current beam state output by the actual particle accelerator and / or the virtual particle accelerator. The virtual accelerator is a digital twin model simulating the actual accelerator operating state according to the deviation data of the beam commissioning process. If the iteration commissioning process meets the iteration termination condition, a first beam commissioning result is generated according to the current beam state, otherwise, the iteration commissioning process is repeatedly executed until the iteration commissioning process meets the iteration termination condition. The iteration termination condition is that the current beam state meets the set commissioning target and / or the number of iteration commissioning reaches the iteration threshold. The whole process divides the accelerator beam commissioning process into a reproducible process that is fixed in stages, so that each process can be directly connected to the actual particle accelerator and / or the virtual particle accelerator on the basis of setting the commissioning target. It can not only use the virtual accelerator to complete the early parameter prediction, but also quickly land to the actual device verification without additional process adaptation, reducing the risk of trial and error of the actual device, and providing clear data support for each iteration process, which can monitor the process progress and effect in real time, and ensure the accuracy of the beam commissioning result.

[0063] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the specific embodiments of the present application can be implemented in accordance with the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0064] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0065] Figure 1 This is one of the flowcharts of the beam tuning method for the particle accelerator provided in this application;

[0066] Figure 2 This is the second flowchart of the beam tuning method for the particle accelerator provided in this application;

[0067] Figure 3 This is a schematic diagram of the process for creating process configuration information in one embodiment of this application;

[0068] Figure 4 This is a flowchart illustrating the creation phase debugging module in one embodiment of this application;

[0069] Figure 5 yes Figure 1 A flowchart illustrating a specific implementation method for step 102;

[0070] Figure 6 This is a flowchart illustrating the iterative debugging process of a particle accelerator in one embodiment of this application;

[0071] Figure 7 This is the third flowchart of the beam tuning method for the particle accelerator provided in this application;

[0072] Figure 8 This is a flowchart illustrating the construction process of a virtual particle accelerator in one embodiment of this application;

[0073] Figure 9 This is a schematic diagram of the beam tuning device of a particle accelerator in one embodiment of this application;

[0074] Figure 10 This is a schematic diagram of the device structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0075] The invention will now be discussed with reference to several exemplary embodiments. It should be understood that these embodiments are described merely to enable those skilled in the art to better understand and thus implement the invention, and are not intended to imply any limitation on the scope of the invention.

[0076] As used herein, the terms "comprises," "comprising," "includes," "including," "has," "having," "contains," "containing," or variations thereof, are intended to be open-ended terms that mean "including, but not limited to." The term "based on" is intended to be open-ended terms that means "based, at least in part, on." The terms "one embodiment" and "an embodiment" are intended to be open-ended terms that mean "at least one embodiment." The term "another embodiment" is intended to be an open-ended term that means "at least one other embodiment."

[0077] In the actual beam commissioning process, the traditional beam commissioning method mainly relies on experience operation, and the process highly depends on expert experience. However, as an complex, nonlinear and strongly coupled system, the running state of the accelerator will drift with time, temperature and other factors, which makes the results of expert experience commissioning highly dependent on the device state and personal judgment, and it is difficult to reproduce the same beam commissioning results, and it is difficult to guarantee the accuracy of the beam commissioning results.

[0078] To solve this problem, the embodiment provides a beam commissioning method of a particle accelerator, as shown in the method can be applied to a beam commissioning platform, including the following steps: Figure 1

[0079] 101. Create process configuration information of the particle accelerator according to the beam commissioning task.

[0080] In the embodiment, the beam commissioning task is to convert the fuzzy requirements for optimizing the beam performance of the accelerator into specific, quantifiable, bounded and verifiable target optimization tasks. As a specific implementation of the beam commissioning task, the process configuration information can decompose the fuzzy requirements of the beam commissioning task into executable steps, parameter ranges, device interface rules, etc., so that the beam commissioning task can be directly landed to the actual particle accelerator and / or virtual particle accelerator.

[0081] ​The process configuration information at least includes variable information and algorithm information for configuring the particle accelerator. The variable information at least includes observation variables and control variables. The observation variables are feedback indicators for monitoring the beam state and can directly reflect the main stream performance and operating state. The data is usually from a detector or a monitoring system and is the core basis for determining whether the debugging is up to standard. For example, beam energy, beam intensity, beam spot size and position, etc. The control variables are operating parameters for regulating the beam state and are usually actively adjustable parameters used to accurately regulate the observation variables to gradually approach the debugging target. For example, magnet parameters, magnetic field strength, etc. The algorithm information is used to convert the beam debugging target into executable parameter calculation, state prediction, and process control rules. Through automatic / semi-automatic logic linkage of observation variables and control variables, the beam state is accurately optimized, and the actual / virtual particle accelerator is completely adapted to the docking requirements. The specific algorithm information can include but is not limited to AI training algorithms (such as reinforcement learning, deep learning, etc.) and various optimization algorithms (such as genetic algorithm, particle swarm optimization, Bayesian optimization, etc.).

[0082] The execution subject of the embodiment can be a beam debugging device or equipment of a particle accelerator, which can be configured on a beam debugging platform. By integrating accelerator basic data, debugging task boundaries, algorithm selection rules, etc. in one time, without scattered analysis or subsequent supplement of core settings, the pre-stage cost is reduced, and the debugging efficiency and accuracy are greatly improved.

[0083] 102. On the basis of setting the debugging target, the particle accelerator is iteratively debugged using the process configuration information to generate setting information of the control variables according to the observation variables obtained in each iteration.

[0084] In the embodiment, the setting of the debugging target is the core anchor point of the iterative debugging process. The essence is to convert the core requirements of beam debugging into quantifiable and verifiable beam performance indicators and standard reaching standards, i.e., what state needs to be reached to be qualified.

