Method for optimizing position of magnetic shielding ring for suppressing stray loss of transformer winding end

By optimizing the position of the magnetic shielding ring, and based on leakage magnetic field distribution data and magnetic mirror effect, the problems of stray loss at the winding end and concentrated leakage magnetic field leakage were solved, thereby improving the operational reliability and energy efficiency of the transformer.

CN123116743APending Publication Date: 2026-11-17NANJING ZHENGRUI POWER TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611626148.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-20
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies cannot adapt to differences in winding structure when designing the position of the magnetic shielding ring, leading to concentrated leakage flux and local overheating problems. Furthermore, traditional methods are difficult to effectively suppress stray losses at the ends of transformer windings.

Method used

By acquiring leakage magnetic field distribution data at the ends of transformer windings, the interception efficiency of the shielding ring is evaluated based on vector direction characteristics. The additional eddy current loss is quantified by combining magnetic mirror effect. The optimal installation position is selected by using net benefit conditions and saturation margin constraints, and the position of the magnetic shielding ring is optimized.

Benefits of technology

It improves the operational reliability and energy efficiency of the transformer across the entire load range, reduces stray losses, and avoids localized overheating and magnetic flux leakage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123116743A_ABST
    Figure CN123116743A_ABST
Patent Text Reader

Abstract

The application discloses a magnetic shielding ring position optimization method for inhibiting stray loss of a transformer winding end, and comprises the following steps: obtaining leakage magnetic field distribution data of a transformer winding end region; evaluating the trapping efficiency of a shielding ring based on the vector direction characteristics in the leakage magnetic field distribution data, and screening out a shielding ring candidate position domain according to the evaluation; quantifying the additional eddy current loss penalty of each position in the candidate position domain on the winding based on the magnetic mirror effect, and performing lower bound constraint screening according to the net income condition; evaluating the local saturation breakdown risk of the inner edge of the shielding ring, and performing upper bound constraint screening according to the saturation margin constraint; and in the comprehensive feasible position domain, the effective attraction range of the shielding ring under multiple load conditions is combined to determine the optimal installation position by optimization with the minimum total stray loss as the target. The application establishes a three-constraint countermechanism of the trapping efficiency, the winding penalty and the saturation risk, is beneficial to realizing the global analytical optimization of the shielding ring position, and improves the operation reliability of the transformer in the full load interval.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of power transformer design and optimization technology, and in particular, it is a method for optimizing the position of magnetic shielding rings to suppress stray losses at the ends of transformer windings. Background Technology

[0002] During the operation of large power transformers, leakage magnetic fields exist in the winding end region. Once leakage flux penetrates the air or insulating medium and enters the surface of magnetically conductive structural components such as the tank wall and clamps, eddy currents are induced in localized areas, leading to stray losses and localized overheating. To suppress this phenomenon, a high-permeability magnetic shielding ring can be installed above the winding end. By providing a low magnetic reluctance path, the leakage flux is intercepted and redirected, further improving the transformer's operational reliability and energy efficiency. The installation position of the shielding ring directly determines its ability to intercept leakage flux and its effectiveness in controlling overall losses.

[0003] Current shielding ring placement design typically relies on a combination of electromagnetic field numerical simulation and empirical rules. Engineers can use finite element method (FEM) software to calculate the leakage magnetic field distribution in the unshielded state, extract the extreme points of leakage flux density amplitude, and tend to place the shielding ring in the region of highest magnetic field strength. When determining the installation gap, a fixed empirical safety distance is usually used, such as no less than 50mm from the winding end face, to avoid insulation or heating side effects caused by the shielding ring. Simultaneously, when assessing the operational safety of the shielding ring itself, the design criterion is primarily that the average magnetic flux density of the ring cross-section does not exceed the material's saturation limit.

[0004] The technical system still needs improvement to address the issue that empirical spacing rules cannot adapt to differences in winding structure. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a method for optimizing the position of the magnetic shielding ring to suppress stray losses at the ends of transformer windings.

[0006] The following technical solutions can be adopted:

[0007] A method for optimizing the position of a magnetic shielding ring to suppress stray losses at the ends of transformer windings includes:

[0008] Obtain leakage magnetic field distribution data in the transformer winding end region;

[0009] The interception efficiency of the shielding ring is evaluated based on the vector direction characteristics in the leakage magnetic field distribution data, and the candidate location domain of the shielding ring is selected accordingly (interception efficiency).

[0010] Based on the magnetic mirror effect, the additional eddy current loss penalty generated by each position in the candidate position domain of the shielding ring on the winding is quantified. The candidate position domain of the shielding ring is screened by lower bound constraint according to the net benefit condition to obtain the candidate position domain after mirror constraint.

[0011] The risk of local saturation breakdown at the inner edge of the shielding ring within the candidate location domain after mirror constraint is evaluated. Based on the saturation margin constraint, the candidate location domain after mirror constraint is screened by upper bound constraint to obtain the comprehensive feasible location domain.

[0012] Within the comprehensive feasible location domain, and considering the effective attraction range of the shielding ring under multiple load conditions, optimization calculations are performed with the goal of minimizing total stray loss to determine the optimal installation location of the magnetic shielding ring.

[0013] The beneficial effects of this invention are:

[0014] This approach promotes global analytical optimization of the shielding ring position, improving the operational reliability of the transformer across the entire load range. The related technical effects will be described in detail below with reference to specific embodiments. Attached Figure Description

[0015] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0016] Figure 1 A flowchart is shown for a method to optimize the position of a magnetic shielding ring to suppress stray losses at the ends of transformer windings;

[0017] Figure 2 A flowchart is shown to evaluate the interception efficiency of a shielding ring based on vector direction characteristics in leakage magnetic field distribution data;

[0018] Figure 3 A flowchart is shown to illustrate the additional eddy current loss penalty of the winding at each position within the candidate position domain of the shielding ring based on the magnetic mirror effect.

[0019] Figure 4 A flowchart is shown to filter candidate location domains of a shielding ring based on a lower bound constraint according to a net benefit condition, and to obtain candidate location domains after mirror constraint.

[0020] Figure 5 A flowchart is shown to evaluate the risk of local saturation breakdown at the inner edge of the shielding ring within the candidate location domain after mirror constraint.

[0021] Figure 6 A flowchart is shown to calculate the edge aggregation factor at the inner corner of the shielding ring based on the geometric dimensional parameters and equivalent chamfer features of the shielding ring cross-section.

[0022] Figure 7 A flowchart is shown to perform upper bound constraint screening on candidate location domains after mirror constraint based on saturation margin constraint to obtain a comprehensive feasible location domain. Detailed Implementation

[0023] To enable those skilled in the art to better understand the present invention, the technical solutions 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 only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0024] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a predetermined order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in sequences other than those illustrated or described herein. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0025] In response to the problems existing in the prior art, the inventors, through searching, research and analysis, found that the prior art has the following limitations:

[0026] Because the leakage magnetic field at the winding end has a highly complex spatial vector distribution, and the path of leakage magnetic flux propagating to structural components is affected by the load fluctuations of the transformer under multiple operating conditions, the existing design methods based on scalar amplitude extrema and static empirical rules are still insufficient when dealing with complex electromagnetic environments.

