A strip-shaped electron beam gun design method based on intelligent collaborative optimization

Abstract

This invention belongs to the field of microwave vacuum electronic device design technology, specifically relating to a strip-beam electron gun design method based on intelligent collaborative optimization. The method includes: S1, selecting an initial model from a database according to the design objectives; S2, establishing the model and extracting electron beam performance parameters through a CST and MATLAB co-simulation platform; S3, calculating the objective evaluation function; S4, optimizing the electron gun structural parameters using an intelligent collaborative optimization algorithm; S5, performing performance evaluation and redundancy analysis; and S6, outputting the final parameters and storing them in a database. This invention achieves full automation of the strip-beam electron gun design process. By intelligently adjusting the optimization strategy through algorithms, it significantly improves design efficiency and accuracy, effectively optimizes the laminar flow and uniformity of the electron beam, and supports both single-sided and double-sided compression strip-beam electron gun configurations.

Description

Technical Field

[0001] This invention belongs to the field of microwave electronics and vacuum electrotechnology, and specifically relates to a design method for a strip electron gun based on intelligent collaborative optimization. Background Technology

[0002] Vacuum electronic devices have broad application prospects in fields such as long-range imaging, radar, active denial systems, satellite communications, and deep space exploration. With the increasing demands for high power, high frequency, and high efficiency, vacuum electronic devices are gradually developing towards miniaturization and compactness. However, device miniaturization requires electron beam miniaturization, and traditional cylindrical electron beams face challenges in miniaturization, including increased space charge force, greater difficulty in electron beam focusing, and decreased power capacity. Therefore, using strip-shaped electron beams with a large aspect ratio has become an effective means to solve these problems.

[0003] Compared to cylindrical electron beams, strip electron beams significantly reduce the space charge force between electrons under the same operating conditions, thereby reducing the design complexity of the focusing system. They also offer higher current density, increasing the electron gun's current and output power. Furthermore, strip electron beams effectively reduce the current density at the emitting surface, extending the cathode material's lifespan and demonstrating significant performance advantages.

[0004] As the core component of a strip-beam device, the laminar flow properties of the electron beam it generates directly determine the device's performance. There are two main design methods: directly generating a strip-beam using an elliptical or rectangular cathode; and using a conventional electron gun to generate a circular electron beam, which is then compressed into a strip-beam using a magnetic field. Current designs show that the strip-beam obtained by compression after emission from a conventional electron gun is not ideal, and it also places high demands on the design structure, making it quite challenging.

[0005] Currently, the optimization of strip electron guns mainly relies on manual optimization. However, manual optimization struggles to achieve good laminar flow characteristics in the optimized strip electron gun, making it impossible to guarantee optimal operating conditions. Furthermore, it is time-consuming, inefficient, and difficult to correlate results from multiple designs. Therefore, developing a method to optimize strip electron gun design has significant technical and practical value. Summary of the Invention

[0006] To address the aforementioned problems or shortcomings, this invention proposes a design method for a strip electron gun based on intelligent collaborative optimization.

[0007] The design method for a strip electron gun based on intelligent collaborative optimization includes the following steps:

[0008] S1. Based on the input electron gun design objectives, including cathode voltage, current, narrow side radius, wide side radius, and narrow side and wide side dimensions of the electron injection waist, the existing model with the closest structure is automatically selected from the electron gun design database as the initial optimization model using a multi-objective programming algorithm.

[0009] S2. Based on the initial model, a three-dimensional model of the electron gun is automatically established using the CST and MATLAB co-simulation platform. The simulation mesh is adaptively divided according to the cathode and anode dimensions, with a finer mesh in key areas such as the cathode emitting surface and anode aperture, and a sparse mesh in uniform field regions. The CST is controlled by MATLAB to perform calculations, and the simulation results are read to extract the actual current I. act Note the narrow side dimension Wt act Width dimension Ww act Electron velocity distribution and density distribution parameters;

[0010] S3. Based on the extracted electron beam performance parameters, calculate the target evaluation function. This evaluation function consists of three parts: the target parameter function, the density distribution function, and the laminar flow function. The specific expression is as follows:

[0011] ;

[0012] In the formula, Fit opt Fit is a target parameter function used to evaluate the deviation between the macroscopic parameters of the electron beam and the design target. qua Fit is a density distribution function used to evaluate the uniformity of the electron beam density distribution. lam C is a laminarity function used to evaluate the laminarity of the electron beam trajectory; i It is a constant, where i=1,2,3, used to adjust the weights of each density parameter.