[0085] Specifically, the debugging target can be mapped to the observation variables and the control variables to obtain complete process configuration information. Then, the process configuration information is imported into the particle accelerator control system, and the system automatically loads the debugging target, variable mapping, rule algorithm, and constraint boundary to complete the hardware interface adaptation of the observation variables and the control variables, so as to ensure that the algorithm can directly read the observation data and output adjustment instructions. The system automatically triggers the observation variable element to collect observation variable data according to the configured collection frequency. Based on the debugging target in the process configuration information, the observation variable data is compared with the target value, the deviation is calculated, and the optimal adjustment scheme of the control variables, i.e., the setting information of the control variables, is determined according to the optimization logic of the algorithm and in combination with the constraint rules.

[0086] Taking the specific scenario of "15GeV beam energy optimization" as an example, the debugging target is set as the beam energy to be stable at 15GeV±0.2GeV. The target is achieved if this range is met for three consecutive iterations. In the process configuration information, the observed variable is the beam energy, which is collected by the energy detector. The control variable is the accelerating cavity voltage, with an adjustment range of 3-5MV and a single step size ≤0.15MV. The selected algorithm is the Bayesian optimization algorithm. In the first iteration of the corresponding iterative debugging process, the observed variable showed that the current beam energy was 14.5 GeV, which was 0.5 GeV lower than the target lower limit. Based on the observed energy deviation of -0.5 GeV and combined with the voltage adjustment rules configured in the process, the Bayesian algorithm generated the control variable setting information to increase the accelerating cavity voltage from the current 4.0 MV to 4.15 MV. The accelerator updated its parameters according to this setting information, and after stable operation, it entered the second iteration. The observed variable showed that the current beam energy was 14.8 GeV, which was 0.2 GeV lower than the target lower limit. The algorithm combined the correlation data of voltage +0.15 MV and energy +0.2 GeV from the previous round and generated the control variable setting information to increase the accelerating cavity voltage from 4.15 MV to 4.25 MV. After the accelerator updated its parameters and stabilized, it entered the third iteration. The observed variable showed that the current beam energy was 15.05 GeV, which was within 15 GeV ± 0.2 GeV. The target range for GeV was set so that the accelerator voltage remained constant at 4.25 MV. This meant that the third iteration had met the target, and the iteration debugging was terminated. The final control variable setting was set to an accelerator cavity voltage of 4.25 MV.

[0087] In this embodiment, the setting information of the control variables does not rely on fixed presets, but is dynamically generated based on the real observation data of each iteration, which can accurately respond to changes in the beam state and make the correction of deviations more targeted.

[0088] 103. Apply the setting information of the control variables to the actual particle accelerator and / or the virtual particle accelerator to obtain the current beam state output by the actual particle accelerator and / or the virtual particle accelerator.

[0089] Understandably, continuously adjusting the accelerator's beam state according to the set beam tuning target is the core of iterative tuning. The control variable settings are calculated through process configuration information during the iterative tuning process. Only when these settings are synchronized to the accelerator can the actual beam state be changed, providing real data for the next round of observations and allowing iterative tuning to continue. In other words, only after the control parameter settings are synchronized to the accelerator can the beam state gradually move from the current deviation value towards the target value, preventing a disconnect between algorithmic decisions and actual tuning.

[0090] In this embodiment, the virtual accelerator is a digital twin model simulating the actual accelerator operating state obtained by modeling the deviation data of the beam commissioning process. Its working principle is to establish a virtual accelerator environment in a computer program, including the structure of the accelerator, the electromagnetic field distribution and other key elements. By inputting control variables such as particle species, energy settings, acceleration field strength and magnetic field strength, the program calculates the particle trajectory and energy change in this virtual environment based on electromagnetic, relativistic and other physical theories. The specific control variable setting information can be accessed to the virtual particle accelerator in the form of instructions, and the system automatically maps the control variables to the virtual particle accelerator components to enable the virtual particle accelerator to call the built-in beam dynamics model based on the control variables, replicate the actual particle accelerator operating environment, simulate the beam trajectory at the preset time step, and real-time calculate the beam energy, emittance, intensity and other core parameters, and output the current beam state accordingly.

[0091] Similarly, the control variable setting information can be accessed to the actual particle accelerator in the form of instructions, and sent to the corresponding control element through a dedicated control bus. After receiving the instruction, the control element quickly responds through the built-in closed-loop adjustment module to accurately adjust its operating parameters to ensure that the actual output is consistent with the control variable information. After the control element completes parameter adjustment, the actual particle accelerator enters the preset stabilizer, and the beam completes motion state adaptation in the new electric and magnetic field environment. After the stabilization period ends, the built-in observation device of the accelerator is started to collect data at a preset frequency to obtain the current beam state.

[0092] It should be noted that the control variable setting information is synchronously issued to the actual particle accelerator and / or virtual particle accelerator in the form of instructions, and both are executed in parallel to adjust and simulate. At the same time, the virtual particle accelerator can first output the prediction result to quickly predict the parameter effect of the virtual accelerator and verify the real state of the actual accelerator, calibrate the model through the data linkage of the two, reduce the risk, and at the same time improve the efficiency and accuracy of the commissioning.

[0093] 104a, if the iterative commissioning process meets the iteration termination condition, a first beam commissioning result is generated according to the current beam state. 104b, otherwise, repeat the iterative commissioning process until the iterative commissioning process meets the iteration termination condition.

[0094] In the embodiment, the iteration termination condition is that the current beam state meets the set debugging target and / or the number of iteration debugging reaches the iteration number threshold. The specific determination reference is based on the current beam state output by the actual particle accelerator and / or the virtual particle accelerator, combined with the number of iteration debugging. If the current beam state meets the index requirements of the set debugging target for the beam state and / or the current number of debugging reaches the iteration number threshold, the setting information of the control variable is correspondingly taken as the effective control variable, facilitating subsequent reuse or optimization. Otherwise, the optimization algorithm obtains the observation variable through the next iteration debugging combined with the deviation and the process configuration information, and generates the setting information of the control variable.