[0027] On the one hand, the proximity placement strategy makes it difficult to achieve better magnetic flux interception efficiency of the shielding ring in the cross-sectional normal direction, and the fixed spacing rules cannot adapt to the differences in winding structure of transformers with different capacities, which can easily lead to unexpected local overheating.

[0028] On the other hand, safety assessments based on average criteria are difficult to capture the extreme non-uniformity of the microscopic magnetic field distribution inside the shielding ring. Under extreme operating conditions, there is still a problem of concentrated leakage magnetic flux caused by local magnetic reluctance abrupt changes.

[0029] To overcome the above shortcomings, combined with Figures 1 to 7 The present application will be further illustrated with reference to specific embodiments.

[0030] Accordingly, some nouns or terms appearing in the description of the embodiments of this application shall be interpreted as follows:

[0031] In the following text, the magnetic shielding ring may also be referred to as the shielding ring.

[0032] In this application, a method for optimizing the position of a magnetic shielding ring to suppress stray losses at the ends of a transformer winding may specifically include:

[0033] Step 101: Obtain leakage magnetic field distribution data in the transformer winding end region;

[0034] In practice, a two-dimensional axisymmetric finite element calculation model of the transformer winding end region can be established to obtain the required data.

[0035] This modeling method is applicable to the middle phase of a three-phase transformer and the side areas far from the corners of the tank.

[0036] Accordingly, the actual ampere-turn distribution of the high-voltage winding and the low-voltage winding is applied as the excitation source, the core is set to adopt a nonlinear magnetization curve, the winding region is equivalent to an anisotropic continuous medium, and the boundary condition is taken as the outer edge of the computational domain as the parallel magnetic field line boundary.

[0037] After solving the steady-state magnetic field under the rated load current condition, the magnetic field components at all grid nodes in the winding end region are extracted as leakage magnetic field distribution data.

[0038] Step 102: Evaluate the interception efficiency of the shielding ring based on the vector direction characteristics in the leakage magnetic field distribution data, and select candidate location domains for the shielding ring accordingly (interception efficiency).

[0039] In this step, vector direction features are introduced for spatial dimensionality reduction and initial screening. That is, the radial and axial components in the leakage magnetic field are used to calculate the degree of magnetic flux deflection, thereby quantifying the potential of a unit cross-section shielding ring to intercept leakage magnetic flux and reduce eddy current losses in structural components at different spatial locations;

[0040] In this case, a set of spatial coordinates with potential values ​​higher than a preset threshold can be extracted to form an initial search space.

[0041] Step 103: Based on the magnetic mirror effect, quantify the additional eddy current loss penalty generated by each position in the candidate position domain of the shielding ring on the winding, and perform lower bound constraint screening on the candidate position domain of the shielding ring according to the preset net benefit condition to obtain the candidate position domain after mirror constraint.

[0042] In addition, while the shielding ring intercepts the leakage flux, its high permeability surface will generate a magnetic mirror effect, which will increase the eddy current loss of the winding itself.

[0043] By using the plane magnetic mirror theory to make the distributed ampere-turns equivalent, the superposition enhancement coefficient of the mirror field on each conductor layer can be calculated, thereby quantifying the safety penalty on the winding side.

[0044] By comparing the reduction in structural losses caused by the shielding ring with the increase in additional winding losses, invalid installation distances that do not reduce overall losses but increase are eliminated, thus completing the lower bound constraint on the candidate positions.

[0045] Step 104: Evaluate the risk of local saturation breakdown at the inner edge of the shielding ring within the candidate location domain after mirror constraint, and perform upper bound constraint screening on the candidate location domain after mirror constraint according to the preset saturation margin constraint to obtain the comprehensive feasible location domain.

[0046] In this step, an upper bound constraint is further implemented to address the operational reliability of the shielding ring itself. When magnetic flux enters the finite cross-section of the shielding ring, an edge-gathering effect will occur at the inner corner points.

[0047] Optionally, the local peak magnetic flux density at this point can be quantified by analytical mapping to assess the risk of corner points entering saturation first and triggering cascading failures caused by concentrated magnetic flux leakage.

[0048] Based on this, positions with sufficient safety margins are reserved to form a comprehensive feasible position domain that balances high interception efficiency, low winding penalty, and no saturation risk.

[0049] Step 105: Within the comprehensive feasible location domain, determine the effective attraction range of the shielding ring based on the shielding ring's geometric parameters and the material's nonlinear characteristics under various load conditions. Combining the effective attraction range of the shielding ring under multiple load conditions, perform optimization calculations with the goal of minimizing total stray loss to determine the optimal installation position of the magnetic shielding ring.

[0050] Considering that transformers operate under varying loads, the spatial distribution of the leakage magnetic field will contract or expand with the load current. By introducing a nonlinear saturation correction to quantify the actual magnetic flux attraction capability of the shielding ring under different loads, potential off-target failures under light loads can be calculated.

[0051] Based on this, the residual losses of structural components, additional losses of windings, and losses of the shielding ring itself under various working conditions are weighted and summed. The objective function is calculated by traversing the comprehensive feasible location domain, and the installation coordinates with better robustness under multiple working conditions are output.

[0052] Based on the above embodiments, the specific steps for evaluating the interception efficiency of the shielding ring based on the vector direction characteristics in the leakage magnetic field distribution data, and for selecting candidate location domains of the shielding ring accordingly, are further described in detail. In one possible implementation, the steps include:

[0053] The interception efficiency of the shielding ring is evaluated based on the vector direction characteristics in the leakage magnetic field distribution data, including:

[0054] Step 201: Extract the radial component and axial component of the leakage magnetic field from the leakage magnetic field distribution data;

[0055] In practice, the node data file output by the finite element simulation can be parsed, and the vector components of the leakage magnetic field can be separated for each spatial grid node in the winding end region.

[0056] In some scenarios, the axial range of the area under investigation covers 0 to 2 times the radial thickness of the winding above the winding end face, and the radial range covers the outer edge of the core yoke to the inner surface of the tank wall.

[0057] The specific ranges mentioned above can be set and adjusted by technical personnel based on general standards or current specifications. The values ​​given in this application are only examples and not the only limitations; they will not be elaborated further below.

[0058] Step 202: Determine the magnetic flux turning angle at each location based on the radial component and axial component of the leakage magnetic field, and construct the magnetic flux turning angle spatial spectrum.

[0059] Accordingly, the ratio of the radial component to the axial component at each node is calculated, and the arctangent of this ratio is obtained. This operation transforms discrete vector direction information into a continuous scalar angular field, reflecting the geometric characteristics of the transition of magnetic flux from axial to radial flow.

[0060] The specific calculation formula is as follows:

[0061] θ(r,z)=arctan(B _r (r,z) / B _z (r,z));

[0062] In the formula, θ(r,z) is the magnetic flux steering angle at spatial coordinates (r,z), and B _r (r,z) represents the radial component of the leakage magnetic field at this coordinate, B _z (r,z) represents the axial component of the leakage magnetic field at this coordinate, and arctan is the arctangent function.

