A multi-physical field coupling modeling method for a double-stator synchronous generator

Abstract

This application relates to the field of dynamic modeling technology, specifically to a multiphysics coupling modeling method for a dual-stator synchronous generator. The method includes: simulating and modeling the multiphysics of the dual-stator synchronous generator and discretizing the mesh; analyzing the magnetic pull force, simulated temperature, and simulated stress per unit area for each discretized element; determining the distribution differences of magnetic pull force, simulated temperature, and simulated stress among the elements to obtain the spatial distribution deviation and multiphysics difference of each element, thereby obtaining the distribution anomaly degree of each element; calculating the synergistic influence degree of each element to obtain the anomaly significance of each element; constructing four-dimensional feature points for each element using the anomaly significance, magnetic pull force per unit area, simulated temperature, and simulated stress; and obtaining characteristic outlier elements in the finite element physical model based on these features, thereby achieving structural optimization of the dual-stator synchronous generator. This application can improve modeling accuracy and optimize the structure of the dual-stator synchronous generator.

Description

Technical Field

[0001] This application relates to the field of dynamic modeling technology, specifically to a multiphysics coupling modeling method for a dual-stator synchronous generator. Background Technology

[0002] Currently, to improve the power generation performance of dual-stator synchronous generators, it is necessary to comprehensively consider the coupling between multiple physical fields for modeling and analysis, thereby providing technical reference for the structural optimization of dual-stator synchronous generators. Existing technologies generally take the dual-stator synchronous generator as the object, establish a three-dimensional geometric model, and perform multi-physics coupling simulation, mainly including electromagnetic field simulation, temperature field simulation, and mechanical structure response simulation. Then, the characteristic outlier units of the dual-stator synchronous generator under the multi-physics coupling effect during the simulation are analyzed and identified, providing technical reference for the structural optimization of dual-stator synchronous generators.

[0003] Existing technologies often only consider the data characteristics of a single physical field to analyze and identify the outlier elements of a dual-stator synchronous generator, without fully considering the synergistic effect between electromagnetic field, temperature field, and mechanical structure response simulation. This results in poor accuracy in identifying the outlier elements of the dual-stator synchronous generator during the simulation process, thus affecting the subsequent optimization design of the dual-stator synchronous generator structure. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a multiphysics coupling modeling method for dual-stator synchronous generators, thereby resolving the existing issues.

[0005] The multiphysics coupling modeling method for a dual-stator synchronous generator proposed in this application adopts the following technical solution: One embodiment of this application provides a multiphysics coupling modeling method for a dual-stator synchronous generator, including the following steps: By simulating and modeling the multiphysics field of a dual-stator synchronous generator and discretizing the mesh, the magnetic pull force per unit area, simulated temperature, and simulated stress of each discretized element were obtained. The degree of difference in the distribution of magnetic pull force per unit area, simulated temperature and simulated stress among the units is analyzed to obtain the spatial distribution deviation and multi-physics field difference of each unit, and then the distribution anomaly of each unit is obtained. By utilizing the synergistic correlation characteristics between the distribution anomaly degree of each unit and magnetic pull, simulated temperature, and simulated stress, the synergistic influence degree of each unit is calculated, and then the anomaly significance of each unit is obtained by combining the distribution anomaly degree. By constructing four-dimensional feature points for each element based on anomaly saliency, magnetic pull per unit area, simulated temperature, and simulated stress, and identifying anomalous four-dimensional feature points, characteristic outlier elements in the finite element physical model are obtained, thereby achieving structural optimization of the dual-stator synchronous generator.

[0006] Preferably, the normalized magnetic pull, simulated temperature and simulated stress of each unit are used to form the three-dimensional feature points of each unit, and clusters are obtained by clustering. For any unit, other units that belong to the same cluster as the unit are regarded as similar units of the unit.