[0095] For example, the current beam state is 15.8 GeV, which is within the set debugging target range of 15.5-16 GeV, and the accelerator voltage is maintained at 4.25 MV as the effective control variable.

[0096] Compared with the current existing technology which relies on experience to operate the particle accelerator for beam debugging, the particle accelerator beam debugging method provided by the embodiment of the present application creates process configuration information of the particle accelerator according to the beam debugging task. The process configuration information at least includes variable information and algorithm information for configuring the process of the particle accelerator, and the variable information at least includes observation variables and control variables. On the basis of the set debugging target, the process configuration information is used to iteratively debug the particle accelerator to generate setting information of the control variable according to the observation variable obtained in each iteration debugging. The setting information of the control variable is applied to the actual particle accelerator and / or the virtual particle accelerator to obtain the current beam state output by the actual particle accelerator and / or the virtual particle accelerator. The virtual accelerator is a digital twin model simulating the running state of the actual accelerator obtained by modeling the deviation data of the beam debugging process. If the iteration debugging process meets the iteration termination condition, a first beam debugging result is generated according to the current beam state. Otherwise, the iteration debugging process is repeatedly executed until the iteration debugging process meets the iteration termination condition. The iteration termination condition is that the current beam state meets the set debugging target and / or the number of iteration debugging reaches the iteration number threshold. The entire process divides the accelerator beam debugging process into a reproducible process that is fixed in stages, so that each process can be directly connected to the actual particle accelerator and / or the virtual particle accelerator on the basis of the set debugging target. The virtual accelerator can be used to complete the preliminary parameter prediction, and can be quickly landed to the actual device verification without additional process adaptation, reducing the risk trial and error of the actual device, and providing clear data support for each iteration process, real-time monitoring of process progress and effect, and ensuring the accuracy of the beam debugging result.

[0097] In actual application scenarios, the process configuration information is a product of the combination of basic configuration and specific debugging tasks, and needs to be based on the inherent properties of the accelerator structure, around the beam debugging target, inherent iterative logic, variable rules and judgment standards, so that the prompt process can be automated and landed. Further, in the above embodiment, as shown in Figure 2 Before step 101, the method further includes the following steps:

[0098] 201, constructing the basic configuration information of the particle accelerator.

[0099] In this embodiment, the basic configuration information at least includes element information and structure information used for basic configuration of the particle accelerator. The element information includes element types, which are classified and defined according to the particularity of various elements in the particle accelerator, and can provide a preset element template for subsequent operations. The element information also includes accelerator basic information, i.e. establishing the file name of the particle accelerator to be debugged, including the name, description and main parameters. The structure information defines the physical layout and functional association of the elements, ensures that the instructions can accurately act on the target elements during debugging, and specifically includes the physical structure layout, such as the spatial trend of the accelerator body and the spatial position of each system, and the logical structure association, such as classifying elements into acceleration modules, focusing modules, measurement modules and other modules with clear connection logic according to functions.

[0100] Specifically, taking the basic configuration information as the created accelerator framework, the accelerator structure can be managed in segments, and elements can be created in the segments. On the one hand, the accelerator framework can divide and organize the physical area of the accelerator, i.e. add, modify or delete each segment in the accelerator, to provide the function of managing the structure of the accelerator. On the other hand, the accelerator framework can instantiate specific accelerator elements in the defined segments, and give them special identifiers and operation parameters. These elements can be created based on the element types in the accelerator framework.

[0101] Correspondingly, the above step 101 specifically includes: 202, creating the process configuration information of the particle accelerator according to the beam debugging task in the accelerator structure constructed by using the basic configuration information.

[0102] After completing the basic configuration of the accelerator structure, the process configuration information of the particle accelerator is created according to the beam commissioning task. According to the beam commissioning task, the elements and their related attributes that the beam commissioning task focuses on can be selected, and different roles can be assigned to the elements. These roles can be divided into preset elements (parameters remain unchanged during the commissioning process), control elements (parameters will be adjusted by the algorithm), and observation elements (outputs will be observed and analyzed by the algorithm). This step can also be automatically selected in batches through task attributes to improve efficiency. According to the nature and goal of the beam commissioning task, the most suitable automatic algorithm can be selected, including but not limited to AI algorithms and various optimization algorithms. The selected algorithm will adjust the control variables according to the observed variables in the subsequent iterative commissioning process.

[0103] Correspondingly, in the accelerator structure constructed using the basic configuration information, the element information of the particle accelerator is determined; the element information of the particle accelerator is configured according to the beam commissioning task, and the variable information of the particle accelerator in the commissioning process is obtained; the variable information of the particle accelerator in the commissioning process is configured according to the beam commissioning task, and the algorithm information of the particle accelerator in the commissioning process is obtained.

[0104] The above creation process can refer to the process shown in Figure 3 , which creates an accelerator in element classification, manages the accelerator in segments, creates elements in accelerator segments or adds elements according to element classification, combines element attributes in segments and provided algorithms, creates at least one module suitable for stage commissioning according to the beam commissioning task, and creates process configuration information of the particle accelerator according to the module combination of stage commissioning. Through the above process, the entire initial configuration process from defining the basic data of the accelerator to preparing the beam commissioning task and selecting the algorithm can be completed, which fully prepares for the subsequent process commissioning and iterative commissioning optimization.

[0105] Specifically, in the process of creating a stage commissioning module, referring to the process shown in Figure 4 , the required elements and their attributes can be selected through the configured stage first. Then, according to the requirements of the beam commissioning task, the selected elements are classified into preset elements, control elements or observation elements. Here, the element type can also be automatically selected in batches according to the properties of the beam commissioning task. Finally, according to the characteristics of the commissioning task, the appropriate algorithm is selected to execute the task, which can include AI training algorithm or other optimization algorithm.