[0063] Step 203: Calculate the magnetic flux density amplitude distribution by combining the radial component and the axial component of the leakage magnetic field;

[0064] By squaring the radial and axial components separately, adding them together, and then taking the square root of the sum, the comprehensive magnetic field strength of each grid node is obtained, providing basic amplitude data for subsequent loss potential assessment.

[0065] Step 204: Calculate the interception efficiency prediction factor based on the magnetic flux steering angle spatial spectrum and magnetic flux density amplitude distribution.

[0066] This step can be used to fuse magnetic flux direction information with amplitude information to construct a single scalar field to characterize the shielding effectiveness at a local location. The specific multi-step calculation logic is implemented through the following sub-steps.

[0067] Optionally, the product of the magnitude in the magnetic flux density magnitude distribution and the sine value of the corresponding magnetic flux turning angle spatial spectrum is obtained;

[0068] When evaluating the ability of a shielding ring to intercept leakage flux, an engineering approximation with the radial component as the dominant interception component is introduced. In engineering practice, the axial height of the shielding ring cross-section is usually not less than its radial width; under this condition, the above approximation has high accuracy.

[0069] When the cross-sectional shape does not meet this condition, technicians can introduce a cross-sectional aspect ratio correction factor into the sinusoidal weight term for adaptive adjustment.

[0070] Under these conditions, the radial projection component of the magnetic flux entering the inner cylindrical surface of the shielding ring dominates. The sine value of the steering angle can be extracted and multiplied by the amplitude to extract the effective magnetic flux density component that can actually penetrate the cross-section of the shielding ring.

[0071] Optionally, the square of the product is calculated, and the square is normalized using the preset maximum leakage magnetic flux density at the winding end to obtain the interception efficiency prediction factor.

[0072] Alternatively, the square of the product is calculated, and the square is normalized using the maximum value in the magnetic flux density amplitude distribution to obtain the interception efficiency prediction factor.

[0073] Since the eddy current loss of a structural component that can be reduced by a shielding ring at a certain spatial location is proportional to the square of the local magnetic field strength generated by the intercepted magnetic flux, the above effective magnetic flux density components are squared to reflect the potential for reducing eddy current loss.

[0074] Next, the square of the maximum amplitude of leakage magnetic flux density in the entire field is used as a reference benchmark, and the calculation results are divided to obtain the dimensionless prediction factor.

[0075] The specific calculation formula is as follows:

[0076] η(r,z)=(B(r,z)×sin(θ(r,z))) 2 / B max 2 ;

[0077] In the formula, η(r,z) is the interception efficiency prediction factor at spatial coordinates (r,z), B(r,z) is the magnetic flux density amplitude at that coordinate, θ(r,z) is the magnetic flux steering angle at that coordinate, sin is the sine function, and B maxThis represents the maximum leakage magnetic density in the winding end region.

[0078] Optionally, the spatial region where the interception efficiency prediction factor is greater than the preset efficiency threshold is extracted to form the candidate location domain of the shielding ring.

[0079] Traverse all field nodes and add the coordinates of the predicted factor values ​​that exceed the threshold standard to the candidate set;

[0080] In some scenarios, if the interception efficiency at the predetermined location is lower than the preset efficiency threshold, even if installing the shielding ring does not cause additional losses or other side effects on the winding side, the stray losses of the structural components that it can reduce are not enough to cover the material and assembly costs of the shielding ring.

[0081] The magnetic shielding ring position optimization method provided by this invention, by constructing an interception efficiency prediction factor based on the leakage magnetic field vector direction characteristics, suppresses the problem of the traditional scalar amplitude positioning method ignoring the magnetic flux deflection angle, and can screen out candidate spaces with high loss reduction potential.

[0082] On the other hand, the additional eddy current loss penalty on the winding at each position within the candidate location domain of the shielding ring, based on the magnetic mirror effect, can be:

[0083] Step 301: Equivalently treat the distributed ampere-turns at the ends of the transformer windings as a concentrated line current source;

[0084] In practice, the field sources distributed in the multi-layer windings need to be simplified into an equivalent excitation model to facilitate subsequent analytical calculations.

[0085] The location of the concentrated line current source can be determined in the following way:

[0086] Obtain the ampere-turns and geometric thickness parameters of each conductor layer at the end of the transformer winding;

[0087] The weighted center position of each conductor layer is calculated by weighting the ampere-turns of each conductor layer in combination with the geometric thickness parameters.

[0088] The location of the weighted center is determined as the location of the concentrated line current source.

[0089] For example, the specific calculation formula is as follows:

[0090] D=d _0 +(Σ(NI _i ×(Σ(t _j +g _j )+t _i / 2))) / Σ(NI _i );

[0091] In the formula, D is the distance from the equivalent line source to the inner side of the shielding ring, and d _0NI is the normal distance from the inner side of the shielding ring to the center of the outermost conductor on the winding end face. _i Let t be the ampere-turn value of the i-th conductor. _j Let g be the thickness of the j-th conductor layer. _j t is the interlayer insulation spacing. _i Let be the thickness of the i-th conductor layer, and Σ represent the summation operation over the corresponding layer interval.

[0092] In some scenarios, the summation range of the inner layer summation Σ(tj+gj) can be from j=1 to j=i-1, representing the cumulative value of the conductor thickness and interlayer insulation spacing of each layer from the outermost layer of the winding end face to the i-th layer.

[0093] When i=1, the summation of the inner layer is 0.

[0094] Step 302: The inner surface of the shielding ring at the candidate position of the shielding ring is equivalent to a plane magnetic mirror, and a mirror line current source is established at a position symmetrical with respect to the plane magnetic mirror.

[0095] The magnetic field boundary condition for the surface of a high-permeability shielding ring requires that the magnetic field lines be tangential to the surface. This effect is equivalent to the magnetic field distribution generated by placing a mirror current in the same direction at a mirror-symmetrical position. By treating the inner edge of the shielding ring as a planar magnetic mirror, an equivalent mirror source with the same direction as the real concentrated line current source is established on the other side at a distance D from the mirror surface.

[0096] Step 303: For each layer of conductors at the winding end, determine the mirror image coefficient based on the material properties of the planar magnetic mirror;

[0097] Accordingly, the in-plane equivalent relative permeability of the shielding ring laminate structure is extracted. This equivalent relative permeability is obtained by multiplying the relative permeability of the silicon steel material at the working magnetic flux density by the lamination coefficient.

[0098] The specific calculation formula is as follows:

[0099] κ _m =(μ _eff -1) / (μ _eff +1);

[0100] In the formula, κ _m μ is the mirror image coefficient. _eff The equivalent relative permeability of the shielding ring lamination.

[0101] Step 304: Combine the mirror coefficient, concentrated line current source and mirror line current source to calculate the superposition coefficient of the additional magnetic field generated by the mirror field at each conductor layer and the original leakage magnetic field; this is used to quantify the enhancement factor of the mirror effect on the local magnetic field of the winding.

