阅读说明：本技术 一种用于电磁场求解的自适应区域分解有限元方法 (Self-adaptive region decomposition finite element method for electromagnetic field solution ) 是由 张云鹏 乔振阳 杨新生 张长庚 王清旋 周岐斌 于 2021-07-09 设计创作，主要内容包括：本发明公开了一种用于电磁场求解的自适应区域分解有限元方法,包括：基于对所求解电磁场问题无先验知识,对所述电磁场整个求解域进行一致剖分,生成初始离散网格；根据目标子区域的数目和所述初始离散网格,对求解任务进行均匀分配并调用区域分解求解器对离散后得到的电磁问题进行求解；基于求解结果,计算用来描述误差信息和三角形网格单元形状尺寸特征的Hessian矩阵和度量张量矩阵；对计算数值结果进行误差分析,若误差不满足预设值,则对网格做自适应处理并再次计算,直到误差满足预设值并输出数值结果。本发明能够在保证求解精度的基础上大幅提高计算效率,且在保证计算精度的同时,降低了所需自由度数目,求解效率有显著的提高。(The invention discloses a self-adaptive region decomposition finite element method for solving an electromagnetic field, which comprises the following steps: based on no prior knowledge of the solved electromagnetic field problem, carrying out consistent subdivision on the whole solved domain of the electromagnetic field to generate an initial discrete grid; according to the number of target sub-regions and the initial discrete grid, uniformly distributing solving tasks and calling a region decomposition solver to solve the electromagnetic problem obtained after dispersion; calculating a Hessian matrix and a measurement tensor matrix which are used for describing error information and shape and size characteristics of the triangular grid cells based on the solving result; and carrying out error analysis on the calculated numerical result, if the error does not meet the preset value, carrying out self-adaptive processing on the grid and calculating again until the error meets the preset value and outputting the numerical result. The invention can greatly improve the calculation efficiency on the basis of ensuring the solving precision, reduces the number of required degrees of freedom while ensuring the calculating precision, and obviously improves the solving efficiency.)

1. An adaptive region decomposition finite element method for electromagnetic field solution, comprising:

based on no prior knowledge of the solved electromagnetic field problem, carrying out consistent subdivision on the whole solved domain of the electromagnetic field to generate an initial discrete grid;

according to the number of target sub-regions and the initial discrete grid, uniformly distributing solving tasks and calling a region decomposition solver to solve the electromagnetic problem obtained after dispersion;

calculating a Hessian matrix and a measurement tensor matrix which are used for describing error information and shape and size characteristics of the triangular grid cells based on the solving result;

and carrying out error analysis on the calculated numerical result, if the error does not meet the preset value, carrying out self-adaptive processing on the grid and calculating again until the error meets the preset value and outputting the numerical result.

2. The adaptive region decomposition finite element method for electromagnetic field solution of claim 1, wherein: the area division of the entire solution domain of the electromagnetic field includes,

and reasonably dividing the solution domain into a plurality of sub solution domains according to the calculated amount based on the discrete grid of the solution problem.

3. An adaptive region decomposition finite element method for electromagnetic field solution as defined in claim 1 or 2, wherein: the electromagnetic field comprises a transient control equation based on a magnetic potential function,

wherein ν is a magnetoresistance ratio, U is a magnetic potential, σ is an electric conductivity, t is time, and J is a current density.

4. An adaptive region decomposition finite element method for electromagnetic field solution as recited in claim 3, wherein: the fully expanded algebraic equation of the electromagnetic problem obtained after the finite element dispersion based on Galerkin comprises,

wherein the diagonal block matrix (A)_ijI ═ j) characterize the contribution of the degrees of freedom in the sub-region i to the coefficient matrix, the non-diagonal block matrix (a)_ijI ≠ j) characterizes the contribution of the degree of freedom in the sub-region j to the sub-region i coefficient matrix, U_iMatrix of unknowns for sub-region i, F_iIs the load matrix for sub-region i.