[0106] Specifically, in the process of iterative commissioning of the particle accelerator, as shown in Figure 5 , step 102 includes the following steps:

[0107] 301. On the basis of setting the commissioning target, the particle beam parameters of the accelerator injection and the objective function of the optimization algorithm are determined.

[0108] 302, iteratively commissioning the particle accelerator using the procedure configuration information to obtain state data of the particle beam parameters in the observation variables during the acceleration process.

[0109] 303, generating setting information of the control variables according to the quantified deviation formed by the state data of the observation variables and the objective function of the optimization algorithm.

[0110] In this embodiment, the objective function of the optimization algorithm is a core formula for quantifying the deviation of the observation variables from the commissioning target, which needs to cover all core indicators, explicitly weight and constraint conditions, and the smaller the function output value, the closer the beam state to the commissioning target. Correspondingly, in the iteration process, the particle beam acceleration process is performed according to the collection nodes and frequencies configured by the procedure, and the state data of the observation variables is collected in real time by the observation device, and the state data of the observation variables is further brought into the objective function of the optimization algorithm to calculate the deviation value of the current iteration round, and the setting information of the control variables in the next round is generated in combination with the procedure configuration constraints.

[0111] Further, the setting information of the control variables given by the optimization algorithm is applied to the actual particle accelerator and / or the virtual particle accelerator. Correspondingly, the system reads the latest results of the actual particle accelerator and the virtual particle accelerator through the interface, and the results of the virtual particle accelerator are displayed in real time on the interface for the operator to observe and debug. Subsequently, the system checks whether the current commissioning state has reached the set commissioning target or the iteration number has reached the iteration number threshold.

[0112] The above-mentioned iterative commissioning process can refer to the procedure shown in Figure 6 The entire particle accelerator commissioning process starts from the initialization configuration, including setting the type and inlet parameters of the particle beam, defining the upper limit of the iteration number of the commissioning, and configuring the objective function of the optimization algorithm. After the configuration is completed, the system starts the commissioning loop. In each iteration, the optimization algorithm observes the state data of the observation elements in the accelerator, and generates setting information of the control variables according to the state data of the observation elements, and further applies the setting information of the control variables to the virtual particle accelerator and the actual particle accelerator. On the one hand, the virtual particle accelerator reads the virtual state data of the observation variables through the interface during execution, and displays the virtual state data to the interface. If the virtual state data meets the set commissioning target or the iteration number reaches the iteration number threshold, the commissioning iteration process is ended, otherwise the iterative commissioning process is repeated. On the other hand, the actual particle accelerator reads the actual state data of the observation variables through the interface during execution. If the actual state data meets the set commissioning target or the iteration number reaches the iteration number threshold, the commissioning iteration process is ended, otherwise the iterative commissioning process is repeated.

[0113] In actual application scenarios, considering that the internal space of an actual particle accelerator is extremely limited and cannot densely arrange multi-dimensional measurement devices, the beam information that can be observed by the actual particle accelerator during the optimization process is a one-dimensional scalar, that is, the variables that can be observed by the actual particle accelerator are limited, and if the complete state of the actual particle accelerator at a certain position after acceleration by a magnet is to be understood, a virtual particle accelerator needs to be used. Correspondingly, as shown in FIG. 1, before step 103, the above method further includes the following steps: Figure 7

[0114] 401. training a model according to the deviation data of the beam commissioning process to obtain a virtual particle accelerator.

[0115] In this embodiment, the virtual particle accelerator is used to simulate the running state of the actual particle accelerator in a virtual environment, and by making the virtual particle accelerator replicate the running characteristics of the actual particle accelerator during model training, the simulation result is close to reality and can replace part of the actual commissioning. Here, the deviation data of the beam commissioning process is the real mapping from the control variable input to the observation variable output of the actual particle accelerator, and the core of model training is to make the virtual particle accelerator learn this mapping relationship, so that when the same number of control variable setting information is input, the deviation between the corresponding output observation variable of the virtual particle accelerator and the output observation variable of the actual particle accelerator gradually decreases, and the virtual simulation is matched with the real running.

[0116] Specifically, a digital twin model is pre-constructed in a virtual environment, and the digital twin model is constructed based on the basic configuration information of the actual particle accelerator; the setting information of the control variable is applied to the actual particle accelerator and the digital twin model respectively, so as to obtain the deviation data of the beam commissioning process by the current beam state output by the actual particle accelerator and the current beam state output by the digital twin model; and the digital twin model is trained according to the deviation data of the beam commissioning process to obtain a virtual particle accelerator.

[0117] The above virtual particle accelerator construction process can refer to the flow shown in FIG. 2. Figure 8 When the system is first run or has no historical data, initial data acquisition is first performed in the process of pre-constructing a digital twin model in a virtual environment. These data may come from some benchmark tests or initial beam tuning. Using these initial data, a neural network correction model is constructed and trained, and the model is used as a digital twin model to learn the inherent error between virtual simulation and real system, laying a foundation for subsequent iterative correction.

[0118] ​Correspondingly, in each iteration of the debugging process, the beam state data of the real particle accelerator and the virtual particle accelerator are continuously recorded, including the control parameters and the observation results, and the data is further cleaned and preprocessed to ensure data quality and consistency for model updating. Then the latest preprocessed data is updated to the training data set, and the neural network correction model is incrementally learned or retrained using the updated data set, which means that the model will continuously adjust its internal parameters according to the latest bias data, thereby gradually reducing the bias between the virtual and the real.

[0119] It can be understood that, in order to ensure the accuracy of the neural network correction model obtained by training, the neural network correction model can be precision evaluated, that is, the degree of agreement between the prediction results and the actual accelerator performance. If the precision does not meet the preset standard requirement, the process is cycled back to the data collection stage, waiting for the data generated by the next beam commissioning to continue to update the neural network correction model. If the precision has met the standard, it is considered that the high-precision digital twin model has been successfully constructed, and the constructed digital twin model can be used as a reliable platform for subsequent optimization and debugging.