[0102] Furthermore, a first distance from each conductor layer to the concentrated current source is determined, and a second distance from each conductor layer to the mirror current source is also determined. Based on the ratio of the first distance to the second distance, and the mirror coefficient, the superposition coefficient of each conductor layer is determined.

[0103] For example, β _i =1+κ _m ×(δ _i / (2D-δ _i ));

[0104] In the formula, β _i δ is the superposition coefficient of the i-th conductor layer. _i Let 2D-δ be the first distance from the i-th layer conductor to the concentrated line current source. _i It represents the second distance from the i-th layer conductor to the current source of the mirror line.

[0105] In the first operating condition, if the ratio of the radius of curvature of the inner edge of the shielding ring to the minimum gap between the shielding ring and the winding is greater than or equal to the preset curvature threshold, the planar mirror approximation is valid, and the above calculation results are directly adopted.

[0106] In some embodiments, the preset curvature threshold is typically set to 10.

[0107] In the second operating condition, if the ratio of the radius of curvature of the inner edge of the shielding ring to the minimum gap between the shielding ring and the winding is less than the preset curvature threshold, a curvature correction factor is introduced to perform cylindrical mirror correction on the superposition coefficient.

[0108] Next, the ratio of the minimum gap between the shielding ring and the winding to twice the radius of curvature of the inner edge surface is calculated, and 1 is added to this ratio to obtain the curvature correction factor. This correction factor is multiplied into the second term of the above formula to compensate for the additional magnetic field enhancement caused by the surface focusing effect.

[0109] As an alternative, for the outer conductor closest to the shielding ring, the far-field approximation deviation increases due to its proximity to the equivalent line source. Therefore, the layer-by-layer independent mirror method can be used for calculation.

[0110] That is, each conductor layer is treated as an independent distributed field source, and the field strength contribution generated by its own image and the images of other layers at the conductor layer is calculated separately and linearly superimposed to obtain a more accurate local superposition coefficient.

[0111] Step 305: Based on the superposition coefficient of each conductor layer and the pre-obtained original eddy current loss, the additional eddy current loss penalty is calculated by weighting.

[0112] The eddy current loss inside the conductor is proportional to the square of the local magnetic field strength. The superposition coefficient of each conductor layer is squared, and subtracting 1 represents the net increase in loss for that layer. This net increase is multiplied by the original eddy current loss in the unshielded state, and then summed across the entire winding range, taking into account the loss proportion weight of each conductor layer, to obtain the overall additional eddy current loss penalty.

[0113] In some embodiments, P _penalty =Σ(w _i ×((β _i ) 2 -1)×P _eddy_i );

[0114] Among them, P _penalty To add eddy current loss penalty, w _i β represents the weight of the loss proportion of the i-th conductor. _i P is the superposition coefficient of the i-th conductor layer. _eddy_i Let Σ be the original eddy current loss of the i-th layer conductor without a shielding ring, and let Σ represent the summation operation over all winding conductors.

[0115] On the other hand, in order to avoid performing time-consuming finite element simulations for each candidate location, analytical estimation is used to obtain the structural component loss reduction rate.

[0116] Specifically, the candidate location domains of the shielding ring can be screened by lower bound constraints according to preset net benefit conditions to obtain the candidate location domains after mirror constraints. The specific steps include the following:

[0117] Step 401: Obtain the structural component loss reduction rate at each location within the candidate location domain of the shielding ring. This is specifically estimated analytically using the following method:

[0118] Step 4011: Starting from the candidate position of the shielding ring, guided by the vector direction in the leakage magnetic field distribution data, trace along the direction of the magnetic field lines away from the winding until it intersects with the surface of the structural component.

[0119] In this embodiment, the baseline leakage magnetic field vector distribution under unshielded conditions is extracted, and numerical steps are performed along the magnetic field lines using the fourth-order Runge-Kutta integral method. This process simulates the natural flow path of the magnetic flux intercepted by the shielding ring when it is not interfered with.

[0120] Step 4012: The area enclosed by the intersections of all magnetic field line tracing trajectories and the surface of the structural component is determined as the affected structural component area;

[0121] As the tracking process progresses, magnetic field lines will fall onto the surface of structural components such as the tank wall or clamps. The set of intersections formed by the tracking trajectories originating from multiple starting points on the surface of the structural components defines the boundary where the intercepted magnetic flux would originally have generated eddy current losses.

[0122] Step 4013: Integrate the eddy current loss density in the affected structural component area to obtain the local reduction amount;

[0123] Specifically, the eddy current loss density can be a pre-obtained reference eddy current loss density.

[0124] Optionally, the surface eddy current loss density distribution data of the affected area can be extracted from the unshielded baseline simulation results and subjected to area integral calculation. The integral result represents the absolute value of local hot spot loss that can be eliminated if a shielding ring is placed at this location.

[0125] Step 4014: Calculate the ratio of the local reduction amount to the structural component's reference stray loss to obtain the structural component's loss reduction rate.

[0126] For example, by dividing the local reduction obtained from the above integration by the global structural reference stray loss, the dimensionless reduction ratio can be obtained. This analytical estimation method can complete the evaluation of all candidate locations using only one reference simulation result, reducing the computational cost of the traversal process.

[0127] As another possible implementation, when it is necessary to verify a few key candidate locations, the finite element simulation method with ideal shielding conditions can be used.

[0128] Specifically, an isotropic high permeability medium region is established at the candidate location, and the eddy current loss integral of the structural component region is extracted after resolving the magnetic field. The reduction rate is obtained by subtracting it from the unshielded reference value.

[0129] Step 402: Combine the pre-obtained structural component reference stray loss and structural component loss reduction rate to determine the structural component loss reduction amount;

[0130] For example, by multiplying the pre-extracted total stray loss of structural components by the reduction rate mentioned above, the actual protection benefit that the shielding ring can provide at the current candidate location can be quantified.

[0131] Step 403: Define the condition that the reduction in structural component losses is greater than the increase in winding losses caused by the additional eddy current loss penalty as the net benefit condition.

[0132] Since installing a shielding ring will simultaneously reduce structural losses and increase additional winding losses, the installation location only has practical engineering application value when the positive benefits outweigh the negative penalties.

[0133] The specific relational conditions can be described as follows:

[0134] η _shield (r _0 ,z _0 )×P _struct_0 >(F _mirror (r _0 ,z _0 )-1)×Σ(P _eddy_i );

[0135] Where, η _shield (r _0 ,z _0 ) is the coordinate (r) _0 ,z _0 Structural component loss reduction rate at location P _struct_0 For the reference stray loss of structural components, F _mirror (r _0 ,z _0 P is the incremental factor corresponding to the additional eddy current loss penalty. _eddy_i Let Σ be the original eddy current loss of the i-th layer conductor, and let Σ represent the summation operation over each layer of conductors in the winding.

[0136] In the above text, F _mirror (r _0 ,z _0 It can be defined as the mean of the squares of the weighted superposition coefficients of the entire winding.

