A Design Method and System for a Raceway Cross-Section Coil-Dominated Superconducting Magnet

Abstract

The present invention relates to a design method and system for a superconducting magnet dominated by a racetrack-shaped cross-section coil, which includes: parametrically modeling key parts by using the finite element method and setting respective variation ranges for each key part; the key parts include an iron core structure, a cross-section coil distribution, and an end coil distribution; optimizing the magnetic field in the good field region by a combined optimization method, and during the optimization iteration process, each key part changes simultaneously within its respective variation range to seek a possible optimal solution in a large parameter space, thereby completing the magnet design. The present invention can better achieve the optimized design of a racetrack-shaped cross-section magnet and achieve a smaller volume and cost. The present invention can be applied in the technical field of magnet design.

Description

Technical Field

[0001] The present invention relates to the technical field of magnet design, and particularly to a design method and system for a racetrack-shaped cross-section coil-dominated superconducting magnet. Background Art

[0002] The deflection dipole magnet is an important component in a low-temperature high-density nuclear matter measurement spectrometer. For this type of spectrometer, a superconducting magnet with a large acceptance angle and a large internal space in the magnet is generally selected. A large acceptance angle can make full use of various detection devices to expand and enhance the research scope and capabilities of the spectrometer. A large effective field space can accommodate more detectors, improving the detection and identification capabilities of various products.

[0003] In order to achieve a large opening angle and a large space while ensuring a large range of magnetic field uniformity, a racetrack-shaped cross-section coil-dominated superconducting magnet can be used. Through the racetrack-shaped cross-section, a very large opening can be achieved along the beam direction of the magnet. At the same time, the required large-range background magnetic field is generated through the coil distribution.

[0004] The design of the coil-dominated superconducting magnet mainly generates a specific magnetic field by analytically solving the spatial current distribution. For the CCT&DCT type of circular cross-section coil-dominated superconducting magnet, its current distribution can be analytically given. Therefore, when optimizing the coil, the magnetic field can be optimized by changing the parameters in the analytical formula to redefine the coil function. However, for the magnet design with a racetrack-shaped cross-section, which is a non-circular cross-section, the method of designing the magnetic field through the induced current can only numerically solve the coil current distribution under the racetrack-shaped cross-section. Subsequent steps such as discretization and adding a shielding iron core will bring about partial changes in the magnetic field, and the coil distribution needs to be adjusted to achieve the design goal. Summary of the Invention

[0005] Aiming at the above problems, the purpose of the present invention is to provide a design method and system for a racetrack-shaped cross-section coil-dominated superconducting magnet, which can better realize the optimized design of the racetrack-shaped cross-section magnet and achieve a smaller volume and cost.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: A design method for a racetrack-shaped cross-section coil-dominated superconducting magnet, which includes: using the finite element method to perform parametric modeling on key parts and setting the respective change ranges of each key part; the key parts include the iron core structure, the cross-section coil distribution, and the end coil distribution; optimizing the magnetic field in the good field area through a combined optimization method, and during the optimization iteration process, each key part changes simultaneously within its respective change range to seek possible optimal solutions in a large parameter space and complete the magnet design.

[0007] Further, the parametric modeling of the iron core structure is to determine the parameters of the iron core chamfer, the internal grooving of the iron core, and the end padding of the iron core, and finely adjust the magnetic fields at various positions in the good field area by using the detailed structures of the iron core chamfer, the internal grooving of the iron core, and the end padding of the iron core.

[0008] Further, the parametric modeling of the cross-sectional coil distribution refers to normalizing the current distribution of the coils in the racetrack-shaped cross-section, and then performing polynomial fitting on the current distribution to convert the numerical solution of the coil distribution into an analytical expression with parameters.

[0009] Further, the parametric modeling of the end coil distribution refers to introducing a functional offset at the position of the end loop of the coil.

[0010] Further, the magnetic field in the good field area is optimized by the combined optimization method to seek possible optimal solutions in the large parameter space, including:

[0011] Determine the outer contour distribution of the coil according to the preliminary calculation, and determine the coefficients of each term in the current distribution function of the magnet coil cross-section and the end through the discretized current distribution;

[0012] Perform the first-level optimization on each coefficient, and find the preliminary optimal solution within the preset parameter range in the large parameter space as the first-level parameter range;

[0013] Perform the second-level optimization on each coefficient after the first-level optimization, and optimize the first-level parameter range into the second-level parameter range;

[0014] Perform the third-level optimization on the second-level parameter range, and further optimize the second-level parameter range into the third-level parameter range.

