Equivalent circuit of magnetic coupling circuit, conversion method, feature extraction method, manufacturing method of equivalent circuit of magnetic coupling circuit, manufacturing method of system, and manufacturing method of feature extraction device

Abstract

This equivalent circuit of a magnetic coupling circuit is an equivalent circuit of a single-phase magnetic coupling circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, said equivalent circuit being characterized by: including a first impedance provided between a primary-side positive terminal and a primary-side negative terminal, a second impedance provided between a secondary-side positive terminal and a secondary-side negative terminal, a third impedance provided between the primary-side positive terminal and the secondary-side positive terminal, and a fourth impedance provided between the primary-side negative terminal and the secondary-side negative terminal; and the circuit constants of the third impedance and the fourth impedance being equivalent.

Description

Equivalent circuit, conversion method, and feature extraction method for magnetically coupled circuits

[0001] The present disclosure relates to an equivalent circuit of a magnetic coupling circuit, a conversion method, and a feature extraction method.

[0002] In a circuit that includes mutual inductance, it is difficult to determine the mutual inductance acting between, for example, two inductance elements from the actual circuit. Therefore, for example, Patent Document 1 discloses a method of replacing the mutual inductance included in an equivalent circuit of an inductance element using a ferrite material with an inductance element.

[0003] Japanese Patent Application Publication No. 11-312187

[0004] If a circuit is represented by a graph structure consisting of points and lines, with circuit components as points and the wiring connecting the circuit components as lines, for example, in a circuit that includes magnetic coupling between two inductance elements like the one described above (a single-phase magnetically coupled circuit), the inductance elements are connected by spatial coupling, so there are no wirings and it is not possible to define lines.As a result, even if a single-phase magnetically coupled circuit is converted into a graph structure, the magnetic coupling is not reflected in the graph structure, and information degradation occurs when the magnetically coupled circuit is converted into a graph structure.

[0005] Therefore, in order to convert a single-phase magnetically coupled circuit into a graph structure while suppressing information degradation, it is expected that the magnetically coupled circuit will be replaced with an equivalent circuit that expresses spatial coupling as circuit components, and that this equivalent circuit will be converted into a graph structure. However, when processing a magnetically coupled circuit as a graph structure, Kirchhoff's law of conservation of current does not hold, so there was a problem that simply converting a conventional equivalent circuit into a graph structure did not allow for graph processing that took into account the characteristics of the magnetically coupled circuit.

[0006] The present disclosure has been made to solve the above-mentioned problems, and aims to obtain an equivalent circuit of a single-phase magnetic coupling circuit that enables graph processing taking into account the characteristics of the magnetic coupling circuit.

[0007] The equivalent circuit of the magnetic coupling circuit according to the present disclosure is an equivalent circuit of a single-phase magnetic coupling circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, and includes a first impedance provided between the primary-side + terminal and the primary-side − terminal, a second impedance provided between the secondary-side + terminal and the secondary-side − terminal, a third impedance provided between the primary-side + terminal and the secondary-side + terminal, and a fourth impedance provided between the primary-side − terminal and the secondary-side − terminal, and is characterized in that the circuit constants of the third impedance and the fourth impedance are equal.

[0008] According to the present disclosure, it is possible to obtain an equivalent circuit of a single-phase magnetic coupling circuit, which allows graph processing based on the characteristics of the magnetic coupling circuit.

[0009] 5A is a diagram showing an example of a single-phase magnetic coupling circuit according to embodiment 1; FIG. 5B is a diagram showing an example of an equivalent circuit of the single-phase magnetic coupling circuit according to embodiment 1; FIG. 5C is a schematic diagram showing an example of the configuration of a typical conventional transformer; FIG. 5D is a diagram showing an example of a T-type equivalent circuit of the transformer shown in FIG. 3; FIG. 5E is a schematic diagram summarizing the relationship between transformation and inverse transformation between the single-phase magnetic coupling circuit shown in FIG. 1 and the equivalent circuit shown in FIG. 2, where FIG. 5A is the single-phase magnetic coupling circuit shown in FIG. 1 and FIG. 5B is the equivalent circuit shown in FIG. 2; FIG. 5F is a schematic diagram showing the circuit structure and circuit constants of the equivalent circuit according to embodiment 1; FIG. 5G is a diagram showing calculation results by circuit simulation for a circuit of a single-phase transformer including magnetic coupling, a conventional T-type equivalent circuit, and the equivalent circuit shown in FIG. 2, where FIG. 5A shows a single-phase magnetic coupling circuit equipped with an AC power supply and a load resistance, FIG. 7B shows a conventional T-type equivalent circuit of the magnetic coupling circuit shown in FIG. 7A, FIG. 7C shows the equivalent circuit of FIG. 2 equipped with an AC power supply and a load resistance, and FIG. 7D shows the frequency characteristics of the voltage across the load resistance on the secondary side. 1 is a diagram showing an example of a single-phase magnetic coupling circuit according to embodiment 2. FIG. 2 is a diagram showing an example of an equivalent circuit of the single-phase magnetic coupling circuit according to embodiment 2. FIG. 3 is a diagram showing an example of a single-phase magnetic coupling circuit according to embodiment 2. FIG. 4 is a diagram showing an example of a graph structure according to embodiment 3. FIG. 5 is a diagram showing an example of an adjacency matrix in embodiment 3. FIG. 6 is a diagram showing an example of a graph structure according to embodiment 3. FIG. 7 is a diagram showing an example of a graph structure according to embodiment 3. FIG. 8 is a diagram showing an example of a graph structure according to embodiment 3. FIG. 9 is a schematic diagram showing an example of a method of assigning attribute information in embodiment 3. FIG. 10 is a diagram showing an example of a dataset used for a classification problem of a graph neural network according to embodiment 3. FIG. 11 is a diagram showing, as a conventional example, inference accuracy for test data when training is performed without taking magnetic coupling components into consideration. FIG. 12 is a diagram showing inference accuracy when training data and test data are created using the conversion method according to embodiment 3 and the type of circuit component is inferred using the test data. FIG. 13 is a diagram showing inference accuracy for test data when training is performed using a dataset in which the circuit constant of the third impedance is doubled and the fourth impedance is short-circuited in the equivalent circuit according to embodiment 3.

[0010] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. Embodiment 1. A circuit including a single-phase magnetic coupling according to embodiment 1 (hereinafter referred to as a "single-phase magnetic coupling circuit") is a circuit in which two inductance elements (coils) that are insulated against direct current and divided into a primary side and a secondary side are spatially coupled by mutual inductance. In embodiment 1, a method for converting this single-phase magnetic coupling circuit into a graph structure while suppressing information degradation is shown, in which the magnetic coupling circuit is replaced with an equivalent circuit including four impedance elements to convert into a graph structure.

[0011] Single-phase magnetic coupling circuits are used in, for example, circuits that convert the ratio between input voltage and output voltage, such as transformers, isolation transformers in which the input voltage and output voltage are equal and are used for DC insulation, equivalent circuits of inductance elements such as ferrite cores, equivalent circuits of common-mode choke coils, circuits for equivalently reducing residual inductance using magnetic coupling, wireless power transmission circuits, and magnetic levitation in superconductivity. In any of these cases, the single-phase magnetic coupling circuit has a circuit configuration represented by two inductance elements and the mutual inductance or coupling coefficient acting between the two inductance elements. Therefore, in embodiment 1, a case in which the single-phase magnetic coupling circuit is used in a transformer will be described as an example.