[0120] In actual application scenarios, considering that the beam commissioning cost of the actual particle accelerator is high, the real beam data can be obtained using the actual particle accelerator to obtain standardized actual observation variables, and the setting information of the control variables is changed according to the standardized actual observation variables, and the setting information of the changed control variables is applied to the virtual particle accelerator to perform iterative debugging on the virtual particle accelerator according to the setting information of the changed control variables to obtain a second beam commissioning result that meets the iteration termination condition.

[0121] Correspondingly, after the setting information of the control variables is generated according to the observation variables obtained by each iteration of the debugging, the setting information of the control variables is applied to the actual particle accelerator to obtain the current beam state output by the actual particle accelerator; if the intermediate beam state output by the actual particle accelerator does not meet the set debugging target, the setting information of the control variables is changed by the observation variables obtained by the iterative debugging; the setting information of the changed control variables is applied to the virtual particle accelerator to make the virtual particle accelerator perform iterative debugging on the particle accelerator according to the setting information of the changed control variables, until the iteration debugging process meets the iteration termination condition, and a second beam commissioning result is generated according to the current beam state output by the virtual particle accelerator; the setting information of the corresponding control variables in the virtual particle accelerator is applied to the actual particle accelerator according to the current beam state output by the virtual particle accelerator, so as to verify the second beam commissioning result by the current beam state output by the actual particle accelerator.

[0122] First, during the initial commissioning start or continuous online operation of the particle accelerator, the system applies the process configuration information or the current control variables to the actual particle accelerator according to the set commissioning target (such as beam orbit, emittance or energy stability requirements). Through real-time acquisition system, the output state of the actual particle accelerator under the current physical environment is obtained. This step not only serves as the initial data acquisition of the commissioning, but also as the real-time state monitoring during online operation, which is used to capture the deviation of the beam state caused by physical factors such as hardware installation error, environmental temperature drift or magnet hysteresis effect.

[0123] Subsequently, once the intermediate beam state output by the actual particle accelerator is monitored to not meet the set commissioning target, or the beam index is detected to exceed the allowed error range during operation, the system immediately corrects the control strategy using the observed variables (i.e. actual running deviation) obtained. The system maps the control variable setting information containing the true physical deviation to the virtual particle accelerator, triggering the iterative commissioning process in the virtual environment. In this digital twin environment, the algorithm performs multiple rounds of simulation and parameter optimization based on the latest physical state until the virtual commissioning process meets the convergence condition, thereby calculating the second beam commissioning result (i.e. the theoretically optimal control parameters) that can compensate for the current physical deviation.

[0124] Finally, the second beam commissioning result generated in the virtual particle accelerator is issued and applied to the actual particle accelerator as a correction instruction. The actual device executes this set of simulation-optimized control variables to output the corrected current beam state. By comparing the actual output with the virtual prediction result, the commissioning result is verified; if it is verified, the current deviation correction is completed; if the system is in online operation mode, the real-time monitoring of the beam state continues, and once the state drift occurs again, the above-mentioned "monitoring-simulation correction-execution verification" closed-loop process is automatically repeated to realize the automatic commissioning and high-stability operation of the particle accelerator throughout its life cycle.

[0125] In actual application scenarios, in order to avoid the iteration lag of a single virtual accelerator due to system complexity and improve the commissioning efficiency of large-scale accelerators, the virtual accelerator can be split into multiple segmented virtual nodes, such as ion source segment, acceleration segment and extraction segment. Each node corresponds to a region of the actual accelerator, and each node synchronously receives the actual data of the corresponding region and independently iterates, and then the commissioning results are integrated by the integration platform.

[0126] In the process of specifically splitting the virtual accelerator into multiple segmented virtual nodes, the overall virtual accelerator model can be split into multiple functionally independent segmented virtual nodes based on the functional area topology and beam transmission logic of an actual large-scale particle accelerator, with each segmented virtual node corresponding to one physical area of the actual accelerator. Exemplarily, the segmented virtual nodes include, but are not limited to, an ion source virtual node (corresponding to an ion generation and extraction area of the actual accelerator), an acceleration virtual node (corresponding to a superconducting acceleration cavity or radio frequency acceleration segment area), a focusing virtual node (corresponding to a quadrupole / hexapole focusing area), and an extraction virtual node (corresponding to a beam deflection and extraction area). It should be noted that after splitting, each segmented virtual node only loads the beam dynamics equations, device parameter models, and control variable set of the corresponding physical area, and eliminates redundant calculation modules unrelated to the area, thereby achieving lightweight single-node modeling.

[0127] In order to facilitate localized data-driven and independent iteration of the segmented nodes, each segmented virtual node can be pre-configured with a dedicated data acquisition and transmission link. Through sensors deployed in the corresponding physical area of the actual accelerator, real-time operation data of different physical areas can be acquired. Further, the acquired actual data is synchronously issued to the corresponding segmented virtual node as initial parameters and constraint conditions for node iteration debugging. Each segmented virtual node independently carries out iteration calculation based on local data to optimize the control variables of the node until a preset local iteration termination condition is met, for example, the beam transmission efficiency of the area is greater than or equal to 95%.

[0128] Since each node only processes local parameters of the area, high-dimensional parameter coupling calculation of the full system model is avoided, significantly reducing the computational load of the single node, and fundamentally solving the iteration lag problem. However, in order to facilitate node collaborative control, an integrated segmented virtual node system platform can be built to receive the local iteration optimal solution uploaded by each segmented virtual node, including but not limited to control variable configuration parameters and beam characteristic simulation results, and integrate the virtual debugging scheme of the full link according to the beam transmission sequence of the actual accelerator. Then, based on the overall beam transmission constraint rules of the actual accelerator, the integrated virtual debugging scheme is simulated and verified. If the local solution of a segmented node leads to beam characteristics of downstream nodes that do not meet the constraints, for example, the energy of the acceleration node is too high to cause beam divergence of the focusing node, the integrated platform issues adjustment instructions to the segmented node to trigger secondary local iteration of the node until the debugging results of all segmented nodes meet the global constraint requirements.