5. The adaptive region decomposition finite element method for electromagnetic field solution of claim 1, wherein: the discrete equations generated based on the finite element strategy are decomposed according to the sub-regions, the iterative equations of unknown quantities in each sub-region and the whole definition region comprise,

based on the Jacobian method, the iterative equation of the unknowns is expressed as:

this equation is equivalent to:

wherein:

based on:

wherein the superscript n +1/n is the iterative computation step number, R_iFor the restriction operator from the entire domain to sub-region i, M_iMatrix of x M, M_iAnd M is the number of unknown quantities in each sub-region and the whole definition region respectively;

the iterative equation for the unknowns is finally written as:

wherein the content of the first and second substances,

rⁿ＝F-AUⁿ

wherein A is an electromagnetic field rigidity matrix, U is a magnetic potential unknown quantity matrix, F is an electromagnetic field load matrix, and superscript T is a matrix transpose.

6. The adaptive region decomposition finite element method for electromagnetic field solution of claim 1, wherein: the Hessian matrix comprises a set of one or more Hessian matrices,

hessian matrix H_kExpressed as:

H_k＝{H_k,ij}；i/j＝1,2

wherein, U_kFor the kth iterative solution of the magnetic potential, x₁、x₂Representing the independent variables in two directions x and y, respectively, of the two-dimensional space.

7. The adaptive region decomposition finite element method for electromagnetic field solution of claim 1, wherein: said H_kCorresponding metric tensor matrix M_kComprises the steps of (a) preparing a mixture of a plurality of raw materials,

M_k＝V^Tdiag(λ)V

wherein λ and V are each H_kThe eigenvalues and eigenvectors of (c).

8. An adaptive region decomposition finite element method for electromagnetic field solution as recited in any of claims 1, 6, 7, wherein: performing an error analysis on the calculated numerical result includes,

in the metric tensor matrix M determined by error information_kIn the geometric space of (a), the grid quality is determined by the following formula:

where | Δ | is the area of the cell, p (Δ) is the perimeter of the cell, N is the desired number of grid cells, h (N) is the average size of the grid cells, and F (t) (0 ≦ F ≦ 1) is an arbitrary convex function with an extreme point at t ≦ 1.

Technical Field

The invention relates to the technical field of electrical equipment and simulation calculation, in particular to a self-adaptive region decomposition finite element method for solving an electromagnetic field.

Background

With the continuous improvement of the automation level of industrial production in China, the position of electrical equipment in industrial application is more and more important, and the accurate and efficient numerical analysis method plays a key role in the optimal design and safe operation of the electrical equipment. The self-adaptive finite element method, namely, error analysis and discrete adjustment are introduced on the basis of the finite element method, and the method becomes a dominant numerical method of the current electromagnetic analysis with accuracy, flexibility and high efficiency, and is widely applied to solving the actual engineering problem. Based on the results of the error analysis in the solution of a particular problem, the method can produce an optimal grid that meets the accuracy requirements, with as few cells as possible and with nearly uniform distribution of errors. The adaptive finite element method is classified into three types, i.e., r-method, p-method and h-method, according to the discrete adjustment method used. In the h-method, according to the result of error analysis, the grid is orderly encrypted or sparse, and finally an optimal grid with errors close to uniform distribution is generated; in the solving process, encryption and sparse operation are involved aiming at different areas, so that a numerical solution which is accurate enough can be obtained by using as few computing resources as possible.

Due to the fact that nonlinear materials are mostly involved in electromagnetic analysis, part of problems are complex in structure and long in time span, the calculation scale is large, a large amount of calculation resources are needed for processing mesh subdivision and subsequent algebraic system solving, and finally the efficiency of electromagnetic analysis and design is affected. With the progress of computer hardware and parallel technology, large-scale engineering problems can be solved by introducing a regional decomposition idea into a finite element method; the area decomposition method is based on the idea that a solution domain is decomposed into a plurality of overlapped or non-overlapped sub-areas, and after boundary conditions of the sub-problems are reasonably processed, the parallel processing units are used for respectively solving the sub-problems; in the algebraic system solution formed by the region decomposition finite element method, a Krylov subspace method is mostly used in combination with a proper preprocessing operator to improve the calculation efficiency; the area decomposition method has been applied to solving the problem of the low-frequency electromagnetic field, which effectively improves the problem solving efficiency, but the traditional area decomposition finite element method still has the problems of relatively insufficient calculation precision, relatively limited solving efficiency and the like.