[0013] The expression for the objective parameter function is:

[0014] ;

[0015] In the formula, I act It is the actual simulated current, I exp It is the desired current, Wt act It is the actual width of the narrow side of the electronic injection waist, Wt exp It is the expected width of the narrow side of the electronic injection waist, Ww act This is the actual width of the electronic injection waistline, Ww exp It is the expected width of the electronic injection waist, C i It is a constant, where i = 4, 5, 6;

[0016] The density distribution function is as follows:

[0017] ;

[0018] In the formula, V pmean It is the average velocity ratio of electrons at the electron waist, V pmax It is the maximum velocity ratio of electrons at the electron waist, D n It is the narrow-side electron density ratio at the electron waist, D w It is the ratio of the electron width density at the electron waist, C i It is a constant, where i = 7, 8, 9, 10, used to adjust the weights of each density parameter;

[0019] The laminar flow function is as follows:

[0020] ;

[0021] In the formula, , representing the position Z of the trajectory of each electron. mi With the entire electronic injection waist position Z m Deviation; , representing the slope of each electron trajectory at the injection waist; C i It is a constant, where i=11,12, used to adjust the weights of each laminar flow index;

[0022] S4. Employing an intelligent collaborative optimization algorithm, the electron gun structural parameters are optimized using the calculated electron beam performance parameters. The intelligent collaborative optimization algorithm executes the following two stages sequentially:

[0023] S4.1, Fuzzy Control Coarse-tuning Stage: A fuzzy control parameter optimization system based on dynamic adjustment is used to perform preliminary optimization of the key structural parameters of the electron gun. This system includes the following core components:

[0024] S4.1.1, By defining dynamic adjustment coefficients Using formula Calculate the adaptive adjustment factor, where It decreases adaptively during the iteration process. This enables dynamic scaling of the error level, allowing the optimization system to automatically adjust its optimization strategy based on the convergence status.

[0025] S4.1.2 Define a 14-dimensional parameter adjustment space, corresponding to 14 key structural parameters of the electron gun: including cathode parabolic coefficients a1 and a2, cathode-anode distance l1, cathode-focusing electrode distance l2, focusing electrode thickness h1 and h2, focusing electrode opening size A1, B1, A5, and B5, focusing electrode interpolation point position P1, P2, P3, and P4, etc., to achieve comprehensive optimization of the electron gun structure;

[0026] S4.1.3. Employ an If-Else conditional statement structure, using the casecnt variable to control different optimization stages of the program, and dynamically select and adjust strategies based on the magnitude of performance error and convergence status, specifically including:

[0027] Current parameter optimization: Prioritize adjusting the anode-cathode spacing parameter. Based on fuzzy control rules, optimize the current parameter by adjusting the anode-cathode spacing parameter, and dynamically select the adjustment strategy according to the magnitude of the current deviation ΔI. When the current deviation is large, a large step adjustment strategy is adopted; when the current deviation is small, a small step fine adjustment strategy is adopted to ensure that the current parameter quickly converges to the desired value.

[0028] Focusing electrode adjustment: The structural parameters such as the opening size of the narrow and wide focusing electrodes are dynamically adjusted according to the injection waist size deviation ΔWt, ΔWw and density distribution; the shape of the focusing electrode is continuously optimized through the output variables ΔP and ΔF of the fuzzy controller, thereby improving the focusing effect of the electron beam.