[0137] When it is 1, it indicates that there is no additional loss;

[0138] When it is greater than 1, it indicates that the winding loss increases due to the mirror effect.

[0139] In this step, a criterion for global loss optimization is established.

[0140] Step 404: Extract the critical net benefit distance from the shielding ring to the nearest winding conductor when the net benefit condition is met, remove positions in the candidate position domain of the shielding ring that are less than the critical net benefit distance, and update the candidate position domain after mirror constraint.

[0141] When the above net benefit condition is equal, the corresponding gap between the shielding ring and the winding is the critical net benefit distance.

[0142] According to the embodiments of this application, since the structural component loss reduction rate decreases monotonically as the shielding ring moves away from the winding, and the mirror penalty decreases monotonically even faster as the shielding ring moves away from the winding, the difference between the two is negative at the near end and positive at the far end. Therefore, the net benefit condition is equal only for a single critical distance.

[0143] The physical boundaries described above replace traditional empirical safety clearances and can be adaptively adjusted according to changes in transformer capacity and winding structure. Coordinates in the candidate location domain that result in the distance from the toroidal body to the nearest winding conductor being less than this critical value can be eliminated.

[0144] In this step, our method addresses the triangular adversarial relationship in the location optimization process.

[0145] The interception efficiency criterion drives the shielding ring closer to the winding end where the leakage magnetic field is dense, while the mirror penalty criterion drives the shielding ring away from the winding to increase the installation gap. The subsequent saturation margin criterion requires the shielding ring to avoid the flux extremum region. The optimal directions of these three constraints are mutually contradictory, and it is impossible to simultaneously satisfy the extrema at a single location. The lower bound constraint screening in this step lays the foundation for finding a compromise balance point in the multi-dimensional constraint space.

[0146] Alternatively, this method uses the plane magnetic mirror theory to establish a mirror model of the concentrated line current source, quantifies the additional winding losses caused by the shielding ring, and derives the critical net benefit distance that reduces global losses, thus defining the physical lower bound for the installation location.

[0147] According to one aspect of this application, an exemplary scheme for assessing the risk of local saturation breakdown of the inner edge of the shielding ring within the candidate location domain after mirror constraint, and for performing upper bound constraint screening on the candidate location domain after mirror constraint according to a preset saturation margin constraint is further described.

[0148] In one possible implementation, the steps may specifically include:

[0149] Step 501: Based on the geometric dimensions and equivalent chamfer features of the shielding ring cross-section, calculate the edge aggregation factor at the inner corner of the shielding ring, including:

[0150] Step 5011: The local area of ​​the inner edge of the shielding ring is equivalent to a two-dimensional right-angled wedge field, and the magnetic field singularity distribution at the corner of the outer space approaching the inner edge of the shielding ring is derived using the conformal mapping method.

[0151] In some embodiments, when magnetic flux enters the finite cross-section of the shielding ring from the external three-dimensional open space, it can converge via the inner edge surface. At the geometric corners of the inner edge, the magnetic flux path suddenly contracts.

[0152] Furthermore, by using conformal mapping algorithm to mathematically transform the external air domain of the right-angled wedge, it is derived that the magnetic field strength near the corner point exhibits a singular distribution characteristic with a predetermined exponent, thus providing a theoretical basis for quantifying the edge clustering effect.

[0153] Step 5012: Obtain the single-layer lamination thickness of the shielding ring lamination structure and the pre-configured process correction coefficient, and determine the product of the two (single-layer lamination thickness and process correction coefficient) as the equivalent radius of curvature corresponding to the equivalent chamfer feature.

[0154] Or, specifically, the product of half the thickness of a single-layer laminate and the process correction factor is determined as the equivalent radius of curvature corresponding to the equivalent chamfer feature;

[0155] The corner points of the actually manufactured shielding ring have a finite radius of curvature. This radius of curvature is determined by the thickness of the laminate and the manufacturing process.

[0156] Accordingly, half the thickness of a single-layer laminate can be used as a reference and multiplied by the aforementioned process correction factor to compensate for the equivalent chamfer effect caused by inter-laminate insulation coatings, minute air gaps, and assembly deformation.

[0157] When setting the process correction factor, it can be determined by referring to a table or through experimentation based on the specific shielding ring manufacturing process. For example, for a stacked structure using conventional shearing and without impregnation treatment, the factor is set to 1.5.

[0158] For structures that have undergone overall impregnation treatment, the value is between 1.8 and 2.0;

[0159] For laminates produced using laser cutting technology, the value is between 2.0 and 2.5.

[0160] Step 5013: Determine the edge aggregation factor based on the power-law relationship between the radial half-width of the shielding ring cross-section and the equivalent radius of curvature.

[0161] The singular field is truncated at its equivalent radius of curvature as a reference for estimating the peak field strength, and the field strength at the centerline of the cross-section is used as a reference for the average field strength. The ratio of the two reflects the flux concentration factor at the corner points. That is:

[0162] k _edge =((w / 2) / r _tip ) 1 / 3 ;

[0163] Where, k _edge The edge aggregation factor is w, where w is the radial width of the shielding ring cross-section, w / 2 represents the radial half-width, and r is the edge aggregation factor. _tip It is the equivalent radius of curvature.

[0164] Step 5014: Calculate the incident magnetic flux entering the shielding ring based on the leakage magnetic field distribution data, and determine the average magnetic flux density of the cross section in combination with the cross-sectional area of ​​the shielding ring.

[0165] By performing a surface integral operation on the normal component of the magnetic flux density in the leakage magnetic field distribution data along the geometric surface where the inner edge of the shielding ring is located, the total magnetic flux entering the ring can be obtained.

[0166] Next, the total magnetic flux is divided by the product of the geometric area of ​​the shielding ring cross section and the lamination coefficient to obtain the cross-sectional average magnetic flux density assuming uniform magnetic flux distribution.

[0167] Step 502: Multiply the edge aggregation factor by the cross-sectional average magnetic flux density to obtain the local peak magnetic flux density at the inner corner of the shielding ring.

[0168] Based on this, restoring the average magnetic flux density to the true magnetic flux density at the corner points helps to reveal the physical phenomenon that when the average magnetic flux density is far below the saturation point, local corner points may have already entered a high magnetization state.

[0169] It should be understood that this step can be used to correct the limitations of the traditional average magnetic flux density criterion.

[0170] Step 503: Combine the preset saturation magnetic flux density of the shielding ring material with the local peak magnetic flux density to assess the risk of local saturation breakdown.

[0171] For example, the candidate location domains after mirror constraints are filtered by upper bound constraints according to the preset saturation margin constraints to obtain the comprehensive feasible location domains. This can be achieved by the following method:

[0172] Furthermore, the difference between the saturation magnetic flux density and the local peak magnetic flux density of the shielding ring material is calculated;

[0173] The knee point magnetic flux density on the BH curve of silicon steel is obtained as the reference value of the material's saturation magnetic flux density. The above-mentioned local peak magnetic flux density is subtracted to quantify the margin of the current candidate position that causes local magnetic saturation.

