Information processing system and program

Abstract

To provide an information processing system and a program capable of performing information processing on a structure in an industrially advantageous manner.SOLUTION: The information processing system uses a vector including position information and constituent material information and calculates the characteristic of a structure by using at least three pieces of the positional information. The information processing system functions as means for learning the relationship of the characteristic in relation to the vector including the position information and the constituent material information, predicting the characteristic from the vector received from a user by the learning, and outputting the predicted evaluation value to the user.SELECTED DRAWING: Figure 11

Description

[Technical Field]

[0001] The present invention relates to an information processing system and a computer program that are useful for designing structures such as motors. [Background technology]

[0002] Non-Patent Document 1 discloses an automated design system for interior permanent magnet synchronous motors. This automated design system includes a generative model that generates an image showing the shape of the motor rotor, and a predictive model that predicts motor characteristics from the generated image.

[0003] The automated design system of Non-Patent Document 1 includes a generative model 310 that generates an image 400 showing the shape of a motor, and a characteristic prediction model 320 that predicts motor characteristics 450 from the generated image 400. This automated design system can automatically design an optimized motor structure (optimal solution) by repeatedly generating a shape and predicting (evaluating) the characteristics of the generated shape.

[0004] However, in the automatic design system of Non-Patent Document 1, the generative model may generate an unmanufacturable shape (a collapsed shape) as shown in FIG. 10(B), and this shape may be selected as the optimal solution.

[0005] Therefore, when automatically designing a structure such as a motor, it is desirable to prevent the generation of an inappropriate shape such as a collapsed shape.

[0006] Since images have too much freedom of expression, the generative model may generate shapes that cannot be manufactured. Therefore, by adopting a representation format with reduced freedom of expression, it is possible to prevent the generation of inappropriate shapes. [Prior art documents] [Non-patent literature]

[0007] [Non-Patent Document 1] Yusei Shimizu, "Robust Optimization of IPMSM Efficiency Using Deep Generative Models", IEEJ Study Group Materials.SA, IEEJ Static Machine Study Group [ed.], 2022 (80-100), 7-12, 2022-09-30 [Non-patent document 2] Yusei Shimizu, "Technical Commentary by a Motor Researcher," [online], [Retrieved August 7, 2023], Internet<URL:https: / / yuyumoyuyu.com / 2023 / 01 / 15 / paperintroduction2 / > Summary of the Invention [Problem to be solved by the invention]

[0008] An object of the present invention is to provide an information processing system and program that can perform information processing relating to structures and the like in an industrially advantageous manner. [Means for solving the problem]

[0009] As a result of intensive research to achieve the above-mentioned object, the inventors have discovered that an information processing system using vectors, wherein the vectors include positional information and constituent material information, and which uses three or more pieces of positional information, is useful for evaluating the characteristics of structures, etc., and in particular can easily represent the shape of a structure, thereby enabling significantly better information processing of structures, and have also discovered that such an information processing system can solve the above-mentioned conventional problems in one fell swoop. Furthermore, after obtaining the above findings, the present inventors conducted further studies and completed the present invention.

[0010] That is, the present invention relates to the following inventions. [1] An information processing system that uses vectors, wherein the vectors include position information and constituent material information, and the information processing system uses three or more pieces of position information. [2] The information processing system according to [1], wherein the vector information is used to design a structure. [3] The information processing system according to [2], wherein the shape of the structure is expressed using the vector information. [4] The information processing system according to [3], wherein the position information includes information on the x-axis and y-axis. [5] The information processing system according to [3], wherein four or more pieces of location information are used. [6] The information processing system according to [5], wherein the position information further includes z-axis information. [7] The information processing system according to [1], wherein the constituent material information includes air. [8] The information processing system according to [1], wherein the constituent material information includes a magnetic material. [9] The information processing system according to [1], wherein the constituent material information includes physical property information.

[10] The information processing system according to [1], further comprising an evaluation means for evaluating the characteristics of a structure from information on the vector when the vector is received from a user.

[11] An information processing system according to

[10] , which learns the relationship between the vector and the characteristics of the structure, predicts the characteristics from the user's vector through the learning, and outputs the predicted evaluation value to the user.