[0029] Cathode adjustment: This involves comprehensively adjusting parameters such as the cathode parabolic coefficient and the cathode-focusing electrode spacing, based on the electron beam density distribution and laminar flow index Fit. lam Fine-tuning was performed, including optimizing the cathode curvatures a1 and a2, to improve electron emission characteristics and enhance laminar flow.

[0030] S4.2, Reinforcement Learning Fine-Tuning Stage: Using the optimal solution output by the fuzzy control in stage S4.1 as the initial state, the reinforcement learning algorithm is initiated for local fine-tuning. This algorithm uses the electron gun structural parameters as the action space and the target evaluation function obtained through CST simulation calculations and analysis using Matlab as the reward signal. The algorithm explores within the neighborhood of the optimal solution, autonomously learning and outputting parameter fine-tuning actions through continuous interactive trial and error with the CST-MATLAB co-simulation platform to maximize the cumulative reward (i.e., minimize the target evaluation function). This stage can discover complex parameter coupling adjustment patterns not covered by preset fuzzy rules, thereby finding a better-performing design scheme based on the fuzzy control results.

[0031] S4.3 Optimization Result Output: When the convergence condition of the objective evaluation function is met or the maximum number of iterations is reached, the final electron gun structure parameters after two-stage co-optimization are output to ensure that the electron gun structure parameters that meet the design requirements are obtained within a reasonable time.

[0032] S5. Perform comprehensive performance evaluation and uniformity analysis on the optimized electron gun: extract the narrow side dimension Wt of the electron beam at the injection waist position. act Width dimension Ww act Current I actCalculate the relative deviation from the expected value; analyze the uniformity of electron density distribution in the injection waist section, and calculate the narrow side density ratio D. n and the ratio of the width to the density D w The laminar flow properties of the electron beam were evaluated. Simultaneously, assembly error redundancy analysis was performed, and the changes in electron beam parameters under cathode position offset were simulated to ensure that the design tolerance met the engineering requirements.

[0033] S6. If the performance evaluation results meet the design requirements, the final electron gun structural parameters are output, including the coefficients of the cathode parabolic equation, the coordinates of the focusing electrode spline nodes, the anode hole size, and the relative positions of each electrode. Based on the output structural parameters, a three-dimensional electron gun model is automatically generated in CST software and verified through simulation to ensure that the macroscopic parameters and performance of the electron beam meet the design specifications. The successful design is stored in the electron gun database, which stores electron gun structural parameters, design target parameters, and historical optimization records. It supports multi-condition combination queries and matching based on voltage, current, and cathode size, and supports reuse in subsequent designs.

[0034] Furthermore, the present invention also provides the following preferred technical solutions:

[0035] In step S2, the adaptive mesh generation strategy is as follows: the mesh size is automatically calculated based on the cathode emitting surface size, the anode hole size, and the overall size of the electron gun; in the vicinity of the cathode emitting surface, the vicinity of the anode hole, and the region with a large electric field gradient, the mesh size can be dynamically adjusted according to the model size; in the region where the electron gun's main trajectory is distributed, the mesh size is set to 1 / 20 of the anode channel size.

[0036] In step S4.1, the input variables of the fuzzy controller include: current deviation ΔI, narrow side size deviation ΔWt, wide side size deviation ΔWw, and target evaluation function index Fit; the output variables include: focusing electrode-anode distance adjustment ΔL, cathode curvature adjustment ΔR, focusing electrode spline node adjustment ΔP, and focusing electrode opening size adjustment ΔF.

[0037] In step S4.1, a dynamic adjustment coefficient is defined. Using formula Calculate the adaptive adjustment factor to achieve dynamic scaling of the error level, enabling the optimization system to automatically adjust the optimization strategy based on the convergence state. Dynamic adjustment coefficient. The initial value is 1, and it decreases adaptively as the objective evaluation function decreases. When the objective evaluation function Fit < 0.5, Reduced to 0.1, when Fit < 0.2, Further reduce it to 0.01 to achieve an adaptive transition from coarse to fine adjustment.