[0007] Preferably, the process of obtaining the spatial distribution deviation of each unit includes: statistically analyzing the center position of each unit after grid discretization, calculating the coefficient of variation of the distance between the center position of each unit and the center positions of all similar units, calculating the ratio of the maximum distance to the minimum distance between the center position of each unit and the center positions of all similar units, and using the product of the coefficient of variation and the ratio as the spatial distribution deviation of each unit.

[0008] Preferably, the sum of the distances between the three-dimensional feature points of each unit and all the three-dimensional feature points in its respective cluster is used as the multiphysics difference degree of each unit.

[0009] Preferably, the distribution anomaly degree is positively correlated with the spatial distribution deviation degree and the multiphysics field difference degree.

[0010] Preferably, the units closest to the center of each unit are taken as the nearest neighbor units of each unit. The normalized magnetic pull force, simulated temperature, simulated stress and distribution anomaly of each nearest neighbor unit are arranged in ascending order of the distance between the center of each unit and the center of the nearest neighbor unit, respectively, to obtain the magnetic pull force sequence, temperature sequence, stress sequence and distribution anomaly sequence of each unit.

[0011] Preferably, the gray-scale correlation analysis method is used to statistically analyze the correlation coefficient sequences between the distribution anomaly sequence and the magnetic tension sequence, temperature sequence, and stress sequence, respectively. The synergistic influence degree is obtained by the average level of the three correlation coefficient sequences and the degree of difference between the three correlation coefficient sequences.

[0012] Preferably, the relationship for obtaining the synergistic influence degree of each unit is as follows: In the formula, Let i be the degree of synergistic influence of the i-th unit. Let be the mean of the data within the three correlation coefficient sequences corresponding to the i-th unit. Let be the mean of the difference distances between the three correlation coefficient sequences corresponding to the i-th unit. This is a preset division by zero constant.

[0013] Preferably, the relationship for obtaining the anomaly significance of each unit is as follows: In the formula, Let i be the normalized synergy influence of the i-th unit. Let be the significance of the anomaly in the i-th unit. Let be the distribution anomaly degree of the i-th unit.

[0014] Preferably, the isolated forest algorithm is used to identify anomalous four-dimensional feature points, and the units corresponding to the anomalous four-dimensional feature points are used as feature outliers in the finite element physical model after dynamic modeling.

[0015] This application has at least the following beneficial effects: This application constructs three-dimensional feature points based on multi-physics field data after multi-physics field coupling simulation modeling, and accurately analyzes the spatial distribution deviation and multi-physics field difference of each unit based on the distribution characteristics of the three-dimensional feature points. Then, by combining the spatial distribution deviation and multi-physics field difference, the distribution anomaly degree is fused and measured more accurately. Furthermore, this application uses grey relational analysis to extract the characteristics of the correlation coefficient, and then analyzes the synergistic effect of the influence on local parts in the finite element physical model under multi-physics coupling. It fully considers the synergistic effect between electromagnetic field, temperature field and mechanical structure response simulation, and realizes the accurate measurement of the degree of synergistic influence, which is conducive to the accurate identification of characteristic outlier units of dual-stator synchronous generator in the subsequent simulation process. In this application, the normalized synergy influence degree is used to increase the coefficient of the distribution anomaly degree, which more significantly highlights the abnormal characteristics of each element in the finite element physical model. Then, a four-dimensional feature point is constructed by combining the anomaly significance. By using the four-dimensional feature point, the characteristic outlier element of the dual-stator synchronous generator in the simulation process is identified, thereby improving the accuracy of identifying the characteristic outlier element of the dual-stator synchronous generator in the simulation process and avoiding affecting the optimization design of the dual-stator synchronous generator structure. Attached Figure Description

[0016] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating the steps of a multiphysics coupling modeling method for a dual-stator synchronous generator provided in this application. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a multiphysics coupling modeling method for a dual-stator synchronous generator proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, terms such as “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a circuit structure, article, or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the article or device that includes said element. Furthermore, the term “and / or” as used herein includes any and all combinations of one or more of the associated listed items. All technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0020] The following description, in conjunction with the accompanying drawings, details the specific scheme of the multiphysics coupling modeling method for a dual-stator synchronous generator provided in this application.