[12] A program that causes an information processing system for calculating the characteristics of a structure to function as a means for learning the relationship between said characteristics by associating said characteristics with vectors including position information and constituent material information, predicting said characteristics from said vectors received from a user through said learning, and outputting the predicted evaluation value to said user. [Effects of the Invention]

[0011] The information processing system and program of the present invention can perform information processing relating to structures and the like in an industrially advantageous manner. [Brief explanation of the drawings]

[0012] [Figure 1] Figure 1 is a diagram of the design system. [Figure 2] FIG. 2 is a perspective view of the motor. [Figure 3] FIG. 3 is a diagram showing the topology of one pole of the rotor. [Figure 4] FIG. 4 is a flowchart of the graph data generation process. [Figure 5(A)] FIG. 5(A) is an explanatory diagram of the graph data. [Figure 5(B)] FIG. 5(B) is an explanatory diagram of the graph data. [Figure 5(C)] FIG. 5C is an explanatory diagram of the graph data. [Figure 5(D)] FIG. 5(D) is an explanatory diagram of the graph data. [Figure 6(A)] FIG. 6(A) is an explanatory diagram of the graph data. [Figure 6(B)] FIG. 6B is an explanatory diagram of the graph data. [Figure 7] FIG. 7 is an explanatory diagram of the graph data. [Figure 8(A)] FIG. 8A is a diagram showing variations in the generation of coordinates. [Figure 8(B)] FIG. 8B is a diagram showing variations in the generation of coordinates. [Figure 8(C)] FIG. 8C is a diagram showing a variation of the generation of coordinates. [Figure 9(A)] FIG. 9(A) is a diagram showing variations of the evaluator. [Figure 9(B)] FIG. 9B shows a variation of the evaluator. [Figure 10(A)] FIG. 10(A) is a diagram showing an example of a design system that generates an image and an image of a collapsed shape. [Figure 10(B)] FIG. 10(B) is a diagram showing an example of a design system that generates an image and an image of a collapsed shape. [Figure 11] FIG. 11 is a diagram illustrating feature vectors focusing on the vertices of a structure. [Figure 12(A)] FIG. 12A is a diagram illustrating a feature vector focusing on a structure region. [Figure 12(B)] FIG. 12B is a diagram illustrating a feature vector focusing on a structure region. DETAILED DESCRIPTION OF THE INVENTION

[0013] The information processing system of the present invention uses vectors, characterized in that the vectors include position information and constituent material information, and three or more pieces of position information are used.

[0014] The vector used in the present invention is not particularly limited as long as it contains position information and constituent materials. It may be a known vector. Examples of the vector include a vector containing position coordinates and constituent material symbols. Examples of the position coordinates include known coordinates, more specifically, polar coordinates, cylindrical coordinates, or Cartesian coordinates. The constituent material symbols are not particularly limited as long as they represent the constituent materials. Examples of the constituent material symbols include one or a combination of two or more symbols selected from numbers, letters such as the alphabet, and other symbols. In the present invention, it is preferable to design a structure using the vector information. According to this preferred embodiment, the structure can be designed more efficiently with minimal information. Examples of such structures include motors, piezoelectric actuators, RFID (Radio Frequency Identification), cables, high-frequency preheaters, capacitors, magnetic heads, magnetic shields, magnetic gears, magnets, circuit breakers, sensors, electrical heating devices, electrolytic plating, electromagnetic forming, electromagnetic retarders, electromagnetic relays, transformers / reactors, bus bars, generators, heaters, bearings, magnetic couplings, induction heating devices, induction machines, linear solenoids / linear actuators, and wireless power supply systems.

[0015] In the present invention, it is also preferable to represent the shape of the structure using the vector information. According to this preferred embodiment, the shape of the structure can be used more efficiently in information processing. As the shape of the structure, for example, various known shapes may be used. Specifically, shapes that are approximately circular, elliptical, polygonal, or formed by various straight lines and curves may be used.

[0016] In the present invention, the position information preferably includes x-axis and y-axis information. According to this preferred embodiment, the shape of the structure can be more accurately utilized in information processing. Examples of the x-axis and y-axis information include known position coordinates including the x-axis and y-axis.