[0012] Fig. 1 is a diagram showing an example of a single-phase magnetically coupled circuit. Fig. 1 shows a typical magnetically coupled circuit in which two inductance elements are coupled by spatial coupling, and part of the alternating current flowing between the terminals of the primary inductance element of the two inductance elements is transmitted to the secondary inductance element, generating a voltage between the terminals of the secondary inductance element. In the following description, for simplicity, the primary inductance element in the magnetically coupled circuit may be simply referred to as the "primary side," and the secondary inductance element may be simply referred to as the "secondary side."

[0013] In the single-phase magnetic coupling circuit shown in Figure 1, both the primary and secondary sides are shown as open ends. However, for example, by connecting an AC power supply between the + and - terminals on the primary side and connecting a load resistor between the + and - terminals on the secondary side, current flows through the circuit, and power from the primary side is propagated to the secondary side due to magnetic coupling. Conversely, by connecting an AC power supply between the terminals on the secondary side and a load resistor between the terminals on the primary side, a portion of the AC current flowing through the secondary side flows back to the primary side. Furthermore, by connecting an AC power supply between the + terminal on the primary side and the + terminal on the secondary side, as in a common mode choke coil, and connecting a load circuit between the - terminal on the primary side and the - terminal on the secondary side, it has the function of reducing common mode noise from the AC power supply to the load circuit, or from the load circuit to the AC power supply. Because the circuit in Figure 1 is symmetrical, in embodiment 1, the primary side will be described as the input and the secondary side as the output.

[0014] Fig. 2 is a diagram showing an example of an equivalent circuit of the magnetic coupling circuit shown in Fig. 1. This equivalent circuit is configured to include, for example, a first impedance provided between the + terminal and - terminal on the primary side, a second impedance provided between the + terminal and - terminal on the secondary side, a third impedance provided between the + terminal on the primary side and the + terminal on the secondary side, and a fourth impedance provided between the - terminal on the primary side and the - terminal on the secondary side.

[0015] The first impedance is a circuit component connected in parallel between the + and - terminals on the primary side, the second impedance is a circuit component connected in parallel between the + and - terminals on the secondary side, the third impedance is a circuit component connected in series between the + terminal on the primary side and the + terminal on the secondary side, and the fourth impedance is a circuit component connected in series between the - terminal on the primary side and the - terminal on the secondary side.

[0016] The polarity of the magnetic coupling circuit can be changed, for example, by changing the direction of the current, reversing the phase of the current by 180°, or reversing the winding direction of the conductor that forms the inductance, so the + terminal and the - terminal may be inversely related. Therefore, in the first embodiment, the + terminal and the - terminal are used to mean either end of the inductance element. In particular, in the first embodiment, the polarity does not matter, so it is sufficient that one terminal of the primary inductance element and one terminal of the secondary inductance element, and the other terminal of the primary inductance element and the other terminal of the secondary inductance element are connected.

[0017] Looking at the correspondence between the magnetic coupling circuit shown in Fig. 1 and the equivalent circuit shown in Fig. 2, the inductance element on the primary side in the magnetic coupling circuit shown in Fig. 1 corresponds to the first impedance in the equivalent circuit shown in Fig. 2, and the inductance element on the secondary side in the magnetic coupling circuit shown in Fig. 1 corresponds to the second impedance in the equivalent circuit shown in Fig. 2. Furthermore, the mutual inductance in the magnetic coupling circuit shown in Fig. 1 corresponds to the third impedance and the fourth impedance in the equivalent circuit shown in Fig. 2. In other words, the magnetic coupling circuit shown in Fig. 1 and the equivalent circuit shown in Fig. 2 have a structural correspondence relationship.

[0018] However, in the magnetic coupling circuit shown in Fig. 1, the inductances of the primary and secondary inductance elements and the mutual inductance are interdependent. Therefore, the circuit constants of each of the first to fourth impedances in the equivalent circuit shown in Fig. 2 must be determined by the combination of the inductances of the primary and secondary inductance elements and the mutual inductance in the magnetic coupling circuit shown in Fig. 1.

[0019] If the magnetic coupling circuit shown in Fig. 1 can be expressed as the equivalent circuit shown in Fig. 2, the primary-side inductance element and the secondary-side inductance element in the magnetic coupling circuit shown in Fig. 1 will be connected by wiring via the third impedance and the fourth impedance. This makes it possible to express the magnetic coupling circuit shown in Fig. 1 as a graph structure with circuit components as nodes and wiring as edges.

[0020] Fig. 3 is a schematic diagram showing a typical configuration example of a conventional transformer. This transformer is a single-phase, two-wire transformer, and the primary conductor (winding) N １ and secondary conductor (winding) N ２ The primary conductor N is wound around a single annular iron core. １ and secondary conductor N ２ The ratio between the primary and secondary voltages changes depending on the number of times the conductor is wound around the iron core. In this case, if the number of times the conductor is wound around the primary and secondary sides is the same, the ratio between the primary and secondary voltages is 1. Furthermore, if the number of times the conductor is wound around the primary side is less than the number of times the conductor is wound around the secondary side, a boosted voltage is output to the secondary side, and if the number of times the conductor is wound around the primary side is more than the number of times the conductor is wound around the secondary side, a reduced voltage is output to the secondary side.

[0021] Furthermore, by using a material with a high relative magnetic permeability for the iron core, the magnetic flux generated by the current flowing through the primary terminals is less likely to leak outside the iron core, making it possible to bring the ratio of the power input to the primary side to the power output from the secondary side closer to 1. Conversely, by using a material with a low dielectric constant for the iron core, the power input to the primary side is less likely to be transmitted to the secondary side. The amount of coupling between the primary side and the secondary side, which is determined by the dielectric constant of the iron core material, the dimensions of the iron core, or the structure of the iron core, is expressed as a coupling coefficient k. Using the coupling coefficient k, the inductance of the primary side of the transformer can be expressed as L １ , the secondary inductance is L ２ , and the mutual inductance is M, the relationship between them is expressed by the following formula (1). １ and inductance L ２Although each of these means self-inductance, in the first embodiment, they will be simply referred to as inductance.

[0022] In this case, k takes a value between 0 and 1. Furthermore, when defining k including polarity, k can be between -1 and 1, but in the first embodiment, polarity is not important, so k is set to between 0 and 1. In many magnetic coupling circuits including transformers, k close to 1 is an ideal condition, as it means that power on the primary side is transmitted to the secondary side without loss. However, when a large magnetic flux is applied to the iron core, the iron core generates heat, and the heat generated can cause magnetic saturation and a decrease in relative permeability, or the heat generated by the iron core can thermally destroy the iron core (magnetic material).

[0023] For this reason, methods such as creating a gap (gap) in the middle of a ring-shaped iron core to prevent magnetic flux from passing through and thereby prevent magnetic saturation of the iron core, or using a material with low relative magnetic permeability for the iron core to intentionally generate leakage flux, are used. In such cases, k is not "1" but is greater than or equal to 0 and less than 1. Furthermore, k varies depending on the gap structure, the material used for the magnetic body, the temperature of the magnetic body, and other factors. Thus, while k is ideally "1," in actual circuits, it is not "1," such as "0.9" or "0.99." In particular, in cases such as wireless power transmission, where it is difficult to use magnetic bodies and misalignment between inductance elements is likely, k takes a small value, such as "0.1." Therefore, the equivalent circuit of a magnetic coupling circuit must be able to represent k values ​​less than 1.

[0024] Generally, the inductance of the primary side of the transformer is L １ , the secondary inductance is L ２ , the current flowing through the primary inductance element is I １ , the current flowing through the secondary inductance element is I ２ , inductance L １ and inductance L ２If the mutual inductance between the primary and secondary inductance elements is M, j is a complex number, and ω is an angular frequency, the voltage v1 between the terminals of the primary inductance element and the voltage v2 between the terminals of the secondary inductance element are calculated by equations (2) and (3), respectively. Note that in Figures 1 to 3 and Figure 4 onwards, which will be described later, the voltages v1 and v2 are represented by capital letters V1 and V2.