[0038] The method of this invention supports two strip electron gun configurations: single-sided compression and double-sided compression. For the double-sided compression configuration, the cathode adopts a two-dimensional parabolic design, and its surface equation is expressed as:

[0039] ;

[0040] Where a1 and a2 are the parabolic coefficients of the wide side and narrow side, respectively, and are independently adjusted by a fuzzy controller based on conditional judgment to achieve adaptive optimization of the compression ratio of the wide side and narrow side.

[0041] The electron gun database stores the following information: electron gun structural parameters (cathode size, focusing electrode node, anode size), design target parameters (voltage, current, injection waist size), and historical optimization records (feedbackdata file); it supports querying and matching based on multiple conditions such as voltage, current, and cathode size.

[0042] The beneficial effects of this invention are as follows:

[0043] 1. A fuzzy control parameter optimization system based on dynamic precision control is adopted, which implements fuzzy logic through an If-Else conditional judgment structure and introduces dynamic adaptive parameters. It achieves intelligent scaling of error levels, using a larger adjustment step size to quickly approach the target in the early stage of optimization, and automatically switching to fine-tuning in the later stage of optimization, effectively balancing convergence speed and optimization accuracy.

[0044] 2. A 14-dimensional parameter adjustment strategy is proposed. Based on the physical structural characteristics of the strip electron gun, key parameters such as the position of the narrow-side focusing electrode and the opening size are gradually optimized to ensure the uniformity of electron beam distribution and laminar flow at the injection waist position.

[0045] 3. By organically combining multi-objective programming, adaptive mesh generation, and fuzzy control based on dynamic adjustment, and through database management, parametric modeling, and dynamic adjustment mechanisms, the efficient and automated design of high-compression-ratio strip electron guns was achieved, significantly shortening the design cycle and improving design accuracy.

[0046] 4. A reinforcement learning fine-tuning stage is introduced to synergize with the coarse-tuning of fuzzy control. The reinforcement learning algorithm, near the optimal solution provided by fuzzy control, can autonomously learn more refined fine-tuning strategies through the CST-MATLAB co-simulation platform. This overcomes the limitation of fixed-rule optimization potentially getting stuck in local optima, further improving the optimization limit of the electron beam's microscopic performance.

[0047] 5. The proposed method has been verified by simulation and can be used to design various strip electron guns with different compression ratios and current levels. The key indicators such as the injection waist size and electron beam laminar flow properties meet the application requirements of high-power, high-frequency microwave devices. Attached Figure Description

[0048] Figure 1 Overall flowchart of the design method for strip electron gun;

[0049] Figure 2 This is a schematic diagram of the strip-shaped electron gun structure of the present invention;

[0050] Figure 3 Workflow diagram for intelligent collaborative optimization system

[0051] Figure 4 To optimize the iterative convergence curve;

[0052] Figure 5 The image shows the optimized electron beam trajectory; where a represents the electron beam trajectory distribution along the yoz plane (the red dashed line represents the monitor position), b represents the electron beam trajectory distribution along the xoz plane (the red dashed line represents the monitor position), c represents the potential distribution along the yoz plane, and d represents the potential distribution along the xoz plane.

[0053] Figure 6 This is a density distribution map of the waist area of ​​the electron injection. Detailed Implementation

[0054] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0055] See attached document Figure 1 This invention proposes a strip electron gun design method based on intelligent collaborative optimization. The following example illustrates the application of this invention to the design of a strip electron gun with a single-sided compression ratio of 30:

[0056] Design goals: Electron beam voltage 28kV, current 0.2A, beam waist dimensions: narrow side 0.2mm, wide side 1.2mm. The strip-shaped electron gun to be optimized is as follows: Figure 2 As shown.