[0021] This application provides an embodiment of a multiphysics coupling modeling method for a dual-stator synchronous generator. For details, please refer to [link to specific documentation]. Figure 1 This includes the following steps: Step 1: Simulate and model the multiphysics field of the dual-stator synchronous generator and discretize the mesh. Then, simulate the magnetic pull force per unit area, temperature and stress of each discretized element.

[0022] In order to provide technical reference for the optimized design of dual-stator synchronous generators, it is necessary to fully consider the synergistic effect between the simulation of electromagnetic field, temperature field and mechanical structure response of dual-stator synchronous generators during the simulation process, so as to more accurately identify the characteristic outlier units of dual-stator synchronous generators during the simulation process.

[0023] Therefore, this application obtains the main design parameters of a dual-stator synchronous generator through the design scheme of the dual-stator synchronous generator, including rated power, rated voltage, number of poles of external power winding, number of poles of internal power winding, number of poles of internal control winding, inner and outer diameters of external stator, inner and outer diameters of internal stator, inner and outer diameters of rotor, effective shaft length of iron core, number of slots of internal stator, number of slots of external stator, inner stator slot angle, and outer stator slot angle.

[0024] Furthermore, based on the main design parameters of the dual-stator synchronous generator, a dynamic model of the generator was created using CAD software (Computer AidedDesign) to obtain its finite element physical model. The dynamically modeled finite element physical model was then discretized into a mesh (1mm × 1mm) to form a set of finite element elements, providing a three-dimensional geometric model and simulation foundation for subsequent multiphysics simulations of the dual-stator synchronous generator. Next, the dynamically modeled finite element physical model and its set of finite element elements were imported into ANSYS Maxwell 3D software. Electromagnetic field simulations of the dual-stator synchronous generator were performed using ANSYS Maxwell 3D software to solve for the power generation loss and magnetic pull per unit area of ​​each element in the finite element set. ANSYS Maxwell 3D software is an electromagnetic field solver for generators, transformers, actuators, and other electromechanical equipment; it can solve static, frequency domain, and time-varying electromagnetic fields.

[0025] Furthermore, temperature field simulations were performed using ANSYS Steady-State Thermal software. When establishing the thermal model, it was necessary to differentiate the thermal conductivity and specific heat capacity of different materials and define boundary conditions, including defining the convective heat transfer coefficient of the casing exterior, with a value ranging from 20 to 100 W / (m²). 2 ·K), define the heat transfer between the winding end and the air contact surface and the heat conduction between the shaft and the bearing contact surface, and then use the power generation loss of each element in the finite element set as the volume heat source, map it to the corresponding geometric region in the finite element physical model after dynamic modeling, and use ANSYS Steady-State Thermal software to solve for the simulation temperature of each element in the finite element set.

[0026] Then, stress field simulation was performed on the dual-stator synchronous generator. The magnetic pull per unit area and the simulated temperature of each element in the finite element set were imported into ANSYS Mechanical software. The stress field simulation of the dual-stator synchronous generator was performed using ANSYS Mechanical software, and the simulated stress of each element in the finite element set was obtained.

[0027] Thus, based on the above process in this embodiment, the magnetic pull per unit area, simulated temperature, and simulated stress of each element after multiphysics simulation modeling and mesh discretization of the dual-stator synchronous generator can be obtained.

[0028] Step 2: Analyze the degree of difference in the distribution of magnetic pull force per unit area, simulated temperature and simulated stress among the units, obtain the spatial distribution deviation and multi-physics field difference of each unit, and then obtain the distribution anomaly of each unit.

