A control method for an ion cyclotron resonance heating system

Abstract

This invention discloses a control method for an ion cyclotron resonance heating system, belonging to the field of radio frequency heating control technology. This invention trains a coupling impedance prediction model using ICRH system operating parameters and plasma parameters, and outputs future coupling impedance prediction results during ICRH system operation to characterize the changing trend of the coupling state between the ICRH system and the plasma. During ICRH system operation, the signals required for the coupling impedance prediction model are acquired in real time to predict the matching state of the ICRH system. When the prediction results indicate a potential mismatch in the future coupling impedance, the impedance matching module in the ICRH system is preemptively adjusted to ensure that the ICRH system is always in an optimal matching state. This solves the technical problems of adjustment lag and transient mismatch inherent in existing technologies.

Description

Technical Field

[0001] This invention belongs to the field of radio frequency heating control technology, and more specifically, relates to a control method for an ion cyclotron resonance heating system. Background Technology

[0002] In magnetic confinement fusion research devices, ion cyclotron resonance heating (ICRH) systems serve as crucial auxiliary heating methods, widely applied in plasma heating and operational parameter control. An ICRH system typically comprises a power source module, transmission line module, impedance matching module, and antenna section. Its operational efficiency and safety depend to some extent on variations in the coupling impedance between the antenna and the plasma. These impedance changes can lead to increased reflected power and elevated standing wave voltage in the transmission line, thereby affecting system reliability. Therefore, ICRH systems must be able to adapt to variations in coupling impedance; hence, this study investigates a control method for ion cyclotron resonance heating systems.

[0003] Existing ICRH system control methods typically adjust the impedance matching system by combining theory and operational experience, or by monitoring parameters such as reflected power, transmission line voltage, or VSWR for real-time feedback adjustment to achieve impedance matching and system protection. However, matching adjustment methods relying on feedback signals often suffer from adjustment lag when facing changes in coupling impedance, easily leading to transient mismatch problems. This can result in increased reflected power, increased transmission line voltage, or even triggering system protection, thereby affecting heating efficiency and stable operation of the device. Summary of the Invention

[0004] In view of the above-mentioned defects or improvement needs of the prior art, the present invention provides a control method for an ion cyclotron resonance heating system to solve the technical problems of adjustment lag and transient mismatch in the prior art.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a control method for an ion cyclotron resonance heating system, comprising: At the present moment Next, obtain from Time's up The ICRH system operating parameters and plasma parameters at each time point are input into a pre-trained coupling impedance prediction model to obtain... Predicted results of the coupling impedance between the ICRH system and the plasma at the specified time. ; Preset time step; when and When the absolute value of the difference between the two exceeds a preset threshold, it is determined that the ICRH system has a mismatch trend. The impedance matching module in the ICRH system is then adjusted to make the coupling impedance between the ICRH system and the plasma equal to that between the two systems. Matching; The coupling impedance prediction model is obtained by training a deep learning model using a pre-collected training set. The training set includes training samples at different sampling times. The training samples below include: Time's up The ICRH system operating parameters and plasma diagnostic parameters at any given time, and the corresponding labels include: The actual value of the coupling impedance between the ICRH system and the plasma at that time.

[0006] More preferably, the sampling time The labels corresponding to the training samples below also include Empirical value of the coupling impedance between ICRH and plasma at a given time; The training objectives of the above coupling impedance prediction model include: minimizing the weighted sum of prediction loss and empirical loss; The prediction loss is the sum of the difference between the predicted coupling impedance of each training sample in the training set and its actual coupling impedance value. The empirical loss is the sum of the difference between the predicted coupling impedance of each training sample in the training set and its empirical value. in, Empirical value of coupling impedance between ICRH system and plasma at time t. for:

[0007] These are preset coefficients; This refers to the main parallel wavenumber of the antenna in the ICRH system. for The distance of the vanishing layer at any given moment is specifically... ; This refers to the distance between the antenna and the plasma center in the ICRH system. for The fast wave cutoff point of the ICRH system at a given time is calculated using the following formula:

[0008]

[0009]

[0010] This represents the number of particle species in the ICRH system. For the first in the ICRH system The plasma frequency of the particles; The radio frequency of the ICRH system; For the first in the ICRH system The cyclotron frequency of a particle; The parallel refractive index corresponding to the ICRH system; for At this moment, the first in the ICRH system The particle at a distance from the plasma center Density at that location; For the first in the ICRH system The electric charge of the particles; For dielectric tensor; For the first in the ICRH system The mass of the particle; for At this moment, the first in the ICRH system The particle at a distance from the plasma center The magnitude of the magnetic field at that location.

[0011] More preferably, the training set includes training samples at different sampling times when the ICRH system is under different operating conditions.

[0012] More preferably, the ICRH system operating parameters include one or more of the following: incident power, reflected power, transmission line voltage, and transmission line current of the ICRH system.

[0013] More preferably, the plasma parameters include: plasma string average density; the plasma parameters also include: plasma horizontal displacement, plasma vertical displacement and one or more of plasma property parameters; the plasma property parameters include plasma current, plasma ring voltage and plasma magnetic field.

[0014] More preferably, the coupling impedance prediction model includes: a time-series feature extraction module, a spatial feature extraction module, and a mapping module; The temporal feature extraction module is used to extract features from the input ICRH system operating parameters and plasma parameters, excluding the plasma string average density. The spatial feature extraction module is used to extract features of the plasma string average density from the input plasma parameters; The mapping module is used to fuse the features output by the temporal feature extraction module and the spatial feature extraction module, and then map them to the corresponding prediction results of the coupling impedance.

[0015] More preferably, the temporal feature extraction module is an LSTM model; the spatial feature extraction module is a CNN model.

[0016] In a second aspect, the present invention provides a control device for an ion cyclotron resonance heating system, comprising: The prediction module is used to predict the current time. Next, obtain from Time's up The ICRH system operating parameters and plasma parameters at each time point are input into a pre-trained coupling impedance prediction model to obtain... Predicted results of the coupling impedance between the ICRH system and the plasma at the specified time. ; Preset time step; Adjustment module, used when and When the absolute value of the difference between the two exceeds a preset threshold, it is determined that the ICRH system has a mismatch trend. The impedance matching module in the ICRH system is then adjusted to make the coupling impedance between the ICRH system and the plasma equal to that between the two systems. Matching; The coupling impedance prediction model is obtained by training a deep learning model using a pre-collected training set. The training set includes training samples at different sampling times. The training samples below include: Time's up The ICRH system operating parameters and plasma diagnostic parameters at any given time are labeled as follows: The actual value of the coupling impedance between the ICRH system and the plasma at that time.

[0017] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory stores a computer program, and the processor executes the control method provided in the first aspect of the present invention when executing the computer program.

[0018] Fourthly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when executed by a processor, controls the device containing the storage medium to perform the control method provided in the first aspect of the present invention.

[0019] Fifthly, the invention also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the control method provided in the first aspect of the invention.

[0020] In summary, the above-described technical solutions conceived in this invention can achieve the following beneficial effects: 1. This invention provides a control method for an ion cyclotron resonance heating system. A corresponding coupling impedance prediction model is constructed based on the ICRH coupling impedance prediction task. The model is trained using ICRH system operating parameters and plasma parameters. During ICRH system operation, the predicted future coupling impedance is output to characterize the changing trend of the coupling state between the ICRH system and the plasma. During ICRH system operation, the signals required for the coupling impedance prediction model are acquired in real time to predict the matching state of the ICRH system. When the prediction results indicate a potential mismatch in the future coupling impedance, the impedance matching module in the ICRH system is preemptively adjusted to ensure the ICRH system is always in an optimal matching state. This solves the technical problems of adjustment lag and transient mismatch inherent in existing technologies.