[0015] Further, the first-level optimization and the second-level optimization both adopt the genetic algorithm for optimization, and the third-level optimization adopts the particle swarm algorithm for optimization.

[0016] Further, the preset parameter range is preset according to the feasibility of the magnet structure.

[0017] Further, the second-level parameter range is 50% - 150% of the first-level optimization result.

[0018] Further, the third-level parameter range is 90% - 110% of the second-level optimization result.

[0019] A racetrack cross-section coil-dominated superconducting magnet design system, which includes: a first processing module that uses the finite element method to perform parametric modeling on key parts and sets the respective change ranges of each key part; the key parts include an iron core structure, a cross-section coil distribution, and an end coil distribution; a second processing module that optimizes the magnetic field in the good field area through a combined optimization method, and during the optimization iteration process, each key part changes simultaneously within its respective change range to seek possible optimal solutions in a large parameter space and complete the magnet design.

[0020] Due to the above technical solutions adopted by the present invention, it has the following advantages:

[0021] The present invention can better realize the optimized design of such special function magnets and achieve the design goal with a smaller volume and cost. Description of the Drawings

[0022] Figure 1 is a flowchart of a racetrack cross-section coil-dominated superconducting magnet design method in an embodiment of the present invention;

[0023] Figure 2 is a schematic diagram of iron core optimization in an embodiment of the present invention;

[0024] Figure 3 is a schematic diagram of coil optimization in an embodiment of the present invention;

[0025] Reference Signs:

[0026] 1. Outer dimension of the iron core, 2. Pad at the end of the iron core, 3. Hollow inside the iron core, 4. Chamfer of the iron core, 5. Cross-section distribution of the coil, 6. Distribution of the end loop. Detailed Embodiments

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of the present invention fall within the scope of protection of the present invention.

[0028] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they specify the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0029] If the runway cross-section is discretized according to the ideal dipole field, the two-dimensional magnetic field results are relatively ideal, and good magnetic field uniformity can be obtained through simple optimization. However, in the three-dimensional magnetic field optimization, not only the actual size of the magnet needs to be considered, but also issues such as the end loops need to be considered during coil winding. At the same time, at the beam inlet and outlet, there will be a large horizontal current distribution at the coil end positions, and these currents will also affect the magnetic field uniformity in the good field region. Therefore, the present invention provides a design method and system for a superconducting magnet dominated by a runway cross-section coil. The present invention uses a multi-parameter collaborative optimization model. During the optimization process, the iron core structure, cross-section coil distribution, and end coil distribution are respectively modeled, and then the magnetic field in the good field region is optimized through a combined optimization method.

[0030] In an embodiment of the present invention, a design method for a superconducting magnet dominated by a runway cross-section coil is provided. In this embodiment, as Figure 1 shown, the method includes the following steps:

[0031] 1) Use the finite element method to perform parametric modeling on key parts such as the iron core structure, cross-section coil distribution, and end coil distribution, and set the respective change ranges of each key part;

[0032] 2) Optimize the magnetic field in the good field region through a combined optimization method. During the optimization iteration process, each key part changes simultaneously within its respective change range to seek possible optimal solutions in the large parameter space and complete the magnet design.

[0033] In the above step 1), the parametric modeling of the iron core structure is to determine parameters such as the iron core chamfer, internal slotting of the iron core, and end padding of the iron core, and use detailed structures such as the iron core chamfer, internal slotting of the iron core, and end padding of the iron core to finely adjust the magnetic field at various positions in the good field region.

[0034] In the above step 1), the parametric modeling of the cross-section coil distribution refers to normalizing the current distribution of the coils in the runway cross-section, and then performing polynomial fitting on the current distribution to convert the numerical solution of the coil distribution into an analytical expression with parameters, so as to control the cross-section coil distribution during the optimization process. Especially for magnets with a large number of turns, after fitting, the parameter space dimension can be greatly reduced, improving the optimization efficiency.

[0035] In the above step 1), the parametric modeling of the end coil distribution refers to introducing a functional offset at the end loop position of the coil, which can be introduced from the vertical direction or along the horizontal direction of the magnet. By introducing the offset, the influence of the end loop on the magnetic field in the good field region can be offset / enhanced, thereby realizing the optimization of the magnetic field in the good field region. At the same time, for magnets with a large number of turns, by introducing the functional offset, the parameter space dimension can also be greatly reduced, improving the optimization efficiency.