[0029] Because of the synergistic effects among electromagnetic, temperature, and stress fields, existing technologies often only consider the data characteristics of a single physical field to analyze and identify outlier elements in dual-stator synchronous generators. This can easily lead to poor accuracy in identifying outlier elements during simulation, thus affecting the subsequent optimization design of the dual-stator synchronous generator structure. Therefore, it is necessary to fully explore the synergistic effects among electromagnetic, temperature, and stress fields to more accurately identify outlier elements in dual-stator synchronous generators.

[0030] To eliminate the physical dimensions of multiphysics coupling calculations and thus accurately analyze and identify the characteristic outlier elements of a dual-stator synchronous generator, the magnetic pull force, simulated temperature, and simulated stress per unit area of ​​all elements are normalized. The normalization method can be exponential or linear normalization; this embodiment does not specify a particular method. This yields the normalized magnetic pull force, normalized simulated temperature, and normalized simulated stress for each element. Furthermore, based on the normalized magnetic pull force, normalized temperature, and normalized stress of each element, three-dimensional feature points are constructed for each element, i.e., the three-dimensional feature points of the i-th element. , These represent the normalized magnetic pull, normalized simulated temperature, and normalized simulated stress of the i-th unit, respectively.

[0031] Furthermore, in order to analyze the distribution characteristics of multiphysics data, the three-dimensional feature points of all units are used as input to the density peaks clustering algorithm. The truncation distance in the algorithm is chosen so that the average number of neighbors accounts for 2% of the total sample. The density peaks clustering algorithm is used to obtain each cluster.

[0032] Then, each unit corresponding to every other 3D feature point within the cluster to which the 3D feature point of each unit belongs is taken as a similar unit of each unit. The center position of each unit after mesh discretization is obtained in CAD software, where the center position of each unit is the coordinate information of the center point of each unit after mesh discretization. The coefficient of variation of the distance between the center position of each unit and the center positions of all its similar units is calculated, and the ratio of the maximum distance to the minimum distance between the center position of each unit and the center positions of all its similar units is calculated, where the minimum distance is not equal to 0. The product of the coefficient of variation and the Euclidean distance ratio is recorded as the spatial distribution deviation of each unit. It should be noted that Euclidean distance is used in this embodiment to measure the distance between the aforementioned center positions. In actual application scenarios, implementers can also choose other distance calculation methods in the prior art to measure the distance between center positions.

[0033] Among them, the spatial distribution deviation is used to reflect the discrete distribution characteristics of the Euclidean distance between each unit and other similar units in a multiphysics field. The greater the spatial distribution deviation, the more significant the discrete distribution characteristics of the Euclidean distance between the unit and other similar units in a multiphysics field. In this case, the distribution of the center positions of all similar units of the unit is more abnormal and not normally distributed in the local area of ​​the unit.

[0034] Simultaneously, the sum of the Euclidean distances between the three-dimensional feature points of each unit and all three-dimensional feature points within its cluster is calculated and denoted as the multiphysics dissimilarity of each unit. This reflects the difference in three-dimensional features between each unit and other similar units in the multiphysics field. The greater the multiphysics dissimilarity, the smaller the similarity of the three-dimensional features of the unit with other similar units in the multiphysics field, thus significantly demonstrating the anomalous outlier nature of the unit in the multiphysics field data within the local region.

[0035] In general, the multiphysics characteristics of all elements in a local region of a dynamically modeled finite element physical model are quite similar, and the similar elements of each element are normally distributed in the local region of that element. However, if there are outlier elements in the local region, the Euclidean distance between the outlier elements and their similar elements will show an abnormal discrete distribution, and there will be significant differences in the multiphysics characteristics between the outlier elements and their similar elements.

[0036] Based on the above analysis, the distribution anomaly degree of each unit is calculated by combining the spatial distribution deviation and multiphysics difference degree of each unit. The distribution anomaly degree is positively correlated with both the spatial distribution deviation and the multiphysics difference degree. In this embodiment, the positive correlation means that the two variables have the same trend of change; one variable increases as the other increases, and decreases as the other decreases. Specifically, in this embodiment, the distribution anomaly degree of the i-th unit... One calculation formula is: In the formula, The spatial distribution deviation of the i-th unit. Let be the multiphysics difference degree of the i-th unit. and The first and second preset weights are respectively set to 0.5 and 0.5, respectively, in this embodiment.