[0021] 2. The control method for the ion cyclotron resonance heating system provided by the present invention has the following training objectives for the coupling impedance prediction model: minimizing the weighted sum of prediction loss and empirical loss; and through empirical value constraints, enabling the coupling impedance prediction model to learn more physical characteristics of coupling impedance, thereby further improving the generalization of the coupling impedance prediction model.

[0022] 3. Furthermore, in the control method of the ion cyclotron resonance heating system provided by the present invention, the training set includes training samples at different sampling times when the ICRH system is under different operating conditions, so that the coupling impedance prediction model can learn the changing characteristics of coupling impedance under different operating conditions, and the model has strong generalization and robustness.

[0023] 4. Furthermore, in the control method of the ion cyclotron resonance heating system provided by this invention, the coupling impedance prediction model includes: a time-series feature extraction module, a spatial feature extraction module, and a mapping module. Considering that the average density of the plasma string is spatially distributed and covers the entire plasma space, the density near the antenna region has the greatest impact on the coupling impedance, so the spatial feature extraction module is used to extract its spatial features. The characteristics of the ICRH system operating parameters and the other parameters in the plasma parameters, except for the average density of the plasma string, are mainly reflected in the changes over time, so the time-series feature extraction module is used to extract their time-series features. This enables the model to learn the representative features of different input parameters in a targeted manner, further improving the accuracy of model prediction and thus improving control precision.

[0024] 5. The control method for the ion cyclotron resonance heating system provided by this invention utilizes the coupling impedance prediction result as target information to guide the adjustment behavior of the impedance matching module in the ICRH system. This allows the impedance matching module to obtain a favorable adjustment direction and reference information before feedback adjustment intervenes. By combining the predicted target information with feedback control, the problem of feedback control response lag can be alleviated, and transient mismatch caused by rapid changes in coupling state can be reduced. This helps to suppress adverse operating conditions such as sudden increases in reflected power and transmission line overvoltage, thereby improving the stability and safety margin of system operation.

[0025] 6. In the control method for the ion cyclotron resonance heating system provided by this invention, the coupling impedance prediction model structure has good scalability and maintainability. New input parameters can be flexibly introduced or the model structure adjusted and optimized according to changes in the ICRH system's operational requirements and plasma observation conditions, without requiring a complete reconfiguration of the overall control architecture. Simultaneously, the introduction of the coupling impedance prediction model helps reduce the system's reliance on human experience, making the coupling state assessment and matching adjustment process more intelligent and automated, thereby improving the repeatability and long-term stability of the ICRH system. Attached Figure Description

[0026] Figure 1 The flowchart shows the control method of the ion cyclotron resonance heating system provided in Embodiment 1 of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0028] Example 1 A control method for an ion cyclotron resonance heating system includes: S1, at the current moment Next, obtain from Time's up The ICRH system operating parameters and plasma parameters at each time point are input into a pre-trained coupling impedance prediction model to obtain... Predicted results of the coupling impedance between the ICRH system and the plasma at the specified time. ; The preset time step is set to 10ms in one optional implementation.

[0029] S2, when and When the absolute value of the difference between the two exceeds a preset threshold, it is determined that the ICRH system has a mismatch trend. The impedance matching module in the ICRH system is then adjusted to make the coupling impedance between the ICRH system and the plasma equal to that between the two systems. If the impedance matches, then the ICRH system is determined to have no mismatch trend, and the impedance matching module in the ICRH system is not adjusted. In one optional implementation, the above preset threshold value is taken as follows: .

[0030] The coupling impedance prediction model is obtained by training a deep learning model using a pre-collected training set. The training set includes training samples at different sampling times. The training samples below include: Time's up The ICRH system operating parameters and plasma diagnostic parameters at any given time are labeled as follows: The actual value of the coupling impedance between the ICRH system and the plasma at that time.

[0031] This embodiment constructs a corresponding coupling impedance prediction model based on the ICRH coupling impedance prediction task. This model characterizes the intrinsic mapping relationship between ICRH system operating parameters, plasma parameters, and the coupling impedance between the ICRH system and the plasma. By comprehensively modeling the key factors affecting coupling impedance, this coupling impedance prediction model possesses good generalization ability and robustness, maintaining stable prediction performance under different plasma states and parameter perturbation conditions, thus providing a reliable basis for ICRH system operation decisions.