[0057] Figure 3The specific intelligent collaborative optimization algorithm is divided into two parts. Fuzzy control part: Through the fuzzy control module, the input structural parameters are modeled and simulated using the CST-Matlab co-simulation platform. After obtaining the results, the algorithm is processed to determine whether the current structural parameters have reached the optimal solution. If the requirements are not met, the algorithm is re-optimized and modeled to solve the problem; if the requirements are met, the optimal structural parameters are output. Reinforcement learning part: The structural parameters obtained from the fuzzy control part are modeled and simulated using the CST-Matlab co-simulation platform. After obtaining the results, the algorithm is processed to determine whether the current structural parameters have reached the optimal solution. If the requirements are not met, the algorithm is re-optimized and modeled to solve the problem using the reinforcement learning algorithm; if the requirements are met, the optimal structural parameters are output. 1. Input the design objective, and automatically select the initial model with the closest structure from the electron gun database (e.g., a strip-type electron gun with a compression ratio of 10) using a multi-objective programming algorithm.

[0058] 2. Using the initial model as a baseline, the initial objective function value is calculated through the CST and MATLAB co-simulation platform.

[0059] 3. A parameter optimization system based on intelligent collaborative optimization is adopted to optimize the parameter based on the initial objective function value:

[0060] Phase 1: Current parameter optimization: Prioritize adjusting the anode-cathode spacing, and adopt an adaptive adjustment strategy based on the magnitude of the current deviation;

[0061] Phase 2: Structural parameter optimization: Based on the current meeting the standard, adjust the remaining 13 parameters, such as the narrow-side focusing electrode aperture size and the position of the narrow-side focusing electrode interpolation curve control point.

[0062] The optimization results of the algorithm are as follows Figure 4 As shown, the objective function reaches its minimum value after 110 calculations, at which point the objective function value is 0.1125. The result of the 110th calculation is the optimal solution obtained by the intelligent collaborative optimization system.

[0063] 4. Performance Evaluation: The final optimized electron beam result is as follows: Figure 5 a, b, c, d and Figure 6 As shown, its actual current is 0.212 A, the narrow side dimension of the injection waist is 0.198 mm, the wide side dimension is 1.297 mm, the electron beam target evaluation function Fit=0.1125, and the electron beam velocity ratio V at the injection waist position is... xy / V xyz The mean value is 0.019, which meets the design requirements.

[0064] 5. Output the final structural parameters and store the structural parameters that meet the design requirements into the database for reference in subsequent designs.

[0065] In summary, this invention overcomes the limitations of single algorithms or local structural improvements in its optimization logic. It achieves a unified approach of rapid global convergence and refined local optimization through a two-stage collaborative strategy, resulting in significantly higher optimization efficiency and accuracy than existing single optimization methods. In terms of design, it covers the entire process from initial model selection to design reuse, taking into account macroscopic parameter accuracy, microscopic laminar flow, and engineering assembly tolerance, forming a complete technical closed loop. This differs from existing technologies that focus on optimizing a single performance indicator. Regarding applicability, the dedicated design for the strip-shaped electron gun meets the needs of high-power, high-frequency devices, and the two configurations adapt to different application scenarios, offering greater versatility and practicality.

[0066] The above examples are merely for illustrative purposes. The strip electron gun design method based on intelligent collaborative optimization proposed in this invention can be used for strip electron guns with different frequency bands and compression ratios. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this invention are protected by this invention.