[0174] Furthermore, the ratio of the difference to the saturation magnetic flux density of the shielding ring material is defined as the inner edge saturation margin;

[0175] Accordingly, the balance is divided by the material's saturation magnetic flux density to obtain a dimensionless percentage parameter.

[0176] That is, M _sat =(B _sat -B _peak ) / B _sat ;

[0177] Among them, M _sat B represents the inner edge saturation margin. _sat B is the saturation magnetic flux density of the shielding ring material. _peak This represents the local peak magnetic flux density at the inner corner of the shielding ring.

[0178] Furthermore, the condition that the inner edge saturation margin is not lower than the preset safety threshold value is defined as the saturation margin constraint;

[0179] For example, the preset safety threshold can be set to 0.15. This value takes into account the combined effects of transient overcurrent, manufacturing tolerances, and material property dispersion; the allocation details will not be listed item by item.

[0180] Furthermore, after removing the positions in the candidate position domain that do not satisfy the saturation margin constraint after removing the mirror constraint, the comprehensive feasible position domain is updated to obtain the comprehensive feasible position domain.

[0181] When the inner edge saturation margin falls below the threshold, the local permeability at the corner will collapse, which is equivalent to the formation of a high magnetic reluctance crack on the cross-section. This crack will force the originally guided magnetic flux to leak out in a concentrated manner, forming a high-density leakage flux bundle on the surface of the downstream structural component, thereby causing more severe local hot spots than in the unshielded state.

[0182] Eliminating dangerous locations can prevent the risk of cascading failures.

[0183] It should be understood that the location with the highest interception efficiency often corresponds to the region with the highest leakage magnetic flux density. The magnetic flux entering the shielding ring is also the largest in this region, making it most likely to trigger inner edge saturation. That is, the optimal interception location and the saturation danger location overlap in spatial distribution. After this step of screening, we are actually looking for a set of Pareto compromise locations that meet the constraints on the overlapping boundary between the high-efficiency region and the high-saturation region. This set constitutes the final comprehensive feasible location domain.

[0184] In some embodiments, an optional implementation method is further described, which combines the effective attraction range of the shielding ring under multiple load conditions and performs optimization calculations with the goal of minimizing total stray loss to determine the optimal installation position of the magnetic shielding ring; specifically including:

[0185] Step 601, obtain the static attraction radius of the shielding ring, including:

[0186] Accordingly, the radial width and axial height of the shielding ring cross-section are obtained;

[0187] In some embodiments, the geometric parameters of the shielding ring design are read. These parameters determine the current-carrying area of ​​the shielding ring cross-section and are the basis for evaluating its ability to statically attract leakage magnetic flux to external space.

[0188] Accordingly, the static attraction radius is determined based on the square root of the product of radial width and axial height, combined with pre-configured geometric constants;

[0189] The radial width and axial height are multiplied and then squared to obtain the characteristic cross-sectional dimension with standard length dimensions. Next, this characteristic dimension is multiplied by a dimensionless geometric constant calibrated through finite element modeling to obtain the theoretical maximum attraction range of the shielding ring under conditions of no magnetic saturation.

[0190] The specific calculation formula is as follows:

[0191] ρ _att =k _geom ×sqrt(w×h);

[0192] Where, ρ _att Let k be the static attraction radius. _geom Here are the pre-configured geometric constants, w is the radial width of the shielding ring cross-section, h is the axial height of the shielding ring cross-section, and sqrt is the square root operation.

[0193] Step 602: For each load condition, determine the corresponding equivalent relative permeability based on the working magnetic flux density inside the shielding ring.

[0194] Transformers face varying load currents during actual operation, causing changes in leakage magnetic field strength. For multiple preset typical load conditions, the average magnetic flux density within the shielding ring cross-section can be extracted. Combined with the magnetization curve of silicon steel, the secant permeability at the corresponding operating point can be obtained by looking up a table or interpolation, reflecting the nonlinear characteristics of the material.

[0195] Step 603: Using the square root of the ratio of the equivalent relative permeability to the initial permeability of the shielding ring material, the static attraction radius is saturated nonlinearly corrected to obtain the dynamic equivalent attraction radius under each load condition.

[0196] As the load current increases, the magnetic flux density inside the shielding ring rises and approaches the saturation point, causing its actual permeability to decrease, which in turn shrinks the effective attraction range. By calculating the ratio of the current operating permeability to the initial permeability in the linear region, and taking the square root of this ratio as a correction coefficient, which is then multiplied into the static attraction radius, this dynamic shrinkage law with load variation can be quantified.

[0197] That is, ρ _eff =ρ _att ×sqrt(μ _eff (B _avg ) / μ _eff_0 );

[0198] Where, ρ _eff For the dynamic equivalent attraction radius, ρ _att Let μ be the static attraction radius. _eff (B _avg ) represents the equivalent relative permeability at the operating magnetic flux density, μ _eff_0 Let be the initial magnetic permeability of the shielding ring material.

[0199] Step 604: The effective attraction range is characterized by the dynamic equivalent attraction radius;

[0200] The dynamic equivalent attraction radius defines the maximum spatial distance at which the shielding ring can cause the leakage flux deflection angle to exceed a predetermined angle, providing a quantitative boundary for subsequent assessment of the risk of off-target failure under light load conditions.

[0201] According to an embodiment of this application, optimization calculations are performed with the goal of minimizing total stray loss to determine the optimal installation position of the magnetic shielding ring, including:

[0202] Optionally, for each load condition, it is determined whether the dynamic equivalent attraction radius is less than the shortest distance from the main path of leakage flux to the center of the shielding ring;

[0203] In some cases, the main path of leakage flux is extracted based on leakage magnetic field distribution data, and the shortest distance from the main path of leakage flux to the center of the shielding ring is calculated.

[0204] Accordingly, the coordinates of the center line of the main path of leakage flux under each working condition in the unshielded state are extracted, and the spatial distance from the center line to the installation center of the shielding ring is calculated.

[0205] By further comparing the distance with the dynamic equivalent attraction radius mentioned above, it can be determined whether the shielding ring can still effectively cover and intercept the main magnetic flux under the current load level.

[0206] Optionally, if so, the square of the ratio of the dynamic equivalent attraction radius to the nearest distance is defined as the attenuation coefficient, and the attenuation coefficient is used to calculate and correct the residual loss of the structural components under the corresponding working conditions.

[0207] If not, the attenuation coefficient for the corresponding operating condition is 1.

[0208] When the judgment result is yes, it indicates that the leakage magnetic field distribution is narrowed under light load conditions, the shielding ring is outside the main magnetic flux path, and its interception capability is reduced.

[0209] Based on the physical laws of far-field attenuation, the square of the ratio of the two distances is calculated as the attenuation coefficient and applied to the calculation of the residual loss of the structural components to punish the shielding failure phenomenon at this location under light load.

[0210] That is, γ _k =(ρ _eff / D _path ) 2 ;

[0211] Where, γ _k ρ is the attenuation coefficient. _eff For the dynamic equivalent attraction radius, D _path This is the shortest distance from the main leakage flux path to the center of the shielding ring.