[0017] In the present invention, it is preferable to use four or more pieces of position information. According to such a preferred embodiment, the three-dimensional design information of the structure can be used in more detail for information processing.

[0018] In the present invention, it is also preferable that the position information further includes z-axis information. According to this preferred embodiment, the three-dimensional design information of the structure can be used in more detail for information processing. Examples of the z-axis information include known position coordinates including the z-axis.

[0019] In the present invention, it is also preferable that the constituent material information includes air. According to such a preferred embodiment, information processing including the space of the structure can be more easily performed. Examples of the constituent material information include material information symbols.

[0020] In the present invention, it is also preferable that the constituent material information includes a magnetic material. According to such a preferred embodiment, information processing including the properties of the material of the structure can be more easily performed. Examples of the constituent material information include material information symbols.

[0021] In the present invention, the constituent material information preferably includes physical property information. According to such a preferred embodiment, more detailed information processing including characteristic evaluation of the structure can be performed. Examples of the physical property information include magnetic permeability, saturation magnetic flux density, residual magnetic flux density, magnetization, coercive force, loss factor, dielectric constant, resistivity, and conductivity.

[0022] In addition, the present invention preferably includes an evaluation means for evaluating the characteristics of a structure based on information on the vectors when the vectors are received from a user. According to this preferred embodiment, it is possible to evaluate particularly nonlinear characteristics of the structure in more detail. Examples of the evaluation means include finite element analysis.

[0023] In addition, in the present invention, it is preferable to learn the relationship between the vector and the characteristics of the structure, predict the characteristics from the user's vector through the learning, and output the predicted evaluation value to the user. According to this preferred embodiment, the characteristics of the structure can be evaluated in a shorter time. Examples of the learning and prediction means include machine learning methods such as support vector regression, Gaussian process regression, random forest, gradient boosting decision tree, neural network, multilayer perceptron, convolutional neural network, Transformer, Vision Transformer, MLP-Mixer, and graph neural network.

[0024] The present invention will be specifically described below with reference to the drawings, but the present invention is not limited to these specific examples.

[0025] 1 shows a design system 100 according to a preferred embodiment. The design system 100 designs the shape of a structure. The shape to be designed is, for example, a physical two-dimensional or three-dimensional shape of a tangible structure. The shape may be the shape of a part of the structure or the shape of the entire structure.

[0026] As an example, the structure may be an electric machine. Here, electric machines are a general term for rotating electric machines (rotating machines) such as motors and generators, and stationary electric machines (stationary machines) such as transformers, rectifiers, and switching control devices. Electric machines are machines whose electromagnetic properties can be evaluated in the design stage.

[0027] The structure may, for example, comprise a magnet, for example an electric machine such as a motor.

[0028] The design system 100 according to the embodiment includes a generator 110 and an evaluator 120. The generator 110 generates a shape of a structure. The evaluator 120 evaluates characteristics of the structure based on the shape generated by the generator 110.

[0029] The design system 100 may be configured by one or more computers having a processor and a memory. The processor may be a CPU or a GPU. The memory is connected to the processor. The memory may include, for example, a primary storage device and a secondary storage device. The primary storage device may be, for example, a RAM. The secondary storage device may be, for example, a hard disk drive (HDD) or a solid-state drive (SSD). The memory may include a computer program executed by the processor. The processor reads and executes the computer program stored in the memory. The computer program has program code for causing the computer to operate as the design system 100. That is, the computer program has program code for causing the computer to operate as the generator 110 and the evaluator 120.

[0030] The design system 100 can output an optimized shape by repeatedly generating and evaluating a shape. For example, the design system 100 uses the determiner 150 to determine whether the characteristics evaluated (predicted) by the evaluator 120 satisfy predetermined requirements, and if the requirements are satisfied, outputs a shape having those characteristics as an optimized structure. If the characteristics evaluated by the evaluator 120 do not satisfy the predetermined requirements, the design system 100 regenerates and evaluates the shape. The design system 100 automatically outputs an optimized shape by repeatedly generating and evaluating various shapes.