Claims

1. A design method for a strip electron gun based on intelligent collaborative optimization, characterized in that, Includes the following steps: S1. Based on the input electron gun design objectives, including cathode voltage, current, narrow side radius, wide side radius, and narrow side and wide side dimensions of the electron injection waist, the existing model with the closest structure is automatically selected from the electron gun design database as the initial optimization model using a multi-objective programming algorithm. S2. Based on the initial model, the electron gun three-dimensional model is automatically established through the CST and MATLAB joint simulation platform, and the simulation calculation is performed using an adaptive mesh generation strategy to extract the key performance parameters of the electron beam. S3. Based on the extracted electron beam performance parameters, calculate the target evaluation function. This evaluation function consists of a weighted sum of three parts: the target parameter function, the density distribution function, and the laminar flow function. The specific expression is as follows: ； In the formula, Fit opt Fit is a target parameter function used to evaluate the deviation between the macroscopic parameters of the electron beam and the design target. qua Fit is a density distribution function used to evaluate the uniformity of the electron beam density distribution. lam C is a laminarity function used to evaluate the laminarity of the electron beam trajectory; i S4 is a constant, where i=1,2,3, used to adjust the weights of each density parameter; S5 uses an intelligent collaborative optimization algorithm to optimize the electron gun structure parameters using the calculated electron beam performance parameters. The intelligent collaborative optimization algorithm executes the following two stages in sequence: S4.1, Fuzzy Control Coarse-tuning Stage: A fuzzy control parameter optimization system based on dynamic adjustment is used to perform preliminary optimization of the key structural parameters of the electron gun; S4.2 Reinforcement Learning Fine-Tuning Phase: Using the better solution output in S4.1 as the initial state, the reinforcement learning algorithm is started for local fine-tuning. S4.3 When the convergence condition of the objective evaluation function is met or the maximum number of iterations is reached, output the final electron gun structure parameters after two-stage co-optimization. S5. Perform comprehensive performance evaluation and uniformity analysis on the optimized electron gun: extract the narrow side dimension Wt of the electron beam at the injection waist position. act Width dimension Ww act Current I act Calculate the relative deviation from the expected value; analyze the uniformity of electron density distribution in the injection waist section, and calculate the narrow side density ratio D. n The electron beam laminar flow index was evaluated, and the assembly error redundancy analysis was performed. By simulating the changes in electron beam parameters under cathode position offset, the design tolerance was ensured to meet engineering requirements. S6. If the performance evaluation results meet the design requirements, output the final electron gun structural parameters, including the coefficients of the cathode parabolic equation, the coordinates of the focusing electrode spline nodes, the anode hole size, and the relative positions of each electrode. Based on the output structural parameters, automatically generate a three-dimensional electron gun model in CST software and perform verification simulation to ensure that the macroscopic parameters and performance of the electron beam meet the design specifications. Successful designs are stored in an electron gun database, which contains electron gun structural parameters, design target parameters, and historical optimization records. It supports multi-condition queries and matching based on voltage, current, and cathode size, and supports reuse of subsequent designs.

2. The method according to claim 1, characterized in that, In step S2, the adaptive mesh generation strategy is as follows: the mesh size is automatically calculated based on the cathode emitting surface size, the anode hole size, and the overall size of the electron gun; in the vicinity of the cathode emitting surface, the vicinity of the anode hole, and the region with a large electric field gradient, the mesh size can be dynamically adjusted according to the model size; in the region where the electron gun's main trajectory is distributed, the mesh size is set to 1 / 20 of the anode channel size.

3. The method according to claim 1, characterized in that, In step S4, the input variables of the fuzzy controller include: current deviation ΔI, narrow side size deviation ΔWt, wide side size deviation ΔWw, and target evaluation function index Fit; the output variables include: focusing electrode-anode distance adjustment ΔL, cathode curvature adjustment ΔR, focusing electrode spline node adjustment ΔP, and focusing electrode opening size adjustment ΔF.

4. The method according to claim 3, characterized in that, In step S4, the dynamic adjustment coefficient The initial value is 1. During the iteration process, this dynamic adjustment coefficient adaptively decays as the target evaluation function decreases. When the target evaluation function Fit < 0.5, The decay rate is reduced to 0.1 when Fit < 0.

2. The value is further reduced to 0.01, achieving an adaptive transition from coarse to fine adjustment.

5. The method according to claim 3, characterized in that, The method supports two strip electron gun configurations: single-sided compression and double-sided compression. For the double-sided compression configuration, the cathode adopts a two-dimensional parabolic design, and its surface equation is expressed as: ； Where a1 and a2 are the parabolic coefficients of the wide side and narrow side, respectively, and are independently adjusted by a fuzzy controller based on conditional judgment to achieve adaptive optimization of the compression ratio of the wide side and narrow side.

6. The method according to claim 1, characterized in that, In step S4.2, the reinforcement learning algorithm uses the target evaluation function Fit as the reward signal and learns parameter fine-tuning strategies autonomously through interaction with the CST-MATLAB co-simulation environment.

Images