[0036] In step 2) above, the combined optimization method refers to using multiple optimization methods including genetic algorithms and particle swarm algorithms to optimize parameters before, during, and after optimization. By means of a preset combination method, the optimization efficiency is improved.

[0037] In step 2) above, the magnetic field in the good field area is optimized by the combined optimization method to find possible optimal solutions in a large parameter space, including the following steps:

[0038] 2.1) Determine the outer contour distribution of the coil according to preliminary calculations, determine the coefficients in the current distribution function of the cross-section and the end of the magnet coil by discretizing the current distribution, and preset the parameter ranges of each coefficient.

[0039] 2.2) Perform the first-level optimization on each coefficient to find a preliminary optimal solution within the preset parameter range in the large parameter space, and this preliminary optimal solution serves as the first-level parameter range;

[0040] In this embodiment, the first-level optimization is carried out using a genetic algorithm. The preset parameter range is preset according to the feasibility of the magnet structure.

[0041] 2.3) Perform the second-level optimization on each coefficient after the first-level optimization to optimize the first-level parameter range into the second-level parameter range;

[0042] In this embodiment, the second-level optimization is also carried out using a genetic algorithm. The second-level parameter range is 50% - 150% of the first-level optimization result.

[0043] 2.4) Perform the third-level optimization on the second-level parameter range to further optimize the second-level parameter range into the third-level parameter range;

[0044] In this embodiment, the third-level optimization is carried out using a particle swarm algorithm, and the third-level parameter range is 90% - 110% of the second-level optimization result.

[0045] In the above steps, in the magnetic field design of the pure coil with iron core structure, according to the characteristics of different optimization algorithms, different optimization algorithms are selected in different optimization stages, and different parameter setting ranges are selected for different types of optimization parameters, which can significantly improve the optimization efficiency and enhance the optimization effect.

[0046] Example:

[0047] As Figure 2 shown, the iron core details include the edge chamfer Br. The internal intermediate grooving structure of the iron core: groove length L_hole, groove width W_hole, and groove depth H_hole. The iron core end padding structure: padding length L_shim, padding width W_shim, and padding height H_shim.

[0048] After entering the iron core, the current distribution that generates the standard dipole field changes, and it is necessary to model and optimize the cross-sectional coil distribution. The cross-sectional coil part and the normalized current distribution curve are as shown in Figure 3 shown. According to the original current part, polynomial fitting is performed, and the coil distribution function L(n):

[0049] L(n) = a0 + a1*n 1 + a2*n 2 + a3*n 3 + a4*n 4 + a5*n 5 + a6*n 6

[0050] The coil distribution on the cross-section is determined by 7 parameters a0 to a6; n represents the coil number.

[0051] The end coil distribution can also be modeled. The cross-sectional current distribution is dominated by arc segments. For the convenience of modeling and optimization, the following distribution is added to the Z coordinate of the end coil for adjustment:

[0052] Z(c) = Z0 + Z1cos(θ) + Z2cos(2θ)

[0053] The coil end is modeled by three parameters Z0, Z1, and Z2, and θ represents the corresponding angle of the coil. The optimization parameters and the initial range settings of each parameter are shown in Table 1. The coil scheme optimization model is jointly determined by the 16 parameters in Table 1.

[0054] Table 1 Optimization parameters and parameter setting ranges

[0055]

[0056] In order to improve the optimization efficiency of the 16-parameter model, a combination of multiple optimization methods is adopted for step-by-step optimization in the magnetic field optimization process of the coil scheme. The range settings of each round of optimization are based on the nature of the parameters. The process is as follows: Determine the outer contour distribution of the coil according to the preliminary calculation, and determine the coefficients of each term in the current distribution function of the magnet coil cross-section and end by discretizing the current distribution. Subsequently, the first round of genetic algorithm optimization is carried out to find the preliminary optimal solution in the large parameter space. In the second round of genetic algorithm optimization, the parameter range is reduced to (50% - 150%) of the previous optimal solution. Finally, the particle swarm algorithm is used for optimization in a smaller range (parameter range 90% - 110%). In the magnetic field design of this pure coil with iron core structure, according to the characteristics of different optimization algorithms, different optimization algorithms are selected in different optimization stages, and different parameter setting ranges are selected for different nature optimization parameters, which can significantly improve the optimization efficiency and enhance the optimization effect.