[0025] 4 is a diagram showing an example of a T-type equivalent circuit of the transformer shown in FIG. 3. In the equivalent circuit shown in FIG. 4, the first impedance is expressed as j×ω×(L １ -M), the second impedance is j × ω × (L ２ By setting the third impedance to j×ω×M, the equivalent circuit shown in Fig. 4 satisfies the above formula (2). Therefore, the transformer shown in Fig. 3 can be expressed by replacing it with the circuit components in the equivalent circuit shown in Fig. 4. In this case, since the mutual inductance M can be expressed as in the above formula (1), the equivalent circuit shown in Fig. 4 can be expressed by the above formula (2) even when the coupling coefficient k varies.

[0026] However, in an actual transformer, the negative terminal on the primary side and the negative terminal on the secondary side are insulated against direct current, whereas the equivalent circuit shown in Figure 4 has a structure that includes a short circuit, so the structure is different. Since Kirchhoff's law of conservation of current is applied to solving problems related to electrical circuits, even if the negative terminal on the primary side and the negative terminal on the secondary side are shorted in the equivalent circuit shown in Figure 4, the calculation results are obtained as if the structure were insulated on both sides.

[0027] However, when processing a circuit using a graph network or graph neural network, it is not possible to take into account Kirchhoff's law of conservation of current. This is because Kirchhoff's law considers a closed path in which the current flowing out of and the current flowing into a circuit component are equal, assuming that the start and end points are the same circuit component, whereas a graph network or graph neural network only considers the connection between two components in one hidden layer. Even if the number of hidden layers is increased, the law of conservation of current does not hold, and a closed path cannot be considered. Therefore, if the equivalent circuit shown in Figure 4 is converted into a graph structure and processed using a graph network or graph neural network, a circuit including a condition in which the primary side and secondary side are short-circuited, which is not intended by the user, will be processed.

[0028] A commonly conceivable solution to this problem is to arrange the first impedance and the second impedance separately on the positive terminal side and the negative terminal side in the equivalent circuit shown in Figure 4. This prevents the equivalent circuit from becoming a short-circuited structure. However, because the equivalent circuit shown in Figure 4 has a circuit structure (circuit topology) in which a direct current flows between the primary side and the secondary side, there is a problem in that even if the circuit can be calculated correctly using Kirchhoff's laws, it cannot be processed correctly based on circuit theory in a graph network or graph neural network.

[0029] On the other hand, the equivalent circuit shown in Fig. 2 is configured to include a first impedance connected in parallel between the primary terminals, a second impedance connected in parallel between the secondary terminals, and a third impedance and a fourth impedance that represent the mutual inductance between the primary inductance and the secondary inductance. In other words, the equivalent circuit shown in Fig. 2 has a circuit structure similar to the physical structure of a transformer.

[0030] Furthermore, by making the circuit constants of the third impedance and the fourth impedance equal in the equivalent circuit shown in FIG. 2, even when processing a graph network or graph neural network that includes only one of the paths of the third impedance or the fourth impedance, a short circuit between the primary side and the secondary side does not occur, and processing can be performed as correct circuit information based on circuit theory.

[0031] 2 has a circuit structure corresponding to the structure of an actual transformer, it is possible to prevent current from flowing through the third impedance or the fourth impedance depending on the circuit constant of the first impedance or the second impedance, and it is also possible to achieve behavior similar to that of a circuit mounted on an actual transformer. Therefore, the equivalent circuit shown in FIG. 2 has the excellent effect of enabling electrically correct processing even in a graph network or a graph neural network.

[0032] Although the structure of the equivalent circuit shown in Figure 2 can be expressed as a graph structure, it is generally difficult to determine circuit constants that are electrically equivalent to an arbitrarily defined circuit structure. Therefore, to prove that the equivalent circuit shown in Figure 2 is valid, it is necessary to show that electrically equivalent circuit constants can be defined for the proposed circuit structure.

[0033] In the first embodiment, it has been derived that the circuit constants of the single-phase magnetic coupling circuit shown in Fig. 1, the equivalent circuit shown in Fig. 2, and the equivalent circuit based on the conventional example shown in Fig. 4 can be determined so as to have the same electrical characteristics according to Kirchhoff's law. Specifically, assuming that j is a complex number and ω (= 2π × (frequency [Hz])) is an angular frequency, the inductance between the primary terminals in the magnetic coupling circuit shown in Fig. 1 is L １ , the inductance between the secondary terminals is L ２ , the mutual inductance between the primary inductance element and the secondary inductance element is M. In addition, in the equivalent circuit shown in FIG. 2, the circuit constant of the first impedance is Z １ , the circuit constant of the second impedance is Z ２, the circuit constant of the third impedance is Z ３ , the circuit constant of the fourth impedance is Z ４ Then, the circuit constants are calculated using the following equations (4) to (6).

[0034] Furthermore, since the mutual inductance M can be expressed as in the above formula (1), the above formulas (4) to (6) can also be expressed as in the following formulas (7) to (9).

[0035] In particular, since the coupling coefficient k is a real number between 0 and 1, M is 0 or more, and Z １ From Z ４ The numerator of is also 0 or more. ３ and Z ４ The imaginary part of is always greater than or equal to 0. １ and Z ２ are in the denominator (L ２ -M), and (L １ -M), and L １ >L ２ Then, equation (5) holds.

[0036] For example, if k = 1, L_1-M<0, so L １ -M is a negative real number. Therefore, Z ２ The imaginary part of can take a value less than or equal to 0. １ <L ２ In the case of Z １ The imaginary part of can take a value less than or equal to 0. However, L １ and L ２ The imaginary parts of the two expressions are never simultaneously equal to or less than 0.

[0037] The sum of the circuit constants of the third impedance and the fourth impedance is Z ３ and Z ４As long as the sum of the third and fourth impedances is the same, there is no problem with how the circuit constants of the third and fourth impedances are distributed as long as Kirchhoff's law is used. However, when using a graph network or graph neural network, the circuit constants of both the third and fourth impedances must be equal in order to obtain the same results regardless of whether the circuit component is the third or fourth impedance. In this way, in a graph network or graph neural network, by making the circuit constants of the third and fourth impedances equal, the user can obtain the correct results they intended without having to consider the return path.

[0038] Conversely, the inductance and mutual inductance, which are circuit constants in the single-phase magnetic coupling circuit shown in Fig. 1, can be calculated from the equivalent circuit shown in Fig. 2. For example, when the complex number is j, the angular frequency is ω, and the first impedance is Z １ , the second impedance is Z ２ , the third impedance is Z ３ Then, the inductance L between the primary terminals in the magnetic coupling circuit shown in FIG. 1 can be calculated by the following equations (11) to (13): １ , inductance L between the secondary terminals ２ , and the mutual inductance M between the primary and secondary sides can be calculated.

[0039] Fig. 5 shows a schematic diagram summarizing the relationship between the conversion and inverse conversion between the single-phase magnetic coupling circuit shown in Fig. 1 and the equivalent circuit shown in Fig. 2. Fig. 5A shows the single-phase magnetic coupling circuit shown in Fig. 1, and Fig. 5B shows the equivalent circuit shown in Fig. 2.

[0040] As shown in Figures 5A and 5B, there is a reversible transformation relationship between a single-phase magnetic coupling circuit and an equivalent circuit in which the spatial coupling in this magnetic coupling circuit is replaced with circuit components. Therefore, this reversible transformation relationship can be applied to applications other than converting a magnetic coupling circuit into a graph structure. For example, by setting a design target in the equivalent circuit shown in Figure 5B and converting this equivalent circuit into the magnetic coupling circuit shown in Figure 5A using the above equations (11) to (13), it is possible to determine the physical structure or material constants of the actual magnetic coupling circuit.