[0037] Another formula for calculating the distribution anomaly degree can be: , To avoid the product being zero, the value range is [0.1, 0.2], and in this embodiment, the value is 0.2.

[0038] In existing methods for measuring anomalies, feature fusion can be performed using either a weighted summation or a product approach based on two different anomaly factors to achieve a fusion measurement of anomalies. In this invention, spatial distribution deviation and multiphysics difference are two distinct anomaly factors. The first calculation formula uses two identical weights to perform a weighted summation of spatial distribution deviation and multiphysics difference to achieve a fusion measurement of distribution anomaly. The second calculation formula uses a product approach to fuse spatial distribution deviation and multiphysics difference, thereby achieving a fusion measurement of distribution anomaly.

[0039] Based on the above process, it can be understood that the distribution anomaly reflects the abnormal characteristics of the spatial location distribution and multi-physics data distribution between each element and its similar elements. The greater the distribution anomaly, the more significant the abnormal characteristics of the spatial location distribution and multi-physics data distribution between the element and its similar elements. In this case, the element is more likely to be a characteristic outlier element in the finite element physical model after dynamic modeling.

[0040] Step 3: Utilize the synergistic correlation characteristics between the distribution anomaly degree of each unit and magnetic pull, simulated temperature, and simulated stress to calculate the synergistic influence degree of each unit, and then combine the distribution anomaly degree to obtain the anomaly significance of each unit.

[0041] To more fully consider the synergistic effect between the simulation of electromagnetic field, temperature field, and mechanical structure response, the M nearest neighbors (M elements) to the center of each element are selected as the M nearest neighbors in this embodiment (M is set to 30). Furthermore, the normalized magnetic pull, simulated temperature, simulated stress, and distribution anomaly of each nearest neighbor are arranged in ascending order of distance between the center of each element and the center of its nearest neighbors, resulting in the magnetic pull sequence, temperature sequence, stress sequence, and distribution anomaly sequence for each element. This reflects the changing characteristics of multiphysics data and distribution anomaly in the nearest neighbor region of each element in the finite element physical model after dynamic modeling.

[0042] Furthermore, to analyze the synergistic effect among electromagnetic field, temperature field, and mechanical structure response simulations, the magnetic pull force sequence, temperature sequence, stress sequence, and distribution anomaly sequence of each unit are used as inputs for Grey Relational Analysis. The resolution coefficient is set to 0.5. The distribution anomaly sequence represents the parent sequence, and the magnetic pull force sequence, temperature sequence, and stress sequence represent the three sub-sequences. The Grey Relational Analysis method is used to calculate the correlation coefficient between each sub-sequence and the parent sequence at each data point, resulting in a correlation coefficient sequence for each sub-sequence of each unit. Each unit corresponds to three correlation coefficient sequences, which are used to reflect the variation characteristics of the correlation coefficient between each sub-sequence and the parent sequence across all data points.

[0043] Generally, after dynamic modeling, there is a strong correlation between the multiphysics data and the distribution anomaly degree in the neighborhood of each element in the finite element physical model. The more similar the correlation characteristics between the multiphysics data and the distribution anomaly degree, the greater the synergistic effect between the simulation of electromagnetic field, temperature field and mechanical structure response, and the greater the possibility of anomalies in the structural units of the dual-stator synchronous generator.