[0057] After calculating and optimizing the three-dimensional iron core structure and coil distribution, based on various physical requirements in the three-dimensional results, such as the basic requirements of a large range, large air gap, and high uniformity of the magnetic field, further considerations need to be given to internal space, application scenarios, coil winding, iron core processing, Dewar installation, the overall size of the magnet, etc. The determination of the final solution needs to comprehensively consider multiple systems and complete the design through multiple rounds of iteration.

[0058] In an embodiment of the present invention, a racetrack cross-section coil-dominated superconducting magnet design system is provided, which includes:

[0059] A first processing module, which uses the finite element method to perform parametric modeling on key parts and sets the respective change ranges of each key part; the key parts include the iron core structure, cross-section coil distribution, and end coil distribution;

[0060] A second processing module, which optimizes the magnetic field in the good field area through a combined optimization method. During the optimization iteration process, each key part changes simultaneously within its respective change range to seek possible optimal solutions in a large parameter space and complete the magnet design.

[0061] In the above embodiment, in the first processing module, the parametric modeling of the iron core structure is to determine the parameters of the iron core chamfer, the iron core internal groove, and the iron core end padding, and use the detailed structures of the iron core chamfer, the iron core internal groove, and the iron core end padding to finely adjust the magnetic field at various locations in the good field area.

[0062] In the above embodiment, in the first processing module, the parametric modeling of the cross-section coil distribution refers to normalizing the current distribution of the coils in the racetrack cross-section, and then performing polynomial fitting on the current distribution to convert the numerical solution of the coil distribution into an analytical expression with parameters.

[0063] In the above embodiment, in the first processing module, the parametric modeling of the end coil distribution refers to introducing a functional offset at the position of the end coil return line.

[0064] In the above embodiment, in the second processing module, the magnetic field in the good field area is optimized through a combined optimization method to seek possible optimal solutions in a large parameter space, including:

[0065] A parameter determination module, which determines the outer contour distribution of the coil according to preliminary calculations, determines the coefficients of each term in the current distribution function of the magnet coil cross-section and the end through discretizing the current distribution, and pre-sets the parameter range;

[0066] A first-level optimization module, which performs a first-level optimization on each coefficient and finds the preliminary optimal solution within the preset range in the large parameter space as the first-level parameter range;

[0067] The secondary optimization module performs secondary optimization on each coefficient after the primary optimization, and optimizes the primary parameter range into a secondary parameter range;

[0068] The tertiary optimization module performs tertiary optimization on the secondary parameter range, and further optimizes the secondary parameter range into a tertiary parameter range.

[0069] In the above embodiments, both the primary optimization and the secondary optimization are performed using the genetic algorithm.

[0070] In the above embodiments, the secondary parameter range is 50% to 150% of the primary optimization result.

[0071] In the above embodiments, the tertiary optimization is performed using the particle swarm optimization algorithm.

[0072] In the above embodiments, the tertiary parameter range is 90% to 110% of the secondary optimization result.