[0031] The characteristics evaluated by the evaluator 120 are characteristics of the structure having the shape generated by the generator 110. The characteristics may be, for example, characteristics of a rotor or a stationary motor. The characteristics may be, for example, electromagnetic characteristics. If the structure is a motor, the characteristics may be motor characteristics. The motor characteristics may include, for example, any one or a combination of the speed-torque characteristics of the motor, the efficiency of the motor, torque ripple, and iron loss.

[0032] In the following description, the design system 100 according to the embodiment is used to design a motor, as an example. Motors are used in various products such as electric vehicles, drones, aircraft, and industrial robots. The motor to be designed by the design system 100 is, as an example, an interior permanent magnet synchronous motor (IPMSM).

[0033] As shown in Fig. 2, interior permanent magnet synchronous motor 10 is a type of motor that uses permanent magnets 15. Interior permanent magnet synchronous motor 10 includes stator 11 and rotor 12 provided inside stator 11. Permanent magnets 15 are embedded in rotor 12. Interior permanent magnet synchronous motor 10 shown in Fig. 2 is, as an example, an 8-pole, 48-slot distributed winding IPMSM.

[0034] The stator 11 is made of electromagnetic steel and has multiple slots 16 formed in the circumferential direction. Air exists in the slots 16. Therefore, when designing the stator 11, two types of materials must be taken into consideration: electromagnetic steel and air.

[0035] Rotor 12 is constructed by embedding permanent magnets 15 in a rotor body made of electromagnetic steel. In rotor 12, regions (holes) where air exists are formed in positions adjacent to permanent magnets 15. Therefore, when designing rotor 12, three types of materials must be considered: electromagnetic steel, permanent magnets, and air.

[0036] Here, the object of design is rotor 12, and the structure of stator 11 (stator structure) is assumed to be predetermined. The external shape of rotor 12 is also assumed to be predetermined. Therefore, the object of design here is the shape and arrangement of the magnets and air in rotor 12. Note that the external shapes of stator 11 and rotor 12 may also be subject to design.

[0037] Furthermore, since each pole of the rotor 12 has the same structure, when designing the rotor 12, it is sufficient to design one pole of the rotor 12.

[0038] 3(A) to 3(C) show examples of two-dimensional topology for one pole of the rotor 12. The topologies shown in Fig. 3(A) to 3(C) differ in the position, number, shape, etc. of the permanent magnets and holes (air).

[0039] In the design system 100 according to the embodiment, the generator 110 generates and outputs graph data 200 that indicates the two-dimensional shape of one pole of the rotor 12 shown in Fig. 3. The evaluator 120 acquires the graph data 200 output from the generator 110, and evaluates the motor characteristics of a motor (IPMSM) that includes the rotor 12 having the generated shape, based on the graph data 200.

[0040] FIG. 4 shows an example of the procedure of the process S400 for generating the shape of a structure, which is executed by the generator 110. As described above, the generator 110 according to the embodiment generates the shape of the rotor 12 as graph data 200, which is represented by a graph. A graph includes nodes (vertices) and edges (sides). Although a graph has less freedom of expression than an image, it still has sufficient freedom of expression to represent a two-dimensional or three-dimensional shape of a structure. Therefore, a graph is suitable as a representation format for the shape to be generated, and can prevent the generation of an inappropriate shape, such as a collapsed shape.

[0041] As shown in Figures 5(C) and 5(D), graph data 200 can be represented by a matrix representing a graph (see Figure 5(C)) and a feature vector (see Figure 5(D)). In the present invention, three or more such feature vectors are used as the vector. In this way, by using three or more of the vectors, a surface can be constructed, and the characteristics of the structure can be applied appropriately and efficiently to information processing.

[0042] A matrix that represents a graph is, for example, an adjacency matrix or a connection matrix. An adjacency matrix indicates the adjacency relationship between nodes in a graph due to edges (whether nodes are adjacent to each other via edges). An adjacency matrix is ​​a representation format of a graph that focuses on the adjacency relationship between nodes. On the other hand, a connection matrix is ​​a representation format of a graph that focuses on the connection relationship between nodes and edges. A connection matrix indicates the connection relationship between nodes and edges (whether nodes and edges are connected).