[0044] Based on the above analysis, and according to the correlation between the distribution anomaly sequence and the magnetic pull sequence, temperature sequence, and stress sequence, the synergistic influence degree of each unit is calculated. The synergistic influence degree of the i-th unit is... The calculation formula is: In the formula, Let i be the degree of synergistic influence of the i-th unit. Let be the mean of the data within the three correlation coefficient sequences corresponding to the i-th unit. Let be the mean of the difference distances between the three correlation coefficient sequences corresponding to the i-th unit. The difference distance can be measured by DTW distance or Mahalanobis distance; in this embodiment, DTW distance is used. To prevent the denominator from being 0, a pre-defined division constant is used, with a value range of [0.05, 0.1]. In this embodiment, the value is 0.06.

[0045] When measuring features with positive and negative proportional relationships based on existing measurement methods, a fractional approach can be used to measure the feature. In this approach, the mean of the correlation coefficient in the numerator is directly proportional to the measurement index, while the mean of the difference distance in the denominator is inversely proportional to the measurement index, thus achieving a feature measurement of the synergistic influence.

[0046] Understandably, the degree of synergistic influence reflects the synergistic effect of the influence on local parts of the finite element physical model under multi-physics coupling. The greater the degree of synergistic influence, the stronger the synergistic effect of the influence on local parts of the finite element physical model under multi-physics coupling, and the greater the possibility of causing abnormalities in the structural units of the dual-stator synchronous generator.

[0047] Therefore, to more accurately identify the outlier elements of the dual-stator synchronous generator during simulation, the synergistic influence of each element in the finite element set is normalized. This normalization can be achieved using exponential or linear normalization methods to obtain the normalized synergistic influence of each element. Furthermore, based on the normalized synergistic influence and the distribution anomaly, the anomalous significance of the i-th element is calculated. : In the formula, Let be the normalized synergy influence of the i-th unit.

[0048] The formula for calculating anomaly significance uses a normalized synergistic influence factor to gain the coefficient 1 of the distribution anomaly degree, thus more significantly highlighting the anomalous characteristics of each element in the dynamically modeled finite element physical model. Anomaly significance reflects the degree of significance of the anomalous characteristics of each element in the dynamically modeled finite element physical model. The higher the outlier the anomaly significance in the dynamically modeled finite element physical model, the higher the probability that the corresponding element is a characteristic outlier element in the finite element physical model.

[0049] Step 4: Construct four-dimensional feature points for each element by using anomaly saliency, magnetic pull per unit area, simulated temperature, and simulated stress. Identify anomalous four-dimensional feature points to obtain characteristic outlier elements in the finite element physical model, thereby achieving structural optimization of the dual-stator synchronous generator.

[0050] Furthermore, to improve the accuracy of identifying outlier elements of a dual-stator synchronous generator during simulation, a method based on the three-dimensional feature points of each element in the finite element set is proposed. and the significance of anomalies in each unit Construct four-dimensional feature points for each unit. Then, the iForest Isolation Forest method is used to identify the outlier elements of the dual-stator synchronous generator. The four-dimensional feature points of all elements in the finite element set are used as the input of the iForest Isolation Forest method. In this embodiment, the preset number of isolation trees is 120, the preset anomaly ratio is 0.01, and the other preset parameters are taken as the default values ​​of the algorithm. The iForest Isolation Forest method is used to identify the abnormal four-dimensional feature points. The elements corresponding to the abnormal four-dimensional feature points are recorded as the outlier elements in the finite element physical model after dynamic modeling, thereby providing a theoretical basis for the structural optimization of the dual-stator synchronous generator.

[0051] The outlier elements in the finite element physical model exhibit a high degree of abnormality in the changes of multiphysics data during multiphysics simulation. This indicates that the outlier elements are more prone to deformation during multiphysics simulation. Therefore, it is necessary to optimize the structure of the dual-stator synchronous generator to ensure the safety and reliability of the dual-stator synchronous generator in actual operation.

[0052] It is understood that references to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application include the specific features, structures, or characteristics described in connection with that embodiment. Therefore, the appearance of phrases such as "in one embodiment," "in some embodiments," "in other embodiments," or "in still other embodiments" in different parts of this specification does not necessarily refer to the same embodiment, but rather means "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0053] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous. Moreover, the sequence numbers of the steps in the embodiments do not imply a specific order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments in this specification.