[0073] The system provided in this embodiment is used to execute the above method embodiments. For the specific process and detailed content, please refer to the above embodiments and will not be elaborated here.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A design method of a superconducting magnet dominated by a racetrack-shaped cross-section coil, characterized in that, Including: Parametric modeling of key parts is carried out using the finite element method, and the respective change ranges of each key part are set; the key parts include the iron core structure, the cross-sectional coil distribution, and the end coil distribution; The magnetic field in the good field area is optimized by a combined optimization method. During the optimization iteration process, each key part changes simultaneously within its respective change range to seek possible optimal solutions in a large parameter space and complete the magnet design; The optimization of the magnetic field in the good field area by the combined optimization method to seek possible optimal solutions in a large parameter space includes: Determine the coil outer contour distribution according to the preliminary calculation, and determine the coefficients in the current distribution function of the magnet coil cross-section and end by discretizing the current distribution; Perform the first-level optimization on each coefficient, and use the preliminary optimal solution within the preset parameter range in the large parameter space as the first-level parameter range; Perform the second-level optimization on the coefficients after the first-level optimization, and optimize the first-level parameter range to the second-level parameter range; Perform the third-level optimization on the second-level parameter range, and further optimize the second-level parameter range to the third-level parameter range; The parametric modeling of the iron core structure is to determine the iron core chamfer, the iron core internal groove, and the iron core end pad parameters, and use the detailed structures of the iron core chamfer, the iron core internal groove, and the iron core end pad to finely adjust the magnetic fields at various positions in the good field area; The parametric modeling of the cross-sectional coil distribution refers to normalizing the current distribution of the coils in the racetrack-shaped cross-section, and then performing polynomial fitting on the current distribution to convert the numerical solution of the coil distribution into an analytical expression with parameters; The parametric modeling of the end coil distribution refers to introducing a functional offset at the end coil loop position; Cross-sectional coil distribution function : Determined by the coil distribution on the cross-section is determined by 7 parameters, representing the current distribution of the coil cross-section; representing the coil number; The cross-sectional current distribution is dominated by arc segments, and the following adjustments are made on the end coil Z coordinate: Through parameters Model the coil ends with three parameters Indicating the end current distribution Indicating the angle corresponding to the respective coil; The optimization parameters and the initial range settings of each parameter in the coil scheme optimization model are as follows: The iron core edge chamfer Br, and its parameter setting range is 0 - 300; The length of the iron core center groove, and its parameter setting range is 0 - 3000; The width of the iron core center groove, and its parameter setting range is 0 - 2000; The depth of the iron core center groove, and its parameter setting range is 0 - 50; The length of the iron core edge pad, and its parameter setting range is 0 - 300; The width of the iron core edge pad, and its parameter setting range is 0 - 1600; The height of the iron core edge pad, and its parameter setting range is 0 - 150; The coil cross-sectional current distribution, and its parameter setting range is ±10%; The end current distribution, and its parameter setting range is ±10%.

2. The design method of the racetrack-shaped cross-section coil-dominated superconducting magnet according to claim 1, wherein, Both the first-level optimization and the second-level optimization adopt the genetic algorithm for optimization, and the third-level optimization adopts the particle swarm algorithm for optimization.

3. The runway-shaped cross-section coil-dominated superconducting magnet design method according to claim 1, characterized in that The preset parameter range is preset according to the feasibility of the magnet structure.

4. The design method of the racetrack cross-section coil-dominated superconducting magnet according to claim 1, characterized in that, The second-level parameter range is 50% - 150% of the first-level optimization result.

5. The runway-type cross-section coil-dominated superconducting magnet design method according to claim 1, characterized in that, The third-level parameter range is 90% - 110% of the second-level optimization result.

6. A racetrack cross-section coil-dominated superconducting magnet design system for implementing the racetrack cross-section coil-dominated superconducting magnet design method according to any one of claims 1-5, characterized in that, Including: The first processing module, which carries out parametric modeling of key parts using the finite element method and sets the respective change ranges of each key part; the key parts include the iron core structure, the cross-sectional coil distribution, and the end coil distribution; The second processing module optimizes the magnetic field in the good field area through a combined optimization method. During the optimization iteration process, each key part changes simultaneously within its respective change range to seek possible optimal solutions in a large parameter space and complete the magnet design; The optimization of the magnetic field in the good field area through the combined optimization method to seek possible optimal solutions in a large parameter space includes: Determine the outer contour distribution of the coil according to the preliminary calculation, and determine the coefficients in the current distribution function of the magnet coil cross-section and the end through the discretized current distribution; Perform the first-level optimization on each coefficient, and use the preliminary optimal solution within the preset parameter range in the large parameter space as the first-level parameter range; Perform the second-level optimization on each coefficient after the first-level optimization, and optimize the first-level parameter range into the second-level parameter range; Perform the third-level optimization on the second-level parameter range, and further optimize the second-level parameter range into the third-level parameter range; The parametric modeling of the iron core structure is to determine the iron core chamfer, the internal groove of the iron core, and the pad parameters at the end of the iron core, and use the detailed structures of the iron core chamfer, the internal groove of the iron core, and the pad at the end of the iron core to finely adjust the magnetic field at each place in the good field area; The parametric modeling of the cross-section coil distribution refers to normalizing the distribution of the coil current in the racetrack-shaped cross-section, and then performing polynomial fitting on the current distribution to convert the numerical solution of the coil distribution into an analytical expression with parameters; The parametric modeling of the end coil distribution refers to introducing a functional offset at the end coil return position.

Images