Disclosure of Invention

This section is for the purpose of summarizing some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. In this section, as well as in the abstract and the title of the invention of this application, simplifications or omissions may be made to avoid obscuring the purpose of the section, the abstract and the title, and such simplifications or omissions are not intended to limit the scope of the invention.

The present invention has been made in view of the above-mentioned conventional problems.

Therefore, the technical problem solved by the invention is as follows: the traditional regional decomposition finite element method has the problems of relatively insufficient calculation precision, relatively limited solving efficiency and the like.

In order to solve the technical problems, the invention provides the following technical scheme: based on no prior knowledge of the solved electromagnetic field problem, carrying out consistent subdivision on the whole solved domain of the electromagnetic field to generate an initial discrete grid; according to the number of target sub-regions and the initial discrete grid, uniformly distributing solving tasks and calling a region decomposition solver to solve the electromagnetic problem obtained after dispersion; calculating a Hessian matrix and a measurement tensor matrix which are used for describing error information and shape and size characteristics of the triangular grid cells based on the solving result; and carrying out error analysis on the calculated numerical result, if the error does not meet the preset value, carrying out self-adaptive processing on the grid and calculating again until the error meets the preset value and outputting the numerical result.

As a preferred aspect of the adaptive region decomposition finite element method for electromagnetic field solution according to the present invention, wherein: the area division of the whole solving domain of the electromagnetic field comprises the step of reasonably dividing the solving domain into a plurality of sub solving domains according to the calculated amount based on the discrete grid of the solving problem.

As a preferred aspect of the adaptive region decomposition finite element method for electromagnetic field solution according to the present invention, wherein: the electromagnetic field comprises a transient control equation based on a magnetic potential function,

wherein ν is a magnetoresistance ratio, U is a magnetic potential, σ is an electric conductivity, t is time, and J is a current density.

As a preferred aspect of the adaptive region decomposition finite element method for electromagnetic field solution according to the present invention, wherein: the fully expanded algebraic equation of the electromagnetic problem obtained after the finite element dispersion based on Galerkin comprises,

wherein the diagonal block matrix (A)_ijI ═ j) characterize the contribution of the degrees of freedom in the sub-region i to the coefficient matrix, the non-diagonal block matrix (a)_ijI ≠ j) characterizes the contribution of the degree of freedom in the sub-region j to the sub-region i coefficient matrix, U_iMatrix of unknowns for sub-region i, F_iIs the load matrix for sub-region i.

As a preferred aspect of the adaptive region decomposition finite element method for electromagnetic field solution according to the present invention, wherein: decomposing a discrete equation generated based on a finite element strategy according to sub-regions, wherein an iterative equation of unknown quantity in each sub-region and the whole definition region comprises that based on a Jacobi method, the iterative equation of the unknown quantity is expressed as:

this equation is equivalent to:

wherein:

based on:

wherein the superscript n +1/n is the iterative computation step number, R_iFor the restriction operator from the entire domain to sub-region i, M_iMatrix of x M, M_iAnd M is the number of unknown quantities in each sub-region and the whole definition region respectively;

the iterative equation for the unknowns is finally written as:

wherein the content of the first and second substances,

rⁿ＝F-AUⁿ

wherein A is an electromagnetic field rigidity matrix, U is a magnetic potential unknown quantity matrix, F is an electromagnetic field load matrix, and superscript T is a matrix transpose.

As a preferred aspect of the adaptive region decomposition finite element method for electromagnetic field solution according to the present invention, wherein: the Hessian matrix comprises a Hessian matrix H_kExpressed as:

H_k＝{H_k,ij}；i/j＝1,2

wherein, U_kFor the kth iterative solution of the magnetic potential, x₁、x₂Representing the independent variables in two directions x and y, respectively, of the two-dimensional space.

As a preferred aspect of the adaptive region decomposition finite element method for electromagnetic field solution according to the present invention, wherein: said H_kCorresponding metric tensor matrix M_kComprises the steps of (a) preparing a mixture of a plurality of raw materials,

M_k＝V^Tdiag(λ)V

wherein λ and V are each H_kThe eigenvalues and eigenvectors of (c).