[0032] This embodiment constructs a corresponding coupling impedance prediction model based on the ICRH coupling impedance prediction task. The model is trained using ICRH system operating parameters and plasma parameters, and future coupling impedance prediction results are output during ICRH system operation to characterize the changing trend of the coupling state between the ICRH system and the plasma. Simultaneously, during ICRH system operation, the signals required for the coupling impedance prediction model are acquired in real time to predict the matching state of the ICRH system. When the prediction results indicate a potential mismatch in future coupling impedance, the impedance matching module in the ICRH system is preemptively adjusted to ensure the ICRH system is always in an optimal matching state.

[0033] Preferably, in one optional implementation, the training set includes training samples at different sampling times when the ICRH system is under different operating conditions.

[0034] This embodiment collects the training set in the following way: Multiple samples were taken during ICRH system operation or historical discharges under different operating conditions to obtain ICRH system operating parameters and plasma parameters related to ICRH coupling state at each sampling time. The collected parameters were then preprocessed, including one or more of the following: time alignment, anomaly removal, normalization, and resampling. For each sampling time... ,Will Time's up The ICRH system operating parameters and plasma diagnostic parameters at each sampling time within this time period are used as the sampling time. The training samples are used to construct a training set that includes training samples from multiple sampling times.

[0035] The coupling impedance prediction model is trained using a training set, enabling it to learn the characteristics of coupling impedance variation under different operating conditions.

[0036] In one alternative implementation, the training objective of the above-mentioned coupling impedance prediction model includes minimizing the sum of the difference loss between the predicted coupling impedance of each training sample in the training set and its actual coupling impedance value.

[0037] In an alternative implementation method two, the sampling time The labels corresponding to the training samples below also include Empirical value of the coupling impedance between the ICRH system and the plasma at a given time; The training objectives of the above coupling impedance prediction model include: minimizing the weighted sum of prediction loss and empirical loss; The prediction loss is the sum of the difference between the predicted coupling impedance of each training sample in the training set and its actual coupling impedance value. The empirical loss is the sum of the difference between the predicted coupling impedance of each training sample in the training set and its empirical value. in, Empirical value of coupling impedance between ICRH system and plasma at time t. for:

[0038] These are preset coefficients, which are related to the ICRH system. This refers to the main parallel wavenumber of the antenna in the ICRH system. for The distance of the vanishing layer at any given moment is specifically... ; This refers to the distance between the antenna and the plasma center in the ICRH system. for The fast wave cutoff point of the ICRH system at a given time is calculated using the following formula:

[0039]

[0040]

[0041] This represents the number of particle species in the ICRH system. For the first in the ICRH system The plasma frequency of the particles; The radio frequency of the ICRH system; For the first in the ICRH system The cyclotron frequency of a particle; The parallel refractive index corresponding to the ICRH system is a constant; for At this moment, the first in the ICRH system The particle at a distance from the plasma center Density at that location; For the first in the ICRH system The electric charge of the particles; For dielectric tensor; For the first in the ICRH system The mass of the particle; for At this moment, the first in the ICRH system The particle at a distance from the plasma center The magnitude of the magnetic field at that location.

[0042] In the above optional implementation method 2, by using empirical value constraints, the coupling impedance prediction model can learn more physical characteristics of the coupling impedance, thereby further improving the generalization of the model.

[0043] In one alternative implementation, the ICRH system operating parameters include one or more of the following: incident power, reflected power, transmission line voltage, and transmission line current of the ICRH system.

[0044] In one optional implementation, the plasma parameters include: plasma string average density; the plasma parameters also include: plasma horizontal displacement, plasma vertical displacement and one or more of plasma property parameters; the plasma property parameters include plasma current, plasma loop voltage and plasma magnetic field.