[0043] Here, as an example, an adjacency matrix (see FIG. 5(C)) is used as a matrix representing a graph. The adjacency matrix in FIG. 5(C) shows the adjacency relationship between nodes "1," "2," "3," "4," etc. in the graph shown in FIG. 5(B). FIG. 5(B) represents the shape of the rotor 12 shown in FIG. 5(A) as a graph with nodes and edges.

[0044] The adjacency matrix shown in Figure 5(C) indicates that node "1" and node "2" are adjacent, node "2" and node "4" are adjacent, node "4" and node "3" are adjacent, and node "3" and node "1" are adjacent.

[0045] As shown in Fig. 5(D), the feature vector indicates at least the position coordinates of the nodes included in the graph. The position coordinates indicate the positions where the nodes exist in the shape of the rotor 12 (structure). In the feature vector shown in Fig. 5(D), the coordinates of node "1" are represented by (x1, y1), the coordinates of node "2" are represented by (x2, y2), the coordinates of node "3" are represented by (x3, y3), and the coordinates of node "4" are represented by (x4, y4).

[0046] The feature vector shown in Fig. 5(D) has material data indicating the material associated with each node. That is, the feature vector shown in Fig. 5(D) is data in which the position coordinates and material data are associated with the node.

[0047] The material associated with a node is, for example, the material that constitutes the region adjacent to the node in the rotor 12 (structure). The material data indicates the material of the region adjacent to the node. In the feature vector of FIG. 5(D), the material data indicates whether the material of the region adjacent to the node is air, a permanent magnet, or both. In FIG. 5(D), the material data is described as a one-hot vector. Note that although the materials of the rotor 12 also include electromagnetic steel, it is omitted from the material data shown in FIG. 5(D) because the node is always adjacent to electromagnetic steel. By omitting electromagnetic steel, the data size can be reduced. Note that electromagnetic steel may be explicitly indicated in the material data.

[0048] In the feature vector of Figure 5(D), the material data of node "1" is expressed as (a,d) = (1,0), indicating that node "1" is adjacent to the region of air a but not adjacent to the region of magnet d. The material data of node "2" is expressed as (a,d) = (1,1), indicating that node "2" is adjacent to the region of air a and magnet d. The material data of node "3" is expressed as (a,d) = (1,0), indicating that node "3" is adjacent to the region of air a but not adjacent to the region of magnet d. The material data of node "4" is expressed as (a,d) = (1,1), indicating that node "4" is adjacent to the region of air a and magnet d.

[0049] Figure 6 shows another example of a feature vector. In the feature vector of Figure 6, the material data is data d mx Includes. d mxis a variable that represents the magnetization direction, and the direction ranging from -90 degrees to 90 degrees is indicated by a value ranging from 0 to 1. x is an identifier for each magnet included in the rotor 12 for one pole, and takes on values ​​from 1 to X (X is the number of magnets included in the rotor 12 for one pole). When the rotor 12 for one pole has three magnets (x=1, x=2, x=3), the material data is (a, d m1 ,d m2 ,d m3 )

[0050] In the feature vector of Figure 6, the material data of node "1" is (a, d m1 ,d m2 ,d m3 ) = (1,0,0,0), which indicates that node "1" is adjacent to the region of air a and not adjacent to any of the regions of magnets (x = 1 to 3). The material data of node "2" is (a,d m1 ,d m2 ,d m3 )=(1,d m1 ,0,0), node "2" is adjacent to the area of ​​air a and the magnet at x=1, and the magnetization direction of the magnet at x=1 is d m1 The material data of node "3" is (a,d m1 ,d m2 ,d m3 ) = (1,0,0,0), which indicates that node "3" is adjacent to the region of air a and not adjacent to any of the regions of magnets (x = 1 to 3). The material data for node "4" is (a,d m1 ,d m2 ,d m3 )=(1,d m1 ,0,0), node "4" is adjacent to the area of ​​air a and the magnet at x=1, and the magnetization direction of the magnet at x=1 is d m1 This indicates that

[0051] 5 and 6, the nodes of the graph indicate vertices included in the rotor 12 (structure). Here, a vertex refers to a point where two or more lines representing the shape intersect or are connected in a non-linear manner. In this case, the shape of the rotor 12 (structure) is represented by its vertices as nodes of the graph, and the straight lines between the vertices as edges of the graph. In other words, the graphs shown in FIGS. 5 and 6 represent the structure with a focus on the vertices.