As a preferred aspect of the adaptive region decomposition finite element method for electromagnetic field solution according to the present invention, wherein: performing error analysis on the computed numerical result includes determining the metric tensor matrix M from error information_kIn the geometric space of (a), the grid quality is determined by the following formula:

where | Δ | is the area of the cell, p (Δ) is the perimeter of the cell, N is the desired number of grid cells, h (N) is the average size of the grid cells, and F (t) (0 ≦ F ≦ 1) is an arbitrary convex function with an extreme point at t ≦ 1.

The invention has the beneficial effects that: the invention provides a self-adaptive region decomposition finite element method which combines the advantages of a self-adaptive method and a region decomposition method, and can greatly improve the calculation efficiency on the basis of ensuring the solving precision; a flexible additive Schwarz area decomposition method is used as a preprocessing operator of a Krylov subspace method, an anisotropic grid self-adaption method based on a measurement tensor is introduced, and an optimal grid can be generated according to magnetic field distribution after grid quality is analyzed; compared with the traditional region decomposition method, the method provided by the invention has the advantages that the number of required degrees of freedom is reduced while the calculation precision is ensured, and the solving efficiency is obviously improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise. Wherein:

FIG. 1 is a schematic diagram of a basic flow chart of an adaptive region decomposition finite element method for electromagnetic field solution according to an embodiment of the present invention;

FIG. 2 is a schematic diagram illustrating the structure and dimensions of a C-shaped actuator for an adaptive region-decomposition finite element method for electromagnetic field solution according to an embodiment of the present invention;

FIG. 3 is a diagram illustrating the convergence curves of magnetic field energy of a C-shaped actuator of an adaptive region decomposition finite element method for electromagnetic field solution according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a grid adaptation process of a C-shaped actuator of an adaptive region decomposition finite element method for electromagnetic field solution according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a magnetic potential filling equipotential line of a C-shaped actuator of an adaptive region decomposition finite element method for electromagnetic field solution according to an embodiment of the present invention;

FIG. 6 is a schematic diagram of a PMSM model with an adaptive region decomposition finite element method for electromagnetic field solution according to an embodiment of the present invention;

FIG. 7 is a schematic diagram of discrete meshes and regions of a PMSM model using an adaptive region decomposition finite element method for electromagnetic field solution according to an embodiment of the present invention;

fig. 8 is a schematic diagram of a magnetic potential filling equipotential line of a permanent magnet synchronous motor model of an adaptive region decomposition finite element method for electromagnetic field solution according to an embodiment of the present invention.

Detailed Description

In order to make the aforementioned objects, features and advantages of the present invention comprehensible, specific embodiments accompanied with figures are described in detail below, and it is apparent that the described embodiments are a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making creative efforts based on the embodiments of the present invention, shall fall within the protection scope of the present invention.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described and will be readily apparent to those of ordinary skill in the art without departing from the spirit of the present invention, and therefore the present invention is not limited to the specific embodiments disclosed below.

Furthermore, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.

The present invention will be described in detail with reference to the drawings, wherein the cross-sectional views illustrating the structure of the device are not enlarged partially in general scale for convenience of illustration, and the drawings are only exemplary and should not be construed as limiting the scope of the present invention. In addition, the three-dimensional dimensions of length, width and depth should be included in the actual fabrication.

Meanwhile, in the description of the present invention, it should be noted that the terms "upper, lower, inner and outer" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of describing the present invention and simplifying the description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation and operate, and thus, cannot be construed as limiting the present invention. Furthermore, the terms first, second, or third are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.

The terms "mounted, connected and connected" in the present invention are to be understood broadly, unless otherwise explicitly specified or limited, for example: can be fixedly connected, detachably connected or integrally connected; they may be mechanically, electrically, or directly connected, or indirectly connected through intervening media, or may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.

Example 1

Referring to fig. 1, for one embodiment of the present invention, there is provided an adaptive region decomposition finite element method for electromagnetic field solution, including:

s1: based on no prior knowledge of the solved electromagnetic field problem, carrying out consistent subdivision on the whole solved domain of the electromagnetic field to generate an initial discrete grid;

s2: according to the number of target sub-regions and the initial discrete grid, uniformly distributing solving tasks and calling a region decomposition solver to solve the electromagnetic problem obtained after the dispersion;

s3: calculating a Hessian matrix and a measurement tensor matrix which are used for describing error information and shape and size characteristics of the triangular grid cells based on the solving result;

s4: and carrying out error analysis on the calculated numerical result, if the error does not meet the preset value, carrying out self-adaptive processing on the grid and calculating again until the error meets the preset value and outputting the numerical result.