[0041] Thus, according to the first embodiment, by combining two different methods of expression, that is, an actual magnetic coupling circuit and an equivalent circuit of the magnetic coupling circuit, it is possible to perform various processes in the area in which each circuit excels. Therefore, the reversible transformation relationship shown in FIG. 5 contributes to the improvement of efficiency when designing a magnetic coupling circuit or the structural design of a transformer or the like incorporating this magnetic coupling circuit, and can also be used when extracting mutual inductance from an actual magnetic coupling circuit. For example, L １ , L ２To calculate M, it is necessary to determine three unknowns: the first impedance, the second impedance, and the third impedance (since the fourth impedance is equivalent to the third impedance, only one of them needs to be determined) in FIG. 5B. To do so, three independent measurement results are required. For example, the first impedance, the second impedance, and the third impedance can be determined by measuring the impedance seen from the V1 side when the + and - terminals of V2 are shorted, open, and connected at 50 Ω. Furthermore, the measurement accuracy of the first impedance, the second impedance, and the third impedance can be improved by attaching a circuit component other than 50 Ω, such as a resistor, coil, or capacitor, to the V1 terminal and measuring the impedance from V2, or by attaching a circuit component between the + terminal of V1 and the + terminal of V2 and measuring the impedance between the - terminal of V1 and the - terminal of V2. Specifically, since a constraint equation with more than three constraints can be established for three unknowns (variables), it becomes an overdetermined equation. Therefore, it is desirable to determine the first impedance, the second impedance, and the third impedance as a solution of the least squares method using a generalized inverse matrix. From the determined first impedance, second impedance, and third impedance, L is uniquely determined according to the method shown in this embodiment. １ , L ２ , M can be determined. Conventionally, in the method using equations (2) and (3), the mutual inductance M is included and it is necessary to determine the voltages v1, v2 and the currents I1, I2 simultaneously because it depends on the currents I1, I2. １ , L ２ , M are mutually dependent, making it difficult to measure them. On the other hand, by defining Z1, Z2, and Z3 in equations (4), (5), and (6), we can calculate L in equations (11), (12), and (13). １ , L ２ , M, an unprecedented effect can be obtained in that calculations can be performed without requiring special processing because the dependency is resolved.

[0042] 6 is a schematic diagram of the circuit structure and circuit constants of the equivalent circuit shown in FIG. 2. In FIG. 6, the first impedance Z １ to the fourth impedance Z ４The circuit constants up to are expressed as inductance.

[0043] As shown in Fig. 1, the single-phase magnetic coupling circuit is expressed only by the inductance component, and therefore the circuit constants of the equivalent circuit shown in Fig. 2 are also expressed only by the inductance component. For example, in Fig. 6, the primary inductance L １ , secondary inductance L ２ , and the mutual inductance M is a real number, so the first impedance Z １ , second impedance Z ２ , third impedance Z ３ , and a fourth impedance Z ４ can also be treated as a single component with real circuit constants.

[0044] Figure 7 shows the results of calculations using a circuit simulator for a single-phase magnetic coupling circuit, a T-type equivalent circuit of this magnetic coupling circuit, and the equivalent circuit shown in Figure 2. The calculation results show the frequency characteristics of the voltage across the load resistance on the secondary side when a 1 V AC signal is applied to the primary side (input) and the frequency of the input AC signal is changed.

[0045] 7A shows a single-phase magnetic coupling circuit including an AC power supply and a load resistance, FIG. 7B shows a T-type equivalent circuit of the magnetic coupling circuit shown in FIG. 7A, FIG. 7C shows the equivalent circuit of FIG. 2 including an AC power supply and a load resistance, and FIG. 7D shows the frequency characteristics of the voltage across the load resistance on the secondary side.

[0046] 7A, the coupling coefficient k1 between the primary and secondary inductance elements is 0.9. Also, in FIG. 7A, the primary inductance L1a serving as the input is 100 μH, the secondary inductance L2a serving as the output is 50 μH, and the load resistance impedance Ra is 10 Ω.

[0047] In addition, in FIG. 7B, the inductance L1b is "36.1 μH", the inductance L2b is "-13.9 μH", the inductance L3b is "63.6 μH", and the impedance Rb of the load resistor is "10 Ω".

[0048] In addition, in FIG. 7C, the inductance L1c is "-69.7 μH", the inductance L2c is "26.1 μH", the inductance L3c is "7.5 μH", the inductance L4c is "7.5 μH", and the impedance Rc of the load resistor is "10 Ω".

[0049] 7D, the voltage across the load resistor of the circuit shown in FIG. 7A is represented as V(out_a), the voltage across the load resistor of the circuit shown in FIG. 7B is represented as V(out_b), and the voltage across the load resistor of the circuit shown in FIG. 7C is represented as V(out_c1, out_c2). As shown in FIG. 7D, the frequency characteristics of the above three circuits are consistent in both amplitude and phase. In other words, the equivalent circuit shown in FIG. 7C can also represent the single-phase magnetically coupled circuit shown in FIG. 7A with the same accuracy as the conventional T-type equivalent circuit shown in FIG. 7B. The circuit constants can be calculated as follows, for example, using the programming language Python (version 3.9.17):# -------------- python code (start) ----------------------- from math import sqrt # Import libraries L1 = 100 * 1e-6 # 100μH L2 = 50 * 1e-6 # 50μH k = 0.9 # Coupling coefficient M = k * sqrt(L1 * L2) # Mutual inductance print("Initial conditions") print(f"L1: {L1:.3e}, L2: {L2:.3e}, M: {M:.3e}, k: {k:.3e}", end="\n\n") # Equations (4) to (6) z1 = (L1*L2 - M*M) / (L2-M) z2 = (L1*L2 - M*M) / (L1-M) z3 = (L1*L2 - M*M) / (2*M) print("Forward mutual inductance with → without") print(f"z1: {z1:.3e}, z2: {z2:.3e}, z3: {z3:.3e}", end="\n\n") # Equations (11) to (13) L1_inv = (z1*(z2 + 2*z3)) / (z1 + z2 + 2*z3) L2_inv = (z2*(z1 + 2*z3)) / (z1 + z2 + 2*z3) M_inv = z1*z2 / (z1 + z2 + 2*z3) k_inv = M_inv / (sqrt(L1_inv*L2_inv)) print("No mutual inductance in the reverse direction → Yes") print(f"L1_inv: {L1_inv:.3e}, L2_inv: {L2_inv:.3e}, M_inv: {M_inv:.3e}, k_inv: {k_inv:.3e}") # -------------- python code (end) ----------------------- When you run the above program, you will get the following result.# -------------- python code output (start) ----------------------- Initial conditions L1: 1.000e-04, L2: 5.000e-05, M: 6.364e-05, k: 9.000e-01 Forward direction Mutual inductance present → absent z1: -6.965e-05, z2: 2.613e-05, z3: 7.464e-06 Reverse direction Mutual inductance absent → present L1_inv: 1.000e-04, L2_inv: 5.000e-05, M_inv: 6.364e-05, k_inv: 9.000e-01 # -------------- python code output (end) ----------------------- From this output result, L1 and L1_inv, L2 and L2_inv, and M and M_inv are equal, and as a result, k and k_inv are equal, which indicates that the transformations of # equations (4) to (6) and the inverse transformations of # equations (11) to (13) are reversible. Furthermore, it can be seen that this is a general solution that is not dependent on the power supply voltage or load resistance.