[0052] FIG. 7 shows an example of a graph representing a structure, focusing on subregions included in the structure. Here, a subregion is a two-dimensional or three-dimensional partial region that constitutes the structure. A subregion is a small region obtained by dividing a structure into multiple parts, like an element in the finite element method. The shape of a subregion may be one type, or multiple types. A subregion can be set for each material of the structure.

[0053] In FIG. 7, the nodes of the graph indicate subregions included in the rotor 12 (structure). Also in FIG. 7, the nodes of the graph indicate subregions that represent the magnet regions included in the rotor 12 and subregions that represent the air regions present in the rotor 12. Also, the edges of the graph indicate that the subregions represented by the nodes are adjacent to each other. In the graph representation shown in FIG. 7, the electromagnetic steel region that represents most of the rotor 12 is not represented, and the magnet and air regions are represented by nodes and edges. By omitting the representation of the electromagnetic steel region, the data size can be reduced.

[0054] In addition, one magnet area may be represented by multiple nodes (multiple partial areas representing magnets), and one air area (hole) may be represented by multiple nodes (multiple blob areas representing air).

[0055] In the case of a graph representation focusing on a subregion, as shown in Figure 7, the feature vector may include data indicating either or both of the shape and size of the subregion, in addition to the node position coordinates and material data (which may include the magnetization direction) shown in Figures 5 and 6. Therefore, in the case of the graph representation shown in Figure 7, the feature vector tends to be high-dimensional data compared to the graph representations of Figures 5 and 6, but the number of edges can be reduced. Furthermore, in the case of the graph representation in Figure 7, it is easy to add new edges and new shapes can be easily represented, making it suitable for generating new shapes (shapes of new topologies).

[0056] On the other hand, in the case of the graph representations in Figures 5 and 6, although the number of edges increases, the feature vector can be made low-dimensional. Also, in the case of the graph representations in Figures 5 and 6, it is difficult to add new edges, so they are suitable for generating typical shapes (typical topology shapes such as those shown in Figures 3(A) to (C)).

[0057] A preferred embodiment of the present invention is shown in FIGS. 11 and 12. FIGS. 11(A-1) and 11(A-2) each show the same structure, with FIG. 11(A-1) illustrating that the structure is made of materials A, B, and C. FIG. 11(A-2) indicates the vertices of the structure with symbols 1 to 6. FIG. 11(B) shows an adjacency matrix based on FIG. 11(A-2), FIG. 11(C-1) shows a two-dimensional feature vector based on FIG. 11(A-2), and FIG. 11(C-2) shows a three-dimensional feature vector. Another example of a feature vector is shown in FIG. 12. FIG. 12(A) shows a two-dimensional feature vector based on FIG. 11(A-2), and FIG. 12(B) shows a three-dimensional feature vector based on FIG. 11(A-2). Using such preferred feature vectors allows for more efficient information processing related to structures and the like. 11(C-1) and (C-2) are feature vectors focusing on the vertices of the structure, and Fig. 12(A) and Fig. 12(B) are feature vectors focusing on the regions of the structure. The feature vectors focusing on the vertices of the structure make it easier to modify the design related to the vertices of the structure, and the feature vectors focusing on the regions of the structure make it easier to modify the design related to the regions of the structure.

[0058] Returning to Fig. 4, the generation process S400 by the generator 110 will be described. First, the generator 110 selects a generation topology (step S401). A plurality of selectable generation topologies are set in advance in the generator 110. The plurality of selectable generation topologies are, for example, the three rotor topologies shown in Figs. 3(A) to (C).

[0059] The generation topology may be selected by accepting a user operation to select the generation topology, or the generator 110 may automatically determine a generation topology from among a plurality of generation topologies based on an appropriate rule or randomly.