The steps S1-S4 specifically include:

the region division of the whole solution domain of the electromagnetic field comprises the following steps:

an anisotropic grid self-adaption method based on measurement tensor is introduced into an additive Schwarz region decomposition method, the advantages of an h-type self-adaption finite element method and the advantages of a region decomposition method are combined, and a grid with better response to field distribution is generated in the solving process;

specifically, a method based on solving problem discrete grids is adopted to reasonably divide the calculated amount, such as a METIS tool; the method can be more easily adapted to the grid size difference existing in the actual problem dispersion due to no limitation of the geometric size; in addition, the METIS tool can more fully utilize computing resources and has higher parallel rate. On the basis of the generated non-overlapped subdomains, overlapped subdomains can be generated by adding one or more layers of units outwards to each subdomain; therefore, the region division performed for the non-overlapping region decomposition algorithm can be easily applied to the overlapping region decomposition algorithm.

Further, the additive Schwarz regioresolution method comprises:

when a finite element method is used for solving the electromagnetic field problem, the original Maxwell equation is deformed by introducing a potential function and combining a constitutive equation of a material so as to simplify calculation, then finite element dispersion is carried out on the potential function based on a discrete grid of a definition domain and a shape function thereof, and a final algebraic system is formed by combining boundary conditions; in the finite element method for area decomposition, the subproblems constructed in each subarea are coupled by boundary unknowns and corresponding boundary conditions, and finally parallel solution is realized by different solving methods.

The additive Schwarz-based regional decomposition method is a universal, efficient and stable regional decomposition method, can adjust the overlapping area of sub-regions aiming at different problems, can realize stable convergence, and introduces the basic principle of the method by combining an electromagnetic field control equation; for a two-dimensional problem, its transient magnetic field control equation based on a magnetic potential function can be expressed as:

wherein ν is a magnetoresistance ratio, U is a magnetic potential, σ is an electric conductivity, t is time, and J is a current density.

The algebraic equation for the complete expansion of the electromagnetic problem obtained after the finite element dispersion of Galerkin can be expressed as:

AU＝F

wherein A is an electromagnetic field rigidity matrix, U is a magnetic potential unknown quantity matrix, and F is an electromagnetic field load matrix.

After the original solution domain is decomposed into N sub solution domains, the potential function vector is correspondingly divided into N sub vectors, meanwhile, a discrete equation generated based on a finite element method can be decomposed according to the sub-regions, and the number of unknown quantities in each sub-region and the whole definition domain is respectively M_iAnd M, so that an algebraic equation for obtaining a complete expansion of the electromagnetic problem after the finite element dispersion by Galerkin can be rewritten as:

wherein the diagonal block matrix (A)_ijI ═ j) characterize the contribution of the degrees of freedom in the sub-region i to the coefficient matrix, the non-diagonal block matrix (a)_ijI ≠ j) characterizes the contribution of the degree of freedom in the sub-region j to the sub-region i coefficient matrix, U_iMatrix of unknowns for sub-region i, F_iIs the load matrix for sub-region i.

Based on the Jacobian method, the iterative equation of the unknowns is expressed as:

this equation is equivalent to:

wherein:

based on:

wherein the superscript n +1/n is the iterative computation step number, R_iFor the restriction operator from the entire domain to sub-region i, M_iMatrix of x M, M_iAnd M is the number of unknown quantities in each sub-region and the whole definition region respectively;

the iterative equation for the unknowns is finally written as:

wherein the content of the first and second substances,

rⁿ＝F-AUⁿ

wherein, the superscript T is the matrix transposition.

In the calculation process, the inverse matrix calculation of the diagonal block matrix corresponding to each sub-region can be realized by using a parallel processor, and in order to improve the solution efficiency of the iterative equation of the unknown quantity and avoid the potential convergence problem, an algebraic system formed after the region decomposition can be solved by combining a Krylov subspace method and a preprocessing operator, and a conjugate gradient method is adopted in the invention.