[0054] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A multiphysics coupling modeling method for a dual-stator synchronous generator, characterized in that, Includes the following steps: By simulating and modeling the multiphysics field of a dual-stator synchronous generator and discretizing the mesh, the magnetic pull force per unit area, simulated temperature, and simulated stress of each discretized element were obtained. The degree of difference in the distribution of magnetic pull force per unit area, simulated temperature and simulated stress among the units is analyzed to obtain the spatial distribution deviation and multi-physics field difference of each unit, and then the distribution anomaly of each unit is obtained. By utilizing the synergistic correlation characteristics between the distribution anomaly degree of each unit and magnetic pull, simulated temperature, and simulated stress, the synergistic influence degree of each unit is calculated, and then the anomaly significance of each unit is obtained by combining the distribution anomaly degree. By constructing four-dimensional feature points for each element based on anomaly saliency, magnetic pull per unit area, simulated temperature, and simulated stress, and identifying anomalous four-dimensional feature points, characteristic outlier elements in the finite element physical model are obtained, thereby achieving structural optimization of the dual-stator synchronous generator.

2. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 1, characterized in that, The normalized magnetic pull, simulated temperature, and simulated stress of each unit are used to form the three-dimensional feature points of each unit. Clustering is then used to obtain each cluster. For any unit, other units that belong to the same cluster as the given unit are considered as similar units of the given unit.

3. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 2, characterized in that, The process of obtaining the spatial distribution deviation of each unit includes: statistically analyzing the center position of each unit after grid discretization, calculating the coefficient of variation of the distance between the center position of each unit and the center positions of all similar units, calculating the ratio of the maximum distance to the minimum distance between the center position of each unit and the center positions of all similar units, and taking the product of the coefficient of variation and the ratio as the spatial distribution deviation of each unit.

4. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 2, characterized in that, The sum of the distances between the three-dimensional feature points of each unit and all the three-dimensional feature points in its respective cluster is used as the multiphysics difference degree of each unit.

5. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 1, characterized in that, The distribution anomaly degree is positively correlated with the spatial distribution deviation degree and the multiphysics field difference degree.

6. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 3, characterized in that, The units closest to the center of each unit are taken as the nearest neighbor units of each unit. The normalized magnetic pull force, simulated temperature, simulated stress and distribution anomaly of each nearest neighbor unit are arranged in ascending order of the distance between the center of each unit and the center of the nearest neighbor unit, respectively, to obtain the magnetic pull force sequence, temperature sequence, stress sequence and distribution anomaly sequence of each unit.

7. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 6, characterized in that, The gray-scale correlation analysis method was used to statistically analyze the correlation coefficients between the distribution anomaly sequence and the magnetic tension sequence, temperature sequence, and stress sequence, respectively. The synergistic influence was obtained by the average level of the three correlation coefficient sequences and the degree of difference between the three correlation coefficient sequences.

8. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 7, characterized in that, The formula for obtaining the synergistic influence of each unit is as follows: In the formula, Let i be the degree of synergistic influence of the i-th unit. Let be the mean of the data within the three correlation coefficient sequences corresponding to the i-th unit. Let be the mean of the difference distances between the three correlation coefficient sequences corresponding to the i-th unit. This is a preset division by zero constant.

9. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 1, characterized in that, The specific formula for obtaining the anomaly significance of each unit is as follows: In the formula, Let i be the normalized synergy influence of the i-th unit. Let be the significance of the anomaly in the i-th unit. Let be the distribution anomaly degree of the i-th unit.

10. The multiphysics coupling modeling method for a dual-stator synchronous generator as described in claim 1, characterized in that, The isolated forest algorithm is used to identify anomalous four-dimensional feature points, and the elements corresponding to the anomalous four-dimensional feature points are used as feature outliers in the finite element physical model after dynamic modeling.

Images