[0129] In an actual application scenario, the process of beam commissioning of a particle accelerator can be divided into multiple commissioning processes, each of which can be arranged as a commissioning module. An operator can intuitively design the commissioning logic by dragging and connecting different commissioning modules. The commissioning logic supports the sequential execution of beam commissioning tasks, such as first calibrating the orbit and then matching the dispersion. It also supports complex logic control, such as automatically triggering the orbit correction process when the beam transmission efficiency is lower than 90%.

[0130] Further, as a specific implementation of the above method, the embodiment of the present application provides a beam commissioning device of a particle accelerator, as shown in the figure, the device comprises a creating unit 51, a commissioning unit 52, a first obtaining unit 53 and a first generating unit 54. Figure 9

[0131] The creating unit 51 is configured to create process configuration information of the particle accelerator according to a beam commissioning task, wherein the process configuration information at least comprises variable information and algorithm information used for process configuration of the particle accelerator, and the variable information at least comprises observation variables and control variables.

[0132] The commissioning unit 52 is configured to perform iterative commissioning on the particle accelerator using the process configuration information on the basis of a set commissioning target, so as to generate setting information of the control variables according to the observation variables obtained in each iteration.

[0133] The first obtaining unit 53 is configured to apply the setting information of the control variables to an actual particle accelerator and / or a virtual particle accelerator, so as to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator, wherein the current beam state is realized by: first using the actual particle accelerator to obtain real beam data, obtaining standardized actual observation variables, changing the setting information of the control variables according to the standardized actual observation variables, applying the changed setting information of the control variables to the virtual particle accelerator, and performing iterative commissioning on the virtual particle accelerator according to the changed setting information of the control variables to obtain the current beam state, wherein the virtual particle accelerator is a digital twin model of a simulated actual accelerator operating state obtained according to deviation data of the beam commissioning process.

[0134] The first generating unit 54 is configured to generate a first beam commissioning result according to the current beam state if an iteration termination condition is met in the iterative commissioning process, or repeatedly perform the iteration commissioning process until the iteration termination condition is met, wherein the iteration termination condition is that the current beam state meets the set commissioning target and / or the number of iterations reaches an iteration threshold.

[0135] ​Compared with the current existing technology which relies on experience operation to perform beam commissioning on the accelerator, the particle accelerator beam commissioning device provided by the embodiment of the present application creates process configuration information of the particle accelerator according to a beam commissioning task, the process configuration information at least includes variable information and algorithm information used for process configuration of the particle accelerator, and the variable information at least includes observation variables and control variables; on the basis of setting a commissioning target, the particle accelerator is iteratively commissioned using the process configuration information to generate setting information of the control variables according to the observation variables obtained in each iteration commissioning; the setting information of the control variables is applied to an actual particle accelerator and / or a virtual particle accelerator to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator, the virtual accelerator is a digital twin model simulating an actual accelerator operating state obtained by modeling deviation data of the beam commissioning process; if the iteration commissioning process meets an iteration termination condition, a first beam commissioning result is generated according to the current beam state, otherwise, the iteration commissioning process is repeatedly executed until the iteration commissioning process meets the iteration termination condition, and the iteration termination condition is that the current beam state meets the set commissioning target and / or the number of iteration commissioning reaches an iteration number threshold. The whole process divides the accelerator beam commissioning process into a reproducible process which is fixed in stages, so that each process can be directly connected to the actual particle accelerator and / or the virtual particle accelerator on the basis of the set commissioning target, which can not only use the virtual accelerator to complete the early parameter prediction, but also quickly land to the actual device verification without additional process adaptation, reduces the risk trial and error of the actual device, and can provide clear data support for each iteration process, can monitor the process progress and effect in real time, and ensures the accuracy of the beam commissioning result.

[0136] In a specific application scenario, the device further includes:

[0137] The construction unit is configured to, before the process configuration information of the particle accelerator is created according to the beam commissioning task, construct basic configuration information of the particle accelerator, and the basic configuration information at least includes element information and structure information used for basic configuration of the particle accelerator.

[0138] Correspondingly, the creation unit is specifically configured to, in the accelerator structure constructed using the basic configuration information, create the process configuration information of the particle accelerator according to the beam commissioning task.

[0139] In a specific application scenario, the creation unit is specifically further configured to:

[0140] In the accelerator structure constructed using the basic configuration information, the element information of the particle accelerator is determined.

[0141] The element information of the particle accelerator is role configured according to the beam commissioning task to obtain variable information of the particle accelerator in the commissioning process.

[0142] According to the beam commissioning task, algorithm configuration is performed on variable information of the particle accelerator in a commissioning process, to obtain algorithm information of the particle accelerator in the commissioning process.

[0143] In a specific application scenario, the commissioning unit is specifically configured to:

[0144] On the basis of setting a commissioning target, a particle beam parameter injected by the accelerator and an objective function of an optimization algorithm are determined.

[0145] The particle accelerator is iteratively commissioned using the flow configuration information, to obtain state data of an observation variable of the particle beam parameter in an acceleration process.

[0146] According to a quantitative deviation formed by the state data of the observation variable and the objective function of the optimization algorithm, setting information of a control variable is generated.

[0147] In a specific application scenario, the device further includes:

[0148] A model training unit is configured to, before the setting information of the control variable is applied to an actual particle accelerator and / or a virtual particle accelerator to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator, perform model training according to deviation data of the beam commissioning process, to obtain a virtual particle accelerator, which is configured to simulate a running state of the actual particle accelerator in a virtual environment.