Since the accurate solution of many degrees of freedom in electromagnetic analysis is unknown, it is essential to perform effective error evaluation on numerical results, and it is also necessary to adjust the dispersion according to the evaluation results to obtain results satisfying the precision and avoid the waste of computing resources; therefore, the grid self-adaptive algorithm is introduced into the additive Schwarz regional decomposition algorithm to ensure the solving precision and improve the calculation efficiency; in order to avoid suspension points possibly caused by independent adjustment of the sub-area grids, the grids of the whole solution domain are uniformly adjusted based on the calculation result after each step of calculation is completed; in order to obtain a grid with better response to magnetic field distribution, the method adopts an anisotropic grid self-adaption method based on the measurement tensor to adjust the grid in the solving process, and other technologies are consistent with a classical self-adaption finite element method.

Wherein, based on the solution result, calculating a Hessian matrix and a metric tensor matrix for describing error information and shape and size characteristics of the triangular grid cells comprises:

hessian matrix H_kExpressed as:

H_k＝{H_k,ij}；i/j＝1,2

wherein, U_kFor the kth iterative solution of the magnetic potential, x₁、x₂Respectively representing independent variables of two directions x and y in a two-dimensional space;

further, H_kCorresponding metric tensor matrix M_kThe method comprises the following steps:

M_k＝V^Tdiag(λ)V

wherein λ and V are each H_kThe eigenvalues and eigenvectors of (c).

If the system error meets the expected value, the numerical result is output after the solving is finished, otherwise, the grid is further processed in a self-adaptive manner and is calculated again; in a metric tensor M determined by the error information_kIn the geometric space of (a), the grid quality is determined by the following formula:

where | Δ | is the area of the cell, p (Δ) is the perimeter of the cell, N is the desired number of grid cells, h (N) is the average size of the grid cells, and F (t) (0 ≦ F ≦ 1) is an arbitrary convex function with an extreme point at t ≦ 1.

Further, based on the measurement tensor, the grid is subjected to self-adaptive processing through the mobile nodes, the edges are exchanged and the nodes are added and deleted, so that the optimal grid subdivision in the geometric space defined by the measurement tensor is generated; after grid self-adaptation, the ratio of numerical solution precision and the number of degrees of freedom is remarkably improved, and then the regional decomposition and solving steps are returned to the solving domain for re-calculation until the error meets the preset value and a numerical result is output; for transient problems, a step calculation step is added on the basis of the circulation, and the step calculation step can be combined with a self-adaptive time step adjustment method to finish calculation by using as few time steps as possible on the premise of meeting the precision requirement.

The method introduces an anisotropic grid self-adaption method based on measurement tensor into an additive Schwarz region decomposition method, combines the advantages of an h-type self-adaption finite element method and the advantages of a region decomposition method, and generates a grid with better response to field distribution in the solving process; by adopting an anisotropic grid self-adaption strategy based on the metric tensor, the grid quality can be judged based on the numerical solution of the current grid, and the subsequent anisotropic adjustment is carried out to obtain an optimal discrete system; multiple sub-region division in the calculation process is realized through METIS, and the solution tasks are distributed approximately evenly according to the number of the sub-regions and the current grids; in order to further improve the solving efficiency, a region decomposition method is used as a pretreatment operator of a Krylov subspace method in the algebraic equation solving; solving the electromagnetic problem without prior knowledge by using a finite element, roughly and uniformly subdividing the whole solution domain, and then carrying out self-adaptive adjustment on the discrete grid according to the evaluation result of error analysis; the method provided by the invention follows the same logic, and is solved by a region decomposition finite element method after self-adaptive dispersion; in order to simplify the calculation steps and avoid suspension points (mismatching grids among sub-regions) possibly caused by scattered grid operation, error evaluation after the regional decomposition and self-adaptive adjustment of the scattered grids are carried out on the global level; therefore, the invention provides a self-adaptive region decomposition finite element method which combines the advantages of the self-adaptive method and the region decomposition method, and the calculation efficiency can be greatly improved on the basis of ensuring the solving precision; a flexible additive Schwarz area decomposition method is used as a preprocessing operator of a Krylov subspace method, an anisotropic grid self-adaption method based on a measurement tensor is introduced, and an optimal grid can be generated according to magnetic field distribution after grid quality is analyzed; compared with the traditional region decomposition method, the method disclosed by the invention has the advantages that the number of required degrees of freedom is reduced while the calculation precision is ensured, and the solving efficiency is obviously improved.