[0060] Each of a plurality of pre-set generating topologies is set with an adjacency matrix (a matrix representing a graph) and values ​​other than node position coordinates in a feature vector (for example, material data associated with each node). Therefore, when a generating topology is selected in step S401, values ​​other than node position coordinates in the adjacency matrix and feature vector are determined according to the selected generating topology (step S402). In other words, by selecting a generating topology, values ​​other than node position coordinates are determined in graph data 200 indicating the shape of the rotor 12 (structure). In this way, the generating topology indicates graph data in which only node position coordinates are variables (undetermined values). Note that the number of nodes in each generating topology may be a fixed value.

[0061] The generator 110 generates node position coordinates in the selected generated topology (step S403). By generating the position coordinates of each node, graph data 200 (generated data) consisting of an adjacency matrix and a feature vector is generated. An example of a method for generating node position coordinates will be described later.

[0062] The generator 110 verifies the generated graph data (generated data) (step S404). This verification is performed based on predetermined verification criteria to determine whether the generated graph data (generated data) properly represents the shape of the rotor 12. The verification criteria are, for example, criteria that indicate when the shape is inappropriate. Examples of criteria that indicate when the shape is inappropriate include "each edge intersects," "the node position coordinates are outside the specified area," and "the magnet has a shape other than a rectangle."

[0063] Edges (straight lines that define the shape of a structure) should not intersect, so if any edges intersect, the shape is inappropriate.

[0064] The node position coordinates must be within a predetermined designated area in each generated topology, so if they are outside the designated area, the shape is inappropriate. The designated area can be set within the rotor 12.

[0065] The magnets should be rectangular, so if the magnets are shaped other than rectangular, they are improperly shaped.

[0066] If the verification in step S404 determines that the generated shape is inappropriate (problematic) (step S405), the process returns to step S403, and node position coordinates in the selected generated topology are regenerated.

[0067] If the verification in step S404 determines that the generated shape is appropriate (no problem) (step S405), the generator 110 outputs the generated graph data 200 (generated data). Therefore, only appropriate graph data 200 is output from the generator 110. The output graph data 200 is provided to the evaluator 120.

[0068] FIG. 8 shows variations of the generation of coordinates in step S403. FIG. 8(A) shows an example in which the coordinates (x, y) of a node are generated as random numbers based on an appropriate random number seed. FIG. 8(B) shows an example in which coordinates are sampled using a machine learning model such as kernel density estimation or a Gaussian mixture model. FIG. 8(C) shows an example of generation using a deep generative model such as variational autoencoding. Note that data other than coordinates in a feature vector may also be generated along with the coordinates.

[0069] Fig. 9 shows a variation of the evaluator 120 shown in Fig. 1. The property evaluation of the structure based on the graph data 200 may be either of Fig. 9(A) or Fig. 9(B).

[0070] 9(A) includes a first characteristic evaluation model 121 (trained model) that has been machine-learned to output the characteristics of a motor equipped with a rotor 12 having a shape indicated by graph data 200 when the graph data 200 is input. The first characteristic evaluation model 121 has been machine-learned (deep learning) using the graph data 200 indicating the shape of the rotor 12 and characteristic data of the motor equipped with the rotor 12 as learning data.

[0071] 9(A), graph data 200 output from the generator 110 is given to the input of the first characteristic evaluation model 121. When the graph data 200 is input, the first characteristic evaluation model 121 can output motor characteristics.

[0072] The evaluator 120 shown in FIG. 9(B) includes a converter 125 and a second characteristic evaluation model 126. The converter 125 converts the graph data 200 into an image. The image represents the shape of the rotor 12 indicated by the graph data 200 in a predetermined image data format (e.g., bitmap). The second characteristic evaluation model 126 has been machine-learned so that, when an image showing the shape of the rotor 12 is input, it outputs the characteristics of a motor including a rotor 12 having the shape shown in the image. The second characteristic evaluation model 126 has been machine-learned (deep learning) using the image showing the shape of the rotor 12 and characteristic data of a motor including that rotor 12 as learning data.