[0050] Note that the circuit constants of the above-mentioned circuits are merely examples, and other circuit constants may be used. Also, in the above explanation, an example was described in which a single-phase magnetic coupling circuit is used in a transformer, but this magnetic coupling circuit is not limited to transformers, and may be used in any circuit, regardless of application, in which two coils (inductors) are magnetically coupled by mutual inductance, such as an equivalent circuit of a common mode choke coil, an equivalent circuit of a ferrite core, a component for canceling residual inductance, a wireless power transmission circuit, coupling between wirings due to residual inductance, or coupling between circuit components.

[0051] As described above, according to the first embodiment, the equivalent circuit of the magnetic coupling circuit is a single-phase equivalent circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, and includes a first impedance provided between the primary-side positive terminal and the primary-side negative terminal, a second impedance provided between the secondary-side positive terminal and the secondary-side negative terminal, a third impedance provided between the primary-side positive terminal and the secondary-side positive terminal, and a fourth impedance provided between the primary-side negative terminal and the secondary-side negative terminal, wherein the circuit constants of the third impedance and the fourth impedance are equal. This makes it possible to obtain an equivalent circuit of a single-phase magnetic coupling circuit that enables graph processing taking into account the characteristics of the magnetic coupling circuit. In particular, in graph processing in which Kirchhoff's law of conservation of current does not hold, processing taking into account the characteristics of the magnetic coupling circuit can be performed. Furthermore, according to the first embodiment, a circuit including magnetic coupling can be represented as an equivalent circuit that does not include magnetic coupling (connected by wiring). Furthermore, by converting this equivalent circuit into a graph structure, the magnetic coupling circuit can be expressed in a graph structure.

[0052] Also, let j be the complex number, ω be the angular frequency, and L be the inductance of the inductance element on the primary side of the magnetic coupling circuit. １ , the inductance of the inductance element on the secondary side of the magnetic coupling circuit is L ２ When the mutual inductance between the primary inductance element and the secondary inductance element is M, the circuit constant Z of the first impedance is １ , the circuit constant Z of the second impedance ２ , the circuit constant Z of the third impedance ３ , and the circuit constant Z of the fourth impedance ４ is calculated by the above formulas (4) to (6). In this way, in the first embodiment, it is possible to calculate the circuit constants that are electrically equivalent to the magnetic coupling circuit for each of the first to fourth impedances that constitute the equivalent circuit.

[0053] In addition, the inductance L of the inductance element on the primary side of the magnetic coupling circuit １ , the inductance L of the secondary inductance element of the magnetic coupling circuit ２ , and the mutual inductance M between the primary-side inductance element and the secondary-side inductance element can be calculated by the above equations (11) to (13). As a result, in the first embodiment, the circuit constants of the magnetic coupling circuit can be calculated using the circuit constants of each of the first to fourth impedances that make up the equivalent circuit. In other words, it is possible to inversely convert an equivalent circuit that does not include magnetic coupling into a circuit that includes magnetic coupling.

[0054] Embodiment 2. In embodiment 1, an equivalent circuit of a magnetic coupling circuit including a primary-side inductance, a secondary-side inductance, and a mutual inductance was described. In embodiment 2, an equivalent circuit will be described in consideration of parasitic components that arise due to the physical dimensions or structure of an actual magnetic coupling circuit, i.e., residual resistance, parasitic capacitance, and residual inductance.

[0055] Fig. 8 is a diagram showing an example of a single-phase magnetic coupling circuit according to embodiment 2. In this magnetic coupling circuit, impedance X is connected between the + terminal on the primary side and the + terminal on the secondary side, and between the - terminal on the primary side and the - terminal on the secondary side, in comparison with the magnetic coupling circuit shown in Fig. 1. Impedance X can be thought of as, for example, stray capacitance (also called parasitic capacitance) generated by spatial coupling, or a conductance component representing the degree of current leakage, or it can also be thought of as a capacitor with a physical entity provided between the primary side and the secondary side to reduce electromagnetic noise.

[0056] When impedance X is a parasitic capacitance or a capacitor, where C is the capacitance [F] of the parasitic capacitance or capacitor, impedance X is 1 / (j×ω×C). When impedance X is a conductance, where resistance is R [Ω], the circuit constant of impedance X can be R. Furthermore, for example, impedance X may be a parallel or series circuit that combines a capacitor and a conductance, and the circuit constants of these circuits may be any as long as they have circuit constants based on circuit theory.

[0057] When the impedance value of the impedance X is X, the inductance L in the magnetic coupling circuit shown in FIG. １ and L ２ , as well as the mutual inductance M and the first impedance Z in the equivalent circuit shown in FIG. １ , second impedance Z ２ , third impedance Z ３ , and a fourth impedance Z ４ The relationships are expressed by the following equations (14) to (16).

[0058] In the above equation (16), when the impedance value X of the impedance X is sufficiently large and can be considered as insulation, the third impedance Z ３ and the fourth impedance Z ４ is Z shown in the following equation (17) which does not take into account the impedance X shown in the first embodiment. ３ and Z ４ Matches the value of

[0059] Furthermore, since equation (16) can be transformed into the following equation (16'), the circuit components can be decomposed as shown in Fig. 9. This equation shows that it can be decomposed into the impedance X of the parasitic component and the impedance of equation (6) that does not include the parasitic component, and that the parasitic component between the primary side and the secondary side can be realized by simply forming an impedance circuit based on the parasitic component in parallel with the circuit constant of equation (6).

[0060] Furthermore, the equivalent circuit having the above impedance X can be inversely converted into the circuit constants of the magnetic coupling circuit shown in FIG. 8 using the following equations (18) to (20).

[0061] In the above equations (18) to (20), when the impedance value X of the impedance X is sufficiently large and can be regarded as insulation, the equations (18) to (20) become the same as the inverse transformation equations (14) to (16) shown in the first embodiment.

[0062] 10 is a diagram showing an example of a single-phase magnetic coupling circuit according to embodiment 2. This magnetic coupling circuit has a primary-side inductance L １ Impedance X connected in series with １ and the secondary inductance L ２ Impedance X connected in series with ２ and is provided.

[0063] Impedance X １ and X ２ is, for example, a residual resistance, and if the respective residual resistances are R1 [Ω] and R2 [Ω], then the impedance X １ and X ２ The circuit constants are R1 and R2, respectively. Impedance X １ and X ２ In addition to the residual resistance, the residual inductance may be used. In that case, the residual inductance is expressed as L_residual １ , L_residual ２ Then, the impedance X １ and X ２ The impedance values ​​of j×ω×L_residual are １ , j×ω×L_residual ２ In this case, the magnetic coupling circuit shown in FIG. 10 can be expressed as the equivalent circuit shown in FIG. 2. In this case, the circuit constants are the impedance X １ and X ２ The impedance values ​​of X １ , X ２ and the impedance value of the first impedance is Z １, the impedance value of the second impedance is Z ２ , the impedance value of the third impedance is Z ３ , the impedance value of the fourth impedance is Z ４ Then, it can be calculated by the following equations (21) to (23).

[0064] In these equations (21) to (23), the impedance X １ Impedance value X １ and impedance X ２ Impedance value X ２ is sufficiently small, equations (21) to (23) can be expressed as １ and X ２ These equations are the same as the equations (7) to (9) shown in the first embodiment.

[0065] Furthermore, the above impedance X １ and X ２ The equivalent circuit shown in FIG. 2 including the above can be inversely converted into the circuit constants of the magnetic coupling circuit shown in FIG. 10 by the following equations (27) to (29).

[0066] In this equation, the impedance X １ and X ２ When the impedance values ​​of the respective terminals are 0Ω, i.e., short-circuited, the equations (27) to (29) coincide with the following equations (30) to (32). The equations (30) to (32) are the same as the equations (11) to (13) shown in the first embodiment.