[0212] Among them, the residual loss of structural components can be the difference between the pre-obtained reference stray loss of structural components and the loss reduction of structural components at the corresponding location;

[0213] Optionally, the corrected residual losses of structural components, additional eddy current loss penalties, and shielding ring losses under each load condition are summed and weighted by the preset time proportion weights of each load condition to construct the total stray loss objective function.

[0214] The three main loss components are summarized to form the total loss assessment value under a single operating condition. Next, the typical load duration curve of the transformer is obtained, and weights are assigned according to the time proportion of each load interval. The total loss of each operating condition is then weighted and summed.

[0215] In some embodiments, the specific formula can be expressed as:

[0216] F(r _0 ,z _0 )=Σ(λ _j ×(P _stray_j (r _0 ,z _0 )+P _penalty_j (r _0 ,z _0 )+P _ring_j (r _0 ,z _0 )));

[0217] Where F(r) _0 ,z _0 ) is the coordinate (r) _0 ,z _0 The objective function for the total stray loss at point λ is... _j P represents the time percentage weight for the j-th load condition. _stray_j (r _0 ,z _0 P represents the corrected residual loss of the structural component. _penalty_j (r _0 ,z _0 P is the additional eddy current loss penalty. _ring_j (r _0 ,z _0 ) represents the loss of the shielding ring itself, and Σ represents the weighted summation over each load condition.

[0218] Optionally, the total stray loss objective function is calculated by traversing the comprehensive feasible location domain, and the location corresponding to the minimum value of the objective function is determined as the optimal installation location.

[0219] Within the comprehensive feasible location domain retained after multiple constraint reductions, the objective function is calculated point-by-point analytically. Since this optimization process relies only on a small amount of unmasked baseline simulation data, and the entire traversal process is completed through algebraic operations, computational efficiency is improved.

[0220] At this point, outputting the coordinate parameters that make the objective function reach its global minimum can be considered as the final recommended solution that balances high interception, low penalty, and robustness across multiple operating conditions.

[0221] According to an embodiment of this application, the inherent loss of the shielding ring can be calculated in the following way:

[0222] Furthermore, the cross-sectional non-uniformity correction coefficient of the shielding ring and the effective core volume are obtained;

[0223] In addition to extracting the physical volume parameters of the ring, a correction coefficient greater than 1 needs to be introduced to compensate for the additional loss increment caused by the local high magnetic flux density region.

[0224] Furthermore, by combining the average magnetic flux density of the shielding ring cross section under the corresponding load conditions, the shielding ring's own loss can be calculated using a preset alternating magnetization loss formula.

[0225] For example, the conventional alternating magnetization loss formula is used to process the cross-sectional average magnetic flux density data, and the calculation result is multiplied by the cross-sectional non-uniformity correction coefficient and the effective core volume; thus avoiding the problem of underestimation of iron loss caused by simply relying on the average magnetic flux density.

[0226] Among them, the average magnetic flux density of the shielding ring cross section can be calculated based on the leakage magnetic field distribution data and the current ratio of the corresponding load conditions, and determined in combination with the cross-sectional area of ​​the shielding ring.

[0227] Furthermore, the non-uniformity correction coefficient is used to compensate for the loss increment in the local high flux density region caused by the accumulation effect at the edge of the shielding ring;

[0228] Furthermore, a dynamic equivalent attraction radius and distance attenuation mechanism corrected by nonlinear permeability is introduced, quantifying the off-target effect caused by leakage magnetic field contraction under light load. Under the multi-condition weighted optimization framework, it is beneficial to ensure that the final determined installation position can maintain good stray loss suppression performance throughout the full load range of the transformer.

[0229] In a further embodiment, the verification step after determining the optimal installation position of the magnetic shielding ring specifically includes:

[0230] Step 701: Apply the magnetic shielding ring at the optimal installation position to the pre-constructed three-dimensional finite element calculation model of the transformer;

[0231] In other words, while the two-dimensional axisymmetric model has high computational efficiency, it cannot cover the physical boundaries of asymmetric regions such as the corners of the fuel tank. Therefore, the optimal radial and axial coordinates output by the two-dimensional analytical optimization can be obtained, and these position parameters, along with the corresponding shielding ring cross-sectional dimensions, can be mapped into a complete three-dimensional model to complete the construction of the verification environment.

[0232] Step 702: Solve the global magnetic field under multiple load conditions and verify the eddy current loss density on the surface of the structural component and the magnetic flux density of the winding conductor.

[0233] In this step, the finite element method (FEM) algorithm is used to calculate the three-dimensional magnetic field distribution under multiple typical load conditions. Loss distribution data of the tank wall and clamping surface are extracted to confirm that there are no abnormal hot spots caused by inner edge saturation leakage. The criterion is that the maximum eddy current loss density does not exceed the corresponding value under unshielded conditions.

[0234] Simultaneously, magnetic flux density data at the winding conductor closest to the shielding ring is extracted to verify whether the magnetic field enhancement at that location is within the penalty range predicted by the mirror model.

[0235] Step 703: If the verification passes, output the final magnetic shielding ring installation position parameters;

[0236] In other cases, if the verification fails, the optimal installation position is finely adjusted within the comprehensive feasible location domain and then re-verified until the verification passes.

[0237] When the hot spot loss of the structural components and the magnetic flux density of the winding side meet the preset design thresholds under all load conditions, the global effectiveness of the optimal installation position is confirmed.

[0238] In this case, the final parameters, including radial coordinates, axial coordinates, cross-sectional dimensions, and safety margins, can be summarized and output for subsequent engineering manufacturing and assembly.

[0239] As an optional implementation, for the direction where the end of the side phase winding faces the corner of the box wall, due to the more obvious three-dimensional spatial effect, the local magnetic flux density in the corner region is usually higher than the predicted value of the two-dimensional symmetric model, which may cause the verification to fail on the first try.

[0240] Optionally, the preliminary position determined by the two-dimensional optimization can be used as the initial iteration value of the three-dimensional model. A local fine-tuning search can be performed in the three-dimensional space along the radial or axial direction. The global magnetic field solution and verification can be re-executed until all indicators meet the requirements.

[0241] The above describes how conformal mapping technology reveals the edge aggregation effect at the inner corner of the shielding ring. By setting a local saturation margin constraint, the risk of abnormal hot spots caused by concentrated leakage flux under extreme operating conditions is suppressed.

[0242] The specific selections, parameter values, and process steps given in this application are merely exemplary embodiments and are not intended to limit the scope of protection of this invention. Based on the core technical logic of this invention, the devices, parameters, and implementation methods can be equivalently replaced and adjusted according to the actual application scenario.