[0073] 9(B), graph data 200 output from generator 110 is provided to converter 125. Converter 125 converts the provided graph data 200 into an image and outputs it. The image output from converter 125 is provided as an input to second characteristic evaluation model 126. When the image is input, second characteristic evaluation model 126 can output motor characteristics.

[0074] The evaluator 120 in FIG. 9(A) does not require the converter 125 of the evaluator 120 in FIG. 9(B), thereby simplifying the configuration. Furthermore, the evaluator 120 in FIG. 9(B) can utilize an image-based evaluation model 126. Image input is common in characteristic evaluation using machine learning models, and many excellent existing techniques for characteristic evaluation exist. Therefore, by adopting an image-input evaluation model 126 as in FIG. 9(B), it is possible to utilize many existing techniques for characteristic evaluation.

[0075] As the motor characteristic evaluation model 126 for image input shown in Fig. 9(B), for example, a prediction model (deep learning model) described in Non-Patent Documents 1 and 2 can be used. As the motor characteristic evaluation model 121 for graph input shown in Fig. 9(A), one can be used that is constructed by machine learning (deep learning) using the input of the prediction model described in Non-Patent Documents 1 and 2 as graph data 200 instead of an image.

[0076] Note that data other than the graph data 200 or images (for example, operating conditions of the motor, other conditions of the motor) may also be input to the evaluation models 121, 126 in FIGS. 9(A) and 9(B).

[0077] The present invention is not limited to the above-described embodiment and can be modified in various ways. In the above-described embodiment, a motor is described as one suitable example. However, the present invention can be applied to other applications. For example, the present invention can be applied to piezoelectric actuators, RFID (Radio Frequency Identification), cables, high-frequency preheaters, capacitors, magnetic heads, magnetic shields, magnetic gears, magnets, circuit breakers, sensors, electrical heating devices, electrolytic plating, electromagnetic forming, electromagnetic retarders, electromagnetic relays, transformers and reactors, bus bars, generators, heaters, bearings, magnetic couplings, induction heating devices, induction machines, linear solenoids and linear actuators, and wireless power transfer systems. When the present invention is applied to a motor, it has the advantage of being able to express motor design information using only the minimum necessary information. [Industrial Applicability]

[0078] The information processing system of the present invention is particularly useful for designing motors and the like. [Explanation of symbols]

[0079] 10: Interior permanent magnet synchronous motor 11: Stator 12: Rotor 15: Permanent magnet 16: Slot 100: Design System 110: Generator 120: Evaluator 121: Motor characteristic evaluation model 125: Converter 126: Motor characteristic evaluation model 150: Judgment device 200: Graph data 310: Generative Model 320: Property prediction model 400:Image 450: Motor characteristics

Claims

1. An information processing system using vectors, wherein the vectors include position information and constituent material information, and the information processing system uses three or more pieces of position information.

2. 2. The information processing system according to claim 1, wherein the vector information is used to design a structure.

3. 3. The information processing system according to claim 2, wherein the shape of the structure is expressed using the vector information.

4. 4. The information processing system according to claim 3, wherein the position information includes x-axis and y-axis information.

5. The information processing system according to claim 3, wherein four or more pieces of position information are used.

6. The information processing system according to claim 5 , wherein the position information further includes z-axis information.

7. The information processing system according to claim 1 , wherein the constituent material information includes air.

8. 2. The information processing system according to claim 1, wherein the constituent material information includes magnetic materials.

9. The information processing system according to claim 1 , wherein the constituent material information includes physical property information.

10. 2. The information processing system according to claim 1, further comprising evaluation means for evaluating characteristics of a structure from information on said vector when said vector is received from a user.

11. 11. The information processing system according to claim 10, further comprising: learning a relationship between the vector and the characteristics of the structure; predicting the characteristics from the vector of the user through the learning; and outputting the predicted evaluation value to the user.

12. A program that causes an information processing system for calculating the characteristics of a structure to function as a means for learning the relationship between said characteristics by associating said characteristics with vectors including position information and constituent material information, predicting said characteristics from said vectors received from a user through said learning, and outputting the predicted evaluation value to said user.

Images