[0149] In a specific application scenario, the model training unit is specifically configured to:

[0150] A digital twin model is pre-constructed in a virtual environment, and the digital twin model is constructed based on basic configuration information of the actual particle accelerator.

[0151] The setting information of the control variable is respectively applied to the actual particle accelerator and the digital twin model, to obtain deviation data of the beam commissioning process by comparing a current beam state output by the actual particle accelerator with a current beam state output by the digital twin model.

[0152] The digital twin model is trained according to the deviation data of the beam commissioning process, to obtain a virtual particle accelerator.

[0153] In a specific application scenario, the device further includes:

[0154] The second obtaining unit is configured to, on the basis of the set commissioning target, perform iterative commissioning on the particle accelerator using the process configuration information, to generate setting information of a control variable according to an observation variable obtained in each iteration of the commissioning, and then apply the setting information of the control variable to an actual particle accelerator to obtain a current beam state output by the actual particle accelerator;

[0155] The changing unit is configured to, if the intermediate beam state output by the actual particle accelerator does not meet the set commissioning target, change the setting information of the control variable by the observation variable obtained in the iterative commissioning;

[0156] The second generating unit is configured to apply the changed setting information of the control variable to the virtual particle accelerator, so that the virtual particle accelerator performs iterative commissioning on the particle accelerator according to the changed setting information of the control variable, until an iteration termination condition is met in the iterative commissioning process, and generate a second beam commissioning result according to a current beam state output by the virtual particle accelerator;

[0157] The verifying unit is configured to, according to the current beam state output by the virtual particle accelerator, apply the setting information of the corresponding control variable in the virtual particle accelerator to the actual particle accelerator, to verify the second beam commissioning result by the current beam state output by the actual particle accelerator.

[0158] It should be noted that other corresponding descriptions of the functions of the particle accelerator beam commissioning device provided in the embodiment can be referred to the corresponding descriptions in the Figures 1-8 , which will not be described here in detail.

[0159] Based on the method as shown in Figures 1-8 , correspondingly, the embodiment of the present application further provides a storage medium having a computer program stored thereon, and the program is executed by a processor to implement the particle accelerator beam commissioning method as shown in Figures 1-8 .

[0160] Based on such 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, a U disk, a mobile hard disk, etc.), and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the method described in each implementation scenario of the present application.

[0161] Based on the method as shown in Figures 1-8 , and Figure 9To achieve the above object, the virtual device embodiment shown provides an entity device for beam commissioning of a particle accelerator, which can be a computer, a smart phone, a tablet computer, a smart watch, a server, a network device, or the like, and 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 commissioning method of a particle accelerator. Figures 1-8 The particle accelerator beam commissioning method shown.

[0162] Optionally, the entity device can further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, and the like. The user interface can include a display screen (Display), an input unit such as a keyboard (Keyboard), and the like. The optional user interface can further include a USB interface, a card reader interface, and the like. The network interface can optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), and the like.

[0163] In the example embodiment, referring to Figure 10 The above-mentioned entity device includes a communication bus, a processor, a memory, and a communication interface, and can further include an input / output interface and a display device. The various functional units can communicate with each other through the bus. The memory stores a computer program. The processor is used to execute the program stored in the memory to execute the above-mentioned beam commissioning method of a particle accelerator.

[0164] Those skilled in the art can understand that the structure of the entity device for beam commissioning of a particle accelerator provided by the embodiment does not constitute a limitation on the entity device, and can include more or fewer components, or combine certain components, or different component arrangements.

[0165] The storage medium can further include an operating system and a network communication module. The operating system is a program for managing hardware and software resources of the above-mentioned entity device for beam commissioning of a particle accelerator, and supports the running of information processing programs and other software and / or programs. The network communication module is used to realize communication between components inside the storage medium, and communication with other hardware and software in the information processing entity device.

[0166] Through the description of the above embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software and necessary general hardware platforms, or by hardware. By applying the technical solutions of the present application, compared with the existing mode, the accelerator beam commissioning is divided into a stage-solidified reproducible process, so that each process can be directly connected to an actual particle accelerator and / or a virtual particle accelerator on the basis of setting a commissioning target, which can not only complete the early parameter prediction by using a virtual accelerator, but also quickly land to actual device verification without additional adaptation process, reduces the risk of trial and error of the actual device, can let each iteration process have clear data support, can monitor the process progress and effect in real time, and ensures the accuracy of the beam commissioning result.

[0167] Those skilled in the art can understand that the drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the drawings are not necessarily required for implementing the present application. Those skilled in the art can understand that the modules in the device in the implementation scenario can be distributed in the device in the implementation scenario according to the description of the implementation scenario, or can be changed and located in one or more devices different from the implementation scenario. The modules of the above implementation scenario can be combined into one module, or can be further split into multiple sub-modules.

[0168] The above application serial numbers are only for description, and do not represent the advantages and disadvantages of the implementation scenario. The above disclosure is only a few specific implementation scenarios of the present application, but the present application is not limited thereto, and any changes that those skilled in the art can think of should fall within the protection scope of the present application.

Claims

1. A method of beam commissioning of a particle accelerator, characterized by, The method comprises the following steps: creating process configuration information of a particle accelerator according to a beam commissioning task, wherein the process configuration information at least comprises variable information and algorithm information used for process configuration of the particle accelerator, and the variable information at least comprises observation variables and control variables; on the basis of setting a commissioning target, iteratively commissioning the particle accelerator by using the process configuration information to generate setting information of the control variables according to observation variables obtained in each iteration; applying the setting information of the control variables to an actual particle accelerator and / or a virtual particle accelerator to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator, wherein the current beam state is achieved by using the actual particle accelerator to obtain real beam data to obtain standardized actual observation variables, changing the setting information of the control variables according to the standardized actual observation variables, applying the changed setting information of the control variables to the virtual particle accelerator, and iteratively commissioning the virtual particle accelerator according to the changed setting information of the control variables to obtain the current beam state, wherein the virtual particle accelerator is a digital twin model of a simulated actual accelerator operating state obtained according to deviation data of a beam commissioning process; if an iteration commissioning process meets an iteration termination condition, generating a first beam commissioning result according to the current beam state, otherwise, repeatedly executing the iteration commissioning process until the iteration commissioning process meets the iteration termination condition, wherein the iteration termination condition is that the current beam state meets the set commissioning target and / or the number of iterations reaches an iteration threshold.