[0243] It should be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

Claims

1. A method for optimizing the position of a magnetic shielding ring to suppress stray losses at the ends of transformer windings, characterized in that, include: Obtain leakage magnetic field distribution data in the transformer winding end region; The interception efficiency of the shielding ring is evaluated based on the vector direction characteristics in the leakage magnetic field distribution data, and the candidate location domain of the shielding ring is selected accordingly. Based on the magnetic mirror effect, the additional eddy current loss penalty generated by each position in the candidate position domain of the shielding ring on the winding is quantified. The candidate position domain of the shielding ring is screened by lower bound constraint according to the net benefit condition to obtain the candidate position domain after mirror constraint. The risk of local saturation breakdown at the inner edge of the shielding ring within the candidate location domain after mirror constraint is evaluated. Based on the saturation margin constraint, the candidate location domain after mirror constraint is screened by upper bound constraint to obtain the comprehensive feasible location domain. Within the comprehensive feasible location domain, and considering the effective attraction range of the shielding ring under multiple load conditions, optimization calculations are performed with the goal of minimizing total stray loss to determine the optimal installation location of the magnetic shielding ring.

2. The method according to claim 1, characterized in that, The interception efficiency of the shielding ring is evaluated based on the vector direction characteristics in the leakage magnetic field distribution data, including: Extract the radial and axial components of the leakage magnetic field from the leakage magnetic field distribution data; The magnetic flux turning angle at each location is determined based on the radial and axial components of the leakage magnetic field, and a spatial spectrum of the magnetic flux turning angle is constructed. The magnetic flux density amplitude distribution is calculated by combining the radial and axial components of the leakage magnetic field. The interception efficiency prediction factor is calculated based on the spatial spectrum of magnetic flux steering angle and the amplitude distribution of magnetic flux density.

3. The method according to claim 1, characterized in that, The additional eddy current loss penalty on the winding at each position within the candidate location domain of the shielding ring is quantified based on the magnetic mirror effect, including: The distributed ampere-turns at the ends of the transformer windings are equivalent to a concentrated line current source; The inner surface of the shielding ring at the candidate position is equivalent to a plane magnetic mirror, and a mirror line current source is established at a position symmetrical to the plane magnetic mirror. For each layer of conductors at the winding ends, the mirror coefficient is determined based on the material properties of the planar magnetic mirror; By combining the image coefficient, the concentrated line current source, and the image line current source, the superposition coefficient of the additional magnetic field generated by the image field at each conductor layer and the original leakage magnetic field is calculated. Based on the superposition coefficients of each conductor layer and the pre-obtained original eddy current losses, an additional eddy current loss penalty is calculated using a weighted average.

4. The method according to claim 3, characterized in that, Based on the net benefit condition, the candidate location domains of the shielding ring are filtered by lower bound constraints to obtain the candidate location domains after mirror constraints, including: Obtain the structural component loss reduction rate at each location within the candidate location domain of the shielding ring; The amount of structural component loss reduction is determined by combining the pre-obtained reference stray loss and structural component loss reduction rate. The condition that the reduction in structural component losses is greater than the increase in winding losses caused by the additional eddy current loss penalty is defined as the net benefit condition. Extract the critical net benefit distance from the shielding ring to the nearest winding conductor when the net benefit condition is met, remove positions in the candidate position domain of the shielding ring that are less than the critical net benefit distance, and update the candidate position domain after mirror constraint.

5. The method according to claim 1, characterized in that, Assess the risk of local saturation breakdown at the inner edge of the shielding ring within the candidate location domain after mirror constraint, including: Based on the geometric dimensions and equivalent chamfer characteristics of the shielding ring cross-section, the edge aggregation factor at the inner corner of the shielding ring is calculated. The incident magnetic flux entering the shielding ring is calculated based on the leakage magnetic field distribution data, and the average magnetic flux density of the cross section is determined by combining the cross-sectional area of ​​the shielding ring. Multiplying the edge aggregation factor by the cross-sectional average magnetic flux density yields the local peak magnetic flux density at the inner corner of the shielding ring. By combining the saturation magnetic flux density and the local peak magnetic flux density of the shielding ring material, the risk of local saturation breakdown is assessed.

6. The method according to claim 5, characterized in that, Based on the geometric dimensions and equivalent chamfer characteristics of the shielding ring cross-section, the edge aggregation factor at the inner corner of the shielding ring is calculated, including: The local region of the inner edge of the shielding ring is equivalent to a two-dimensional right-angled wedge field, and the magnetic singularity distribution of the external space approaching the inner edge corner of the shielding ring is derived by using the conformal mapping method. Obtain the single-layer lamination thickness of the shielding ring lamination structure and the pre-configured process correction coefficient, and determine the product of the two as the equivalent radius of curvature corresponding to the equivalent chamfer feature; The edge aggregation factor is determined based on the power-law relationship between the radial half-width of the shielding ring cross-section and the equivalent radius of curvature.

7. The method according to claim 1, characterized in that, Based on the saturation margin constraint, the candidate location domains after the mirror constraint are filtered by upper bound constraint to obtain the comprehensive feasible location domains, including: Calculate the difference between the saturation magnetic flux density and the local peak magnetic flux density of the shielding ring material; The ratio of the difference to the saturation magnetic flux density of the shielding ring material is defined as the inner edge saturation margin. The condition that the inner edge saturation margin is not lower than the preset safety threshold value is defined as the saturation margin constraint. After removing the mirror constraint, the candidate location domains that do not satisfy the saturation margin constraint are updated to obtain the comprehensive feasible location domain.

8. The method according to claim 1, characterized in that, The effective attraction range of the shielding ring under multiple load conditions includes: Obtain the static attraction radius of the shielding ring; For each load condition, the corresponding equivalent relative permeability is determined based on the working magnetic flux density within the shielding ring. By using the square root of the ratio of the equivalent relative permeability to the initial permeability of the shielding ring material, the static attraction radius is saturated nonlinearly corrected to obtain the dynamic equivalent attraction radius under various load conditions. The effective attraction range is characterized by the dynamic equivalent attraction radius.

9. The method according to claim 1, characterized in that, The optimal installation location of the magnetic shielding ring is determined by optimization calculations with the goal of minimizing total stray loss, including: For each load condition, determine whether the dynamic equivalent attraction radius is less than the shortest distance from the main leakage flux path to the center of the shielding ring; If so, the square of the ratio of the dynamic equivalent attraction radius to the nearest distance is defined as the attenuation coefficient, and the residual loss of the structural components under the corresponding working condition is calculated and corrected using the attenuation coefficient. The modified residual loss of structural components, additional eddy current loss penalty and shielding ring loss under each load condition are summed, and weighted by the time proportion weight of each load condition to construct the total stray loss objective function. The total stray loss objective function is calculated by traversing the comprehensive feasible location domain, and the location corresponding to the minimum value of the objective function is determined as the optimal installation location.

10. The method according to claim 1, characterized in that, After determining the optimal installation location of the magnetic shielding ring, the following steps are also included: The magnetic shielding ring at the optimal installation position is applied to the pre-constructed three-dimensional finite element calculation model of the transformer; Under multiple load conditions, the global magnetic field is solved to verify the eddy current loss density on the surface of the structural component and the magnetic flux density of the winding conductor. If the verification passes, the final magnetic shielding ring installation position parameters will be output.