[0067] 11 is a diagram showing an example of a single-phase magnetic coupling circuit according to embodiment 2. This magnetic coupling circuit has a primary-side inductance L １ Impedance X connected in series with １ and the secondary inductance L ２ Impedance X connected in series with ２and an impedance X connected between the + terminal on the primary side and the + terminal on the secondary side, and between the - terminal on the primary side and the - terminal on the secondary side.

[0068] This circuit includes all of the parasitic components shown in FIGS. 8 and 10. In this circuit, the impedances X and X １ and X ２ The impedance values ​​of X and X are respectively １ , X ２ The impedance value of the first impedance in the equivalent circuit shown in FIG. １ , the impedance value of the second impedance is Z ２ , the impedance value of the third impedance is Z ３ , the impedance value of the fourth impedance is Z ４ Then, the relationships of the following equations (33) to (35) hold.

[0069] In these equations (33) to (35), the impedance X １ Impedance value X １ and impedance X ２ Impedance value X ２ is considered to be sufficiently small, each of the impedance values ​​from the first impedance to the fourth impedance will be the same as that of the circuit shown in FIG. 8, and if the impedance X is considered to be sufficiently large and in an insulating state, each of the impedances from the first impedance to the fourth impedance will be the same as the impedance of the circuit shown in FIG. 10.

[0070] Furthermore, the equivalent circuit shown in FIG. 2 can be expressed as the impedance X and the impedance X by the following equations (36) to (38). １ and impedance X ２ 11.

[0071] In these equations (36) to (38), when the impedance X is sufficiently large and can be considered as insulation, the inductance L １ and L ２, and the mutual inductance M are equal to the circuit constants in the inverse transformation of the circuit shown in FIG. 10, and the impedance X １ and impedance X ２ is small enough to be considered short-circuited, １ and L ２ , and the mutual inductance M are equal to the impedance at the time of inverse transformation in the circuit shown in FIG. 8. That is, the inductance L １ and L ２ , and the calculation formula for the mutual inductance M are the results that include the results of FIGS. 8 and 10 shown in the first and second embodiments.

[0072] As described above, even under conditions in which parasitic components occur as in an actual magnetic coupling circuit and the magnetic coupling circuit is not an ideal circuit having only magnetic coupling components, the magnetic coupling circuit can be represented by the equivalent circuit shown in Fig. 2. Furthermore, since the equivalent circuit shown in Fig. 2 has a structure that can be compared with the physical structure of an actual magnetic coupling circuit, it is possible to search for circuit constants that satisfy the design goal based on the equivalent circuit shown in Fig. 2, and then perform an inverse transformation from the equivalent circuit shown in Fig. 2 to the magnetic coupling circuit shown in Fig. 1, in which the relative permeability or coupling coefficient can be easily considered, thereby making it possible to compare the design goal with the physical structure. On the other hand, with the equivalent circuit based on the conventional example shown in Fig. 4, the physical structure does not match the structure of the equivalent circuit, making it difficult to consider the parasitic components described above. Therefore, the use of the method shown in this embodiment provides a special effect.

[0073] Therefore, according to the second embodiment, unlike conventional magnetic design, parameter search is not performed by considering the design target as an electromagnetic field simulation based on the finite element method or the like, which requires a huge amount of calculation, but rather by circuit simulation, which requires a calculation amount several orders of magnitude (several thousand to several hundred million times or more) smaller than that of electromagnetic field simulation, or by substitution calculation into algebraic equations, which requires an even smaller amount of calculation than that of circuit simulation, thereby achieving an unprecedented and exceptional effect. １ and L ２, and the mutual inductance M are mutually dependent, so the inductance L １ and L ２ It was necessary to calculate M generated in the physical structure by changing the parameters of the first impedance Z required in the design specifications. １ , second impedance Z ２ , third impedance Z ３ By determining the inductance L １ and L ２ , and the mutual inductance M are uniquely determined, which has the special effect of eliminating the need for parameter search.

[0074] As described above, according to the second embodiment, the complex number is j, the angular frequency is ω, and the inductance of the inductance element on the primary side of the magnetic coupling circuit is L １ , the inductance of the inductance element on the secondary side of the magnetic coupling circuit is L ２ , the mutual inductance between the primary inductance element and the secondary inductance element is M, the impedance connected between the positive terminal of the primary inductance element and the positive terminal of the secondary inductance element is X, and the impedance connected between the negative terminal of the primary inductance element and the negative terminal of the secondary inductance element is X, then the first impedance Z １ , second impedance Z ２ , third impedance Z ３ , and a fourth impedance Z ４ is calculated by the above formulas (14) to (16). As a result, in the second embodiment, in addition to the effect of the first embodiment, even when the magnetic coupling circuit includes a parasitic component, it is possible to calculate a circuit constant that is electrically equivalent to the magnetic coupling circuit for each of the first to fourth impedances that constitute the equivalent circuit.

[0075] In addition, the inductance L of the inductance element on the primary side of the magnetic coupling circuit １ , the inductance L of the secondary inductance element of the magnetic coupling circuit ２, and the mutual inductance M between the primary-side inductance element and the secondary-side inductance element can be calculated by the above equations (18) to (20). As a result, in the first embodiment, even if the magnetic coupling circuit includes a parasitic component, the circuit constants of the magnetic coupling circuit can be calculated using each of the first to fourth impedances that constitute the equivalent circuit.

[0076] Also, let j be the complex number, ω be the angular frequency, and L be the inductance of the inductance element on the primary side of the magnetic coupling circuit. １ , the inductance of the inductance element on the secondary side of the magnetic coupling circuit is L ２ , the mutual inductance between the primary inductance element and the secondary inductance element is M, the impedance value of the impedance connected between the + terminal and - terminal of the primary inductance element is X １ , the impedance value of the impedance connected between the positive and negative terminals of the secondary inductance element is X ２ When this is the case, the impedance value Z of the first impedance １ , the impedance value Z of the second impedance ２ , the impedance value Z of the third impedance ３ , and the impedance value Z of the fourth impedance ４ is calculated by the above formulas (21) to (23). As a result, in the second embodiment, even if the magnetic coupling circuit includes residual resistance or the like, it is possible to calculate the circuit constants that are electrically equivalent to the magnetic coupling circuit for each of the first to fourth impedances that constitute the equivalent circuit.

[0077] In addition, the inductance L of the inductance element on the primary side of the magnetic coupling circuit １ , the inductance L of the secondary inductance element of the magnetic coupling circuit ２, and the mutual inductance M between the primary-side inductance element and the secondary-side inductance element can be calculated by the above equations (27) to (29). As a result, in the second embodiment, even if the magnetic coupling circuit includes residual resistance or the like, it is possible to calculate the circuit constants of the magnetic coupling circuit using the impedance values ​​of the first to fourth impedances that constitute the equivalent circuit.

[0078] Also, let j be the complex number, ω be the angular frequency, and L be the inductance of the inductance element on the primary side of the magnetic coupling circuit. １ , the inductance of the inductance element on the secondary side of the magnetic coupling circuit is L ２ , the mutual inductance between the primary inductance element and the secondary inductance element is M, the impedance value of the impedance connected between the + terminal and - terminal of the primary inductance element is X １ , the impedance value of the impedance connected between the positive and negative terminals of the secondary inductance element is X ２ , the impedance value of the impedance connected between the positive terminal of the primary inductance element and the positive terminal of the secondary inductance element, and the impedance value of the impedance connected between the negative terminal of the primary inductance element and the negative terminal of the secondary inductance element are X, the circuit constant Z of the first impedance is １ , the impedance value Z of the second impedance ２ , the impedance value Z of the third impedance ３ , and the impedance value Z of the fourth impedance ４ is calculated by the above formulas (33) to (35). As a result, in the second embodiment, even if the magnetic coupling circuit includes parasitic components, residual resistance, and the like, it is possible to calculate circuit constants that are electrically equivalent to the magnetic coupling circuit for each of the first to fourth impedances that constitute the equivalent circuit.