[0045] It should be noted that the corresponding input parameters should be selected according to the target characteristics. For example, for the ICRH system part, the ICRH system reflected power signal and transmission line voltage signal should be selected; for the plasma part, the electron density chord integral signal (the plasma chord average density is obtained by converting the electron density chord integral signal) in the low field measurement area of ​​the far-infrared three-wave polarimeter interferometer system should be selected, and the relative error between the model's predicted coupling impedance and the actual coupling impedance on the test set should be less than 20%.

[0046] It should be noted that the coupling impedance prediction model is a deep learning model, which can be a multilayer perceptron model, a convolutional neural network model, a long short-term memory network model, a gated recurrent unit model, or a hybrid model formed by combining the above models, etc., without limitation. Appropriate hyperparameters should be selected to improve the accuracy and stability of the coupling impedance prediction results, including the number of layers in the prediction model, the type and number of neurons in each layer, the number of training epochs, the learning rate, the form of the loss function, and the type of optimizer.

[0047] Preferably, in one optional implementation, the coupling impedance prediction model includes: a time-series feature extraction module, a spatial feature extraction module, and a mapping module; The temporal feature extraction module is used to extract features from the input ICRH system operating parameters and plasma parameters, excluding the plasma string average density. The spatial feature extraction module is used to extract features of the plasma string average density from the input plasma parameters; The mapping module is used to fuse the features output by the temporal feature extraction module and the spatial feature extraction module, and then map them to the corresponding ICRH coupling impedance prediction results.

[0048] Considering that the average plasma string density is a spatial distribution covering the entire plasma space, the density near the antenna region has the greatest impact on the coupling impedance. Therefore, a spatial feature extraction module is used to extract its spatial features. Meanwhile, the characteristics of the ICRH system operating parameters and other plasma parameters besides the average plasma string density are mainly reflected in temporal changes. Therefore, a temporal feature extraction module is used to extract their temporal features. This allows the model to learn representative features of different input parameters in a targeted manner, further improving the accuracy of model predictions and thus improving control precision.

[0049] It should be noted that the aforementioned time-series feature extraction module can be an LSTM model, a Bi-LSTM model, an RNN model, etc., and is not limited here. Preferably, in one optional implementation, the aforementioned time-series feature extraction module is an LSTM model, which processes the input parameters with time characteristics.

[0050] It should be noted that the aforementioned spatial feature extraction module can be a CNN model, a Transformer model, a spatial attention model, etc., and is not limited here. Preferably, in one optional implementation, the aforementioned temporal feature extraction module is a CNN model, which processes input parameters with spatial characteristics.

[0051] It should be noted that the above mapping module is a regression module, which can be a fully connected layer, LSTM, etc. There is no limitation here, but a fully connected layer is preferred.

[0052] It should be noted that there are multiple ways to fuse the features output by the temporal feature extraction module and the spatial feature extraction module. These methods include concatenation along the channel dimension, weighted summation, and gating mechanism fusion. No specific method is specified here. The preferred method is concatenation along the channel dimension.

[0053] In S2 above, when the ICRH system exhibits a mismatch trend, the impedance matching module in the ICRH system is adjusted to make the coupling impedance between the ICRH system and the plasma match... Matching.

[0054] The specific adjustment method of the impedance matching module is determined by the type of impedance matching module. The impedance matching device in the impedance matching module can be a two-stub impedance matching device, a three-stub impedance matching device, etc., which is not limited here. This embodiment takes a two-stub impedance matching device as an example, in order to make the coupling impedance between the ICRH system and the plasma equal to that of the plasma, the impedance matching device is adjusted accordingly. Matching can be achieved by adjusting the equivalent electrical length of the stub, specifically by adjusting the capacitor in the stub impedance matching device. By predicting the changing trend of the coupling impedance, the direction and magnitude of capacitor adjustment, as well as the priority of adjusting which stub first, are determined. The magnitude and priority of the adjustment are dynamically adjusted according to the system operating environment and the required adjustment accuracy, thereby achieving precise matching and improving system stability and response speed.