2. The beam commissioning method of a particle accelerator according to claim 1, wherein, Before the step of creating the process configuration information of the particle accelerator according to the beam commissioning task, the method further comprises: constructing basic configuration information of the particle accelerator, wherein the basic configuration information at least comprises element information and structure information used for basic configuration of the particle accelerator; correspondingly, the step of creating the process configuration information of the particle accelerator according to the beam commissioning task comprises: creating the process configuration information of the particle accelerator according to the beam commissioning task in an accelerator structure constructed by using the basic configuration information.

3. The beam commissioning method of a particle accelerator according to claim 2, wherein, The step of creating the process configuration information of the particle accelerator according to the beam commissioning task in the accelerator structure constructed by using the basic configuration information comprises: determining element information of the particle accelerator in the accelerator structure constructed by using the basic configuration information; performing role configuration on the element information of the particle accelerator according to the beam commissioning task to obtain variable information of the particle accelerator in a commissioning process; performing algorithm configuration on the variable information of the particle accelerator in the commissioning process according to the beam commissioning task to obtain algorithm information of the particle accelerator in the commissioning process.

4. The beam commissioning method of a particle accelerator according to claim 1, wherein, The step of iteratively commissioning the particle accelerator by using the process configuration information on the basis of setting the commissioning target to generate setting information of the control variables according to observation variables obtained in each iteration comprises: on the basis of setting the commissioning target, determining particle beam parameters injected by the accelerator and an objective function of an optimization algorithm. iteratively debug the particle accelerator using the process configuration information to obtain state data of the particle beam parameters in the observation variables during the acceleration process; generate setting information of the control variables according to a quantitative deviation formed by the state data of the observation variables and an objective function of the optimization algorithm.

5. The beam commissioning method of a particle accelerator according to any one of claims 1 to 4, characterized in that, Before the setting information of the control variables is applied to the actual particle accelerator and / or the virtual particle accelerator to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator, the method further comprises: performing model training according to the deviation data of the beam commissioning process to obtain a virtual particle accelerator, the virtual particle accelerator being used to simulate a running state of the actual particle accelerator in a virtual environment.

6. The beam commissioning method of a particle accelerator according to claim 5, wherein, The model training according to the deviation data of the beam commissioning process to obtain the virtual particle accelerator comprises: pre-constructing a digital twin model in a virtual environment, the digital twin model being constructed based on basic configuration information of the actual particle accelerator; applying the setting information of the control variables to the actual particle accelerator and the digital twin model respectively to obtain deviation data of the beam commissioning process by comparing a current beam state output by the actual particle accelerator with a current beam state output by the digital twin model; training the digital twin model according to the deviation data of the beam commissioning process to obtain the virtual particle accelerator.

7. The beam commissioning method of a particle accelerator according to any one of claims 1 to 4, characterized in that, After iteratively debugging the particle accelerator using the process configuration information to generate setting information of the control variables according to observation variables obtained in each iteration of the debugging on the basis of the set commissioning target, the method further comprises: applying the setting information of the control variables to the actual particle accelerator to obtain a current beam state output by the actual particle accelerator; if the intermediate beam state output by the actual particle accelerator does not meet the set commissioning target, changing the setting information of the control variables by the observation variables obtained through the iteration of the debugging; applying the changed setting information of the control variables to the virtual particle accelerator to enable the virtual particle accelerator to iteratively debug the particle accelerator according to the changed setting information of the control variables until an iteration termination condition is met, and to generate a second beam commissioning result according to a current beam state output by the virtual particle accelerator; applying the setting information of the corresponding control variables in the virtual particle accelerator to the actual particle accelerator according to the current beam state output by the virtual particle accelerator to verify the second beam commissioning result by the current beam state output by the actual particle accelerator.

8. A beam commissioning device for a particle accelerator, characterized by, comprises: a creating unit configured to create process configuration information of a particle accelerator according to a beam commissioning task, the process configuration information at least comprising variable information and algorithm information used for process configuration of the particle accelerator, the variable information at least comprising observation variables and control variables; a debugging unit configured to iteratively debug the particle accelerator using the process configuration information to generate setting information of the control variables according to observation variables obtained in each iteration of the debugging on the basis of a set commissioning target; The first obtaining unit is configured to apply the setting information of the control variable to an actual particle accelerator and / or a virtual particle accelerator to obtain a current beam state output by the actual particle accelerator and / or the virtual particle accelerator, wherein the current beam state is obtained by using the actual particle accelerator to obtain real beam data to obtain a standardized actual observation variable, changing the setting information of the control variable according to the standardized actual observation variable, applying the changed setting information of the control variable to the virtual particle accelerator, and performing iterative debugging on the virtual particle accelerator according to the changed setting information of the control variable to obtain the current beam state, wherein the virtual particle accelerator is a digital twin model of a simulated actual accelerator operating state obtained according to deviation data of a beam debugging process. The first generating unit is configured to generate a first beam debugging result according to the current beam state if the iterative debugging process meets an iteration termination condition, or repeatedly perform the iterative debugging process until the current beam state meets the iteration termination condition, wherein the iteration termination condition is that the current beam state meets the set debugging target and / or the number of iterations reaches an iteration threshold. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The processor implements the steps of the beam debugging method of the particle accelerator according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program implements the steps of the beam debugging method of the particle accelerator according to any one of claims 1 to 7 when executed by the processor.

Images