[0079] In addition, the inductance L of the inductance element on the primary side of the magnetic coupling circuit １ , the inductance L of the secondary inductance element of the magnetic coupling circuit２ , and the mutual inductance M between the primary-side inductance element and the secondary-side inductance element can be calculated by the above equations (36) to (38). As a result, in the second embodiment, even if the magnetic coupling circuit includes parasitic components, residual resistance, and the like, it is possible to calculate the circuit constants of the magnetic coupling circuit using the impedance values ​​of the first to fourth impedances that constitute the equivalent circuit.

[0080] Embodiment 3 In the first and second embodiments, an equivalent circuit expressed using the first to fourth impedances has been described. In the third embodiment, an example of converting the above-described equivalent circuit into a graph structure will be described.

[0081] In the third embodiment, in the equivalent circuit expressed using the first impedance to the fourth impedance shown in the first and second embodiments, the first impedance is defined as the first node, the second impedance as the second node, the third impedance as the third node, and the fourth impedance as the fourth node, and the first node and the third node, the first node and the fourth node, the second node and the third node, and the second node and the fourth node are connected by edges, thereby converting the equivalent circuit into a graph structure.

[0082] FIG. 12 is a diagram showing an example of a graph structure according to the third embodiment. This graph structure has first to fourth nodes and edges connecting the nodes. From this graph structure, information about the graph (hereinafter referred to as "graph information") can be obtained. The graph information includes, for example, information about the nodes included in the graph structure and information about the edges included in the graph structure. The graph information may also include information indicating the connection relationships between the nodes (hereinafter referred to as "connection information") and attribute information of the nodes and edges.

[0083] The connection information can be held as text data based on the nodes, for example, as follows: First node: edge 1, edge 2 Second node: edge 3, edge 4 Third node: edge 1, edge 3 Fourth node: edge 2, edge 4

[0084] Furthermore, the connection information can also be held, for example, on an edge-by-edge basis as follows: Edge 1: First node, third node Edge 2: First node, fourth node Edge 3: Second node, third node Edge 4: Second node, fourth node

[0085] Furthermore, the connection information can be stored, for example, as an adjacency matrix as shown in FIG. 13 , or as a connection matrix or a combination of an order matrix and a graph Laplacian matrix. Regardless of the storage format, the connection information can be stored in any format because it is mutually inversely transformable. Furthermore, as shown in FIG. 9 , circuit components may be arranged in parallel with the third impedance and the fourth impedance. In this case, as shown in FIG. 14 , the third impedance (third node) may be decomposed into the third-1 impedance (third-1 node) and the third-2 impedance (third-2 node), and the fourth impedance (fourth node) into the fourth-1 impedance (fourth-1 node) and the fourth-2 impedance (fourth-2 node). When a circuit constant including impedance X is represented by a one-hot vector as shown in FIG. 12 , if impedance X is something other than inductance, the one-hot vector must take on non-zero values ​​for two or more elements because the impedance X includes both inductance components and other types of components, which poses a problem of being unable to be expressed by a one-hot vector. On the other hand, by expressing it as shown in Figure 14, it can be separated into impedance X and inductance components, which has the advantage that it can be described using only one-hot vectors, and this has the advantage that information on the type of nodes between nodes can be separated when processing with a graph neural network.However, there is a disadvantage that as the number of nodes increases, the number of edges to the outside also increases, so you can choose whether to add node 3-2 and node 4-2 depending on the purpose and required accuracy.

[0086] In particular, an adjacency matrix such as that shown in Figure 13 is desirable because it is appropriate as an input format for a graph neural network. The adjacency matrix is ​​a square matrix of (number of nodes) x (number of nodes), where an element is 1 when a connection exists from one node to another node, and 0 when no connection exists. In addition, in the adjacency matrix, for a node where a self-loop exists, the diagonal element of the node is set to 1, and in a graph where no self-loop exists, all diagonal elements are set to 0.

[0087] By converting the equivalent circuit into a graph structure using this method, it becomes possible to process the magnetic coupling circuit using a graph network or graph neural network (graph processing). Furthermore, by inputting the graph information obtained from the graph structure into the graph neural network, feature quantities of the magnetic coupling circuit can be extracted. A feature is a group of values ​​that serve as clues for predicting the correct label of the input training or test data. These feature quantities can be acquired within the framework of a graph neural network. Classification problems can be solved by inputting these feature quantities into a fully connected layer, outputting the same number of values ​​as the number of classes, and applying an activation function used in classification, such as a log softmax function or a softmax function, immediately before the output layer. Regression problems can be solved by inputting these feature quantities into a fully connected layer and outputting a single numerical value. In addition to classification and regression problems with correct labels, it is also possible to combine a graph neural network that extracts feature quantities, such as an autoencoder, with deep learning that restores input data from feature quantities. Self-supervised learning, which involves masking node or edge attributes and predicting the hidden values, is also possible. In this way, features are a group of numbers that abstractly represent the characteristics of input data obtained by applying a nonlinear function to the input data within a deep learning framework. Although the features themselves cannot be understood by humans, by learning them in combination with a loss function that outputs the difference from the correct answer, it is possible to capture the characteristics of the input data from the data.

[0088] In addition, the extracted features can be used to realize many functions, such as predicting the output voltage between the terminals of a magnetic coupling circuit or the current value flowing in the wiring, predicting the frequency characteristics of the voltage or current, predicting the cost of the magnetic coupling circuit, predicting the amount of electromagnetic noise generated, predicting the heat generation of coils or transformers, capacitors, semiconductors, etc., optimizing circuit constants, selecting optimal circuit components including semiconductors, and optimizing the circuit structure.

[0089] In this case, each of the above functions can be realized by providing the graph neural network with graph information obtained by converting the equivalent circuit into a graph structure as described in the third embodiment, and classification (e.g., selecting an integer between 1 and 10), regression data (e.g., calculating a real value between 0 and 1), or waveform (e.g., including regression data of length N as time information) that serves as teacher data for the graph information.

[0090] For example, predicting the output signal between the terminals of a magnetic coupling circuit is waveform prediction, predicting the cost of a magnetic coupling circuit is regression prediction, predicting the amount of electromagnetic noise generated in a magnetic coupling circuit is waveform prediction or a regression problem, predicting the amount of heat generated by a semiconductor in a magnetic coupling circuit is regression prediction, optimizing the circuit constants of a magnetic coupling circuit is regression prediction, predicting the number of layers of a semiconductor or printed circuit board in a magnetic coupling circuit is classification prediction, selecting circuit components including semiconductors in a magnetic coupling circuit is classification prediction, optimizing the circuit structure of a magnetic coupling circuit is a binary classification problem of whether wiring is present or absent, etc. Each of the above functions can be realized using existing graph neural network techniques.

[0091] Furthermore, by applying an autoencoder, a variational autoencoder, a generative adversarial network (GAN), or the like to a graph neural network, a generative model of a magnetic coupling circuit can be constructed.

[0092] 15 is a diagram showing an example of a graph structure according to embodiment 3. In this graph structure, the + terminal on the primary side of the equivalent circuit shown in FIG. 2 is designated as a first terminal node, the − terminal on the primary side is designated as a second terminal node, the + terminal on the secondary side is designated as a third terminal node, and the − terminal on the secondary side is designated as a fourth terminal node. Edges are then connected between the first terminal node and the first node, between the first terminal node and the third node, between the second terminal node and the first node, between the second terminal node and the fourth node, between the third terminal node and the second node, between the third terminal node and the third node, between the fourth terminal node and the second node, and between the fourth terminal node and the fourth node.