[0055] These adjustments effectively correct coupling mismatches caused by changes in plasma state, ensuring the ICRH system remains in optimal matching condition. In this process, feedback control and prediction results work together to guarantee rapid and accurate adjustment. The synergistic adjustment method, combining the coupling impedance prediction model with the ICRH system, effectively improves system response speed and adjustment accuracy.

[0056] This embodiment, by predicting the coupling impedance at the next moment, can determine a clear adjustment direction, enabling the ICRH system to respond quickly and take effective measures when faced with changes in coupling state. Specifically, when a mismatch trend is expected in the future, the adjustment behavior of the impedance matching module in the ICRH system can be controlled in advance, thereby controlling the coupling state of the ICRH system, adjusting the system matching in advance, ensuring that the system is effectively adjusted before changes in coupling state occur, avoiding transient mismatch, and improving the stability of the ICRH system.

[0057] In summary, this embodiment, by integrating a coupling impedance prediction model and a real-time feedback control mechanism, can efficiently adjust the coupling state of the ICRH system. The method includes: predicting the coupling impedance between the ICRH system and the plasma at the next moment using the coupling impedance prediction model, thereby predicting the changing trend of the coupling state and providing forward-looking information for the ICRH system; when the prediction result indicates a mismatch trend in the ICRH system, adjusting the impedance matching module in the ICRH system to match the coupling impedance between the ICRH system and the plasma at the next moment. This method can control the adjustment behavior of the impedance matching model in the ICRH system in advance, thereby avoiding mismatch and ensuring that the ICRH system always maintains an optimal matching state. At the same time, this embodiment also has advantages such as simple design, low development difficulty, strong maintainability, and reduced manpower.

[0058] Compared with the traditional matching adjustment method that relies solely on real-time feedback signals, this embodiment achieves a forward-looking assessment of the coupling state, enabling the matching system to provide a basis for adjustment before significant changes occur in the coupling state, thereby improving the initiative and predictability of the overall adjustment process of the ICRH system.

[0059] Example 2 A control device for an ion cyclotron resonance heating system includes: The prediction module is used to predict the current time. Next, obtain from Time's up The ICRH system operating parameters and plasma parameters at each time point are input into a pre-trained coupling impedance prediction model to obtain... Predicted results of the coupling impedance between the ICRH system and the plasma at the specified time. ; Preset time step; Adjustment module, used when and When the absolute value of the difference between the two exceeds a preset threshold, it is determined that the ICRH system has a mismatch trend. The impedance matching module in the ICRH system is then adjusted to make the coupling impedance between the ICRH system and the plasma equal to that between the two systems. If the impedance matches, then the ICRH system is deemed to have no mismatch trend, and the impedance matching module in the ICRH system is not adjusted. The coupling impedance prediction model is obtained by training a deep learning model using a pre-collected training set. The training set includes training samples at different sampling times. The training samples below include: Time's up The ICRH system operating parameters and plasma diagnostic parameters at any given time are labeled as follows: The actual value of the coupling impedance between the ICRH system and the plasma at that time.

[0060] The relevant technical solutions are the same as those in Embodiment 1 of this invention, and will not be repeated here.

[0061] Example 3 An electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the control method provided in Embodiment 1 of the present invention.

[0062] The relevant technical solutions are the same as those in Embodiment 1 of this invention, and will not be repeated here.

[0063] Example 4 A computer-readable storage medium includes a stored computer program, wherein when the computer program is run by a processor, it controls the device where the storage medium is located to execute the control method provided in Embodiment 1 of the present invention.

[0064] The relevant technical solutions are the same as those in Embodiment 1 of this invention, and will not be repeated here.

[0065] Example 5 A computer program product includes a computer program / instructions that, when executed by a processor, implement the control method provided in Embodiment 1 of the present invention.

[0066] The relevant technical solutions are the same as those in Embodiment 1 of this invention, and will not be repeated here.