Example 2

Referring to fig. 2 to 8, another embodiment of the present invention is shown, in order to verify and explain the technical effects adopted in the method, the embodiment adopts the conventional technical scheme and the method of the present invention to perform a comparison test, and compares the test results by means of scientific demonstration to verify the real effects of the method.

As shown in fig. 2, which is a structure and size diagram of a C-type actuator, a light color region is a winding through which 1000 ampere-turn current flows, a solution domain including the actuator is used in a numerical problem, and a set of uniform initial grids including 3076 triangle units and 1617 degrees of freedom is generated in the solution domain, and the problem is solved in subsequent calculations by respectively using a conventional area decomposition method and an adaptive area decomposition method, and the two methods are compared in terms of efficiency and precision.

In the process of solving by using a traditional region decomposition method, the number of degrees of freedom is increased by the size of an integral limiting unit, so that the solution precision is improved, after eight times of integral encryption, the obtained magnetic field energy basically reaches stability, as shown in fig. 3, a finally generated grid contains 33760 triangular units and 17141 degrees of freedom, and due to the adoption of a consistent subdivision principle, the encryption process can encrypt regions with small errors of calculation results at the same time, so that certain waste of calculation resources can be inevitably caused.

Solving the problem using adaptive region decomposition, the magnetic field being solved after four adaptive operations on the basis of the initial gridThe energy is basically equivalent to the result of the area decomposition method, as shown in fig. 3, in the solving process, the adaptive method adjusts the current discrete grid based on the error analysis result, adds nodes in the area with large error and removes nodes in the area with small error, so as to ensure that the result meeting the precision requirement is obtained on the premise of avoiding the waste of computing resources. A specific mesh self-adaptation process and corresponding sub-region division thereof are given in fig. 4, the finally generated mesh has 16787 triangular units and 8446 degrees of freedom, wherein the number of degrees of freedom is reduced by about 50%, and observation of the generated mesh can find that the meshes around the abrupt permeability boundary and the sharp corner are encrypted, and the mesh size of the region with small magnetic field gradient is relatively large; for the electromagnetic problem, the self-adaptive area decomposition method can obtain the result close to the same precision only by about one half of the calculation resource required by the traditional area decomposition method; taking the magnetic potential result obtained by the area decomposition method as a reference value, and adaptively adapting to the L of the magnetic potential obtained by the area decomposition method²Error of 2.4X 10^-3(ii) a To better show the results obtained, the fill equipotential lines of the magnetic potential are given in fig. 5.

FIG. 6 is a two-dimensional model of a surface-mounted permanent magnet synchronous motor with four pairs of poles and twenty-four slots, which is analyzed for no-load operation by using a conventional finite element method for area decomposition and a finite element method for adaptive area decomposition according to the present invention; the discrete mesh and the corresponding area division after the seven times of uniform mesh encryption convergence on the basis of the initial uniform mesh are shown in fig. 7(a), wherein the discrete mesh comprises 11788 triangular units and 5970 degrees of freedom, the solution domain is divided into fifteen sub-solution domains on the basis of the discrete mesh, and the magnetic potential distribution on the basis of the mesh solution is shown in fig. 8 (a); the mesh generated by five times of adaptive adjustment when the adaptive region decomposition finite element method is adopted for solving and the solved result are respectively shown in fig. 7(b) and 8(b), the results obtained by the two methods have good consistency through comparison, and the result obtained by the region decomposition method is used as a reference value to calculate the L of the obtained magnetic potential by the adaptive region decomposition method²Error of 3.1X 10^-3The accuracy of the method provided by the invention is verified; mesh unit number and freedom after final adaptive adjustmentThe degrees are 7512 and 3801 respectively, which are far smaller than the corresponding parameters of the uniform encryption grid, wherein the number of degrees of freedom is reduced by about 36%, and the calculation scale is remarkably reduced.

It should be noted that the above-mentioned embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.