[0067] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A control method for an ion cyclotron resonance heating system, characterized in that, include: At the present moment Next, obtain from Time's up The ICRH system operating parameters and plasma parameters at each time point are input into a pre-trained coupling impedance prediction model to obtain... Predicted results of the coupling impedance between the ICRH system and the plasma at the specified time. ; Preset time step; when and When the absolute value of the difference between the two exceeds a preset threshold, it is determined that the ICRH system has a mismatch trend. The impedance matching module in the ICRH system is then adjusted to make the coupling impedance between the ICRH system and the plasma equal to that between the two systems. Matching; The coupling impedance prediction model is obtained by training a deep learning model using a pre-collected training set; the training set includes training samples at different sampling times; and the sampling times... The training samples below include: Time's up The ICRH system operating parameters and plasma diagnostic parameters at any given time, and the corresponding labels include: The actual value of the coupling impedance between the ICRH system and the plasma at that time.

2. The control method for the ion cyclotron resonance heating system according to claim 1, characterized in that, Sampling time The labels corresponding to the training samples below also include Empirical value of the coupling impedance between the ICRH system and the plasma at a given time; The training objective of the coupling impedance prediction model includes minimizing the weighted sum of the prediction loss and the empirical loss. The prediction loss is the sum of the difference between the predicted coupling impedance of each training sample in the training set and its actual coupling impedance value. The empirical loss is the sum of the difference loss between the predicted coupling impedance of each training sample in the training set and its empirical value. in, Empirical value of coupling impedance between ICRH system and plasma at time t. for: These are preset coefficients; This refers to the main parallel wavenumber of the antenna in the ICRH system. for The distance of the vanishing layer at any given moment is specifically... ; This refers to the distance between the antenna and the plasma center in the ICRH system. for The fast wave cutoff point of the ICRH system at a given time is calculated using the following formula: This represents the number of particle species in the ICRH system. For the first in the ICRH system The plasma frequency of the particles; The radio frequency of the ICRH system; For the first in the ICRH system The cyclotron frequency of a particle; The parallel refractive index corresponding to the ICRH system; for At this moment, the first in the ICRH system The particle at a distance from the plasma center Density at that location; For the first in the ICRH system The electric charge of the particles; For dielectric tensor; For the first in the ICRH system The mass of the particle; for At this moment, the first in the ICRH system The particle at a distance from the plasma center The magnitude of the magnetic field at that location.

3. The control method for the ion cyclotron resonance heating system according to claim 1, characterized in that, The training set includes training samples at different sampling times when the ICRH system is under different operating conditions.

4. The control method for the ion cyclotron resonance heating system according to claim 1, characterized in that, The ICRH system operating parameters include one or more of the following: incident power, reflected power, transmission line voltage, and transmission line current of the ICRH system.

5. The control method for the ion cyclotron resonance heating system according to claim 1, characterized in that, The plasma parameters include: plasma string average density; the plasma parameters also include: plasma horizontal displacement, plasma vertical displacement and one or more of plasma property parameters; the plasma property parameters include plasma current, plasma loop voltage and plasma magnetic field.

6. The control method for the ion cyclotron resonance heating system according to claim 5, characterized in that, The coupling impedance prediction model includes: a time-series feature extraction module, a spatial feature extraction module, and a mapping module; The time-series feature extraction module is used to extract features from the input ICRH system operating parameters and plasma parameters, excluding the plasma string average density. The spatial feature extraction module is used to extract the features of the average density of plasma strings in the input plasma parameters. The mapping module is used to fuse the features output by the time-series feature extraction module and the spatial feature extraction module, and then map them to the corresponding prediction results of the coupling impedance.

7. The control method for the ion cyclotron resonance heating system according to claim 6, characterized in that, The temporal feature extraction module is an LSTM model; the spatial feature extraction module is a CNN model.

8. An electronic device, characterized in that, include: A memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the control method according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein when the computer program is run by a processor, it controls the device where the storage medium is located to perform the control method according to any one of claims 1-7.

10. A computer program product, characterized in that, It includes a computer program / instruction that, when executed by a processor, implements the control method according to any one of claims 1-7.

Images