[0093] The connection information in this graph structure can be held as text data centered on nodes and terminal nodes, for example, as follows: First terminal node: Edge 5, Edge 6 Second terminal node: Edge 7, Edge 8 Third terminal node: Edge 9, Edge 10 Fourth terminal node: Edge 11, Edge 12 First node: Edge 1, Edge 2, Edge 6, Edge 7 Second node: Edge 3, Edge 4, Edge 10, Edge 11 Third node: Edge 1, Edge 3, Edge 5, Edge 9 Fourth node: Edge 2, Edge 4, Edge 8, Edge 12

[0094] As described above, the connectivity information may be stored in any form as long as it is reversible and can be converted, such as by storing the connectivity information based on edges or as an adjacency matrix.

[0095] In FIG. 15 , an arbitrary circuit is connected between the first terminal node and the second terminal node, and an arbitrary circuit is also connected between the third terminal node and the fourth terminal node. For example, an input power supply is connected between the first terminal node and the second terminal node, and a load circuit is connected between the third terminal node and the fourth terminal node. Alternatively, for example, an input circuit and a load circuit may be connected between the first terminal node and the second terminal node and between the third terminal node and the fourth terminal node, and a bidirectional circuit may be configured in which the input circuit and the load circuit can be selectively switched by switching a semiconductor or other switch, like a regenerative brake. Note that the first terminal node, the second terminal node, the third terminal node, and the fourth terminal node can be replaced with nodes that are circuit components. For example, the first terminal node may be replaced with the positive side of the power supply, and the second terminal node may be replaced with the negative side of the power supply. In this case, the first terminal node and the second terminal node become a single node. Alternatively, a power supply node may be disposed adjacent to the first terminal node and the second terminal node. In this case, the first terminal node and the second terminal node become separate nodes. Furthermore, each terminal node may be divided into two or more nodes. For example, when a resistor and a capacitor are attached to one terminal of the magnetic coupling circuit (here, for simplicity, it is assumed to be the first terminal node), this can be realized by defining edges that connect the resistor and the first node, the resistor and the third node, the capacitor and the first node, the capacitor and the third node, and the resistor and the capacitor, respectively.

[0096] As shown in Figure 16, it is also desirable to assign the above-mentioned attribute information to each node from the first node to the fourth node. Attribute information, including numerical values, can be freely assigned to each node. However, when processing each node as a single graph, it is desirable to standardize the size of the attribute information matrix (the number of rows and columns in the two-dimensional case) and the format (order and size) of the information entered into each matrix element, since this allows the entire graph to be processed collectively, independent of the node. The order refers to, for example, entering resistance components in the first column, inductance components in the second column, and capacitance components in the third column. The size refers to, for example, entering resistance in the order of Ω, inductance in the order of μH, and capacitance in the order of pF.

[0097] In particular, by making the number of elements in the attribute information matrix equal, processing by a graph neural network, which is one type of graph processing, becomes possible. As shown in Figure 16, when the number of elements in the attribute information is three, in addition to connection information between nodes and edges such as: First node: edge 1, edge 2 Second node: edge 3, edge 4 Third node: edge 1, edge 3 Fourth node: edge 2, edge 4, attribute information such as: First node: attribute information 1, attribute information 2, attribute information 3 Second node: attribute information 4, attribute information 5, attribute information 6 Third node: attribute information 7, attribute information 8, attribute information 9 Fourth node: attribute information 10, attribute information 11, attribute information 12 By assigning attribute information to each node, the connection information and attribute information can be stored as text data.

[0098] Furthermore, for example, by storing connection information between nodes and edges as an adjacency matrix, assigning a unique number to each node, and assigning one column of a matrix for each node in the row corresponding to the unique number, attribute information can be stored as tabular data. In particular, in the third embodiment, the first impedance or the second impedance may take a negative value. Furthermore, particularly in graph neural networks, activation functions such as the ReLU function, Sigmoid function, Tanh function, or Softmax function are likely to respond to real numbers between 0 and 1 or integer signals between 0 and 1, so it is desirable to normalize the attribute information to a range between 0 and 1. In this case, taking into account negative circuit constants that occur only under special conditions, such as negative resistance or negative inductance, there is a problem in that the dynamic range of the attribute information when normalized is limited to half the positive value.

[0099] For example, when "10 μH" and "20 μH" are normalized to real numbers between 0 and 1, they become "0.10" and "0.11", but when normalized taking into account negative components, they become "0.100" and "0.105", and the difference between the two becomes smaller, making it more susceptible to information degradation due to rounding errors in processing such as graph neural networks.

[0100] Therefore, by expressing the attribute information corresponding to the circuit constant of the first impedance and the attribute information corresponding to the circuit constant of the second impedance using a first element indicating a positive or negative sign and a second element indicating the absolute value of each impedance, it is possible to prevent a decrease in the accuracy of the attribute information. In particular, there exists a condition in which the circuit constants of the first impedance and the second impedance are negative real numbers. In this case, the first impedance and the second impedance are special circuit components that facilitate the flow of electrical signals, as opposed to circuit components that generally impede the flow of electrical signals (currents). However, like the attribute information corresponding to the circuit constants of the first impedance and the second impedance, the attribute information corresponding to the circuit constants of the third impedance and the fourth impedance may also be expressed using a first element indicating a positive or negative sign and a second element indicating the absolute value of each impedance.

[0101] For example, as shown in FIG. 17, if the circuit constant corresponding to the first node is "10 μH", the circuit constant corresponding to the second node is "-3 μH", the circuit constant corresponding to the third node is "1 μH", and the circuit constant corresponding to the fourth node is "1 μH", the circuit constant corresponding to the second node will be a negative real number. Therefore, the first element of the attribute information for the second node is set to "1". On the other hand, since the circuit constants corresponding to the first node, third node, and fourth node are positive real numbers, the first element of the attribute information for each node is set to "0". Furthermore, by inputting the absolute value of the circuit constant for each node into the second element of the attribute information, the attribute information for each node will be, for example, as follows: First node: 0, 10 μH Second node: 1, 3 μH Third node: 0, 1 μH Fourth node: 0, 1 μH

[0102] As a result, negative elements are not included in the attribute information, preventing a decrease in the amount of information in the attribute information due to a reduced dynamic range and suppressing a decrease in processing accuracy in the graph neural network. Note that, for example, in circuit simulations, when the coupling coefficient k is set to exactly 1, the attribute information may contain "0." In this case, since "0" is neither positive nor negative, the first element representing the sign can be either "0" or "1," but a positive value is preferable. This is because information indicating a negative circuit constant occurs under special conditions, and by making the circuit constant negative when such information occurs under special conditions, it can be processed by a graph neural network or the like as information with a large amount of information, such as information indicating a special component. However, in actual magnetic coupling circuits, the coupling coefficient k never becomes "1." Even in the equations described in the first and second embodiments, a coupling coefficient of "1" diverges. Therefore, by inputting a component as close to "1" as possible, such as "0.9999," the circuit constant is converted to a value greater than "0," which is desirable and prevents the problem of a decrease in the amount of information in the attribute information.

[0103] Furthermore, in the third embodiment, an element with a positive or negative sign is input as the first element, and the absolute value of the impedance is input as the second element. However, the input order may be any, for example, the absolute value of the impedance may be input as the first element, and the element with a positive or negative sign may be input as the second element. Also, while Fig. 17 shows an example in which the attribute information is made up of elements of two matrices, the number of elements in the attribute information is not important. In particular, since multiple types of components are used in a magnetic coupling circuit, when separating the types using a one-hot vector, it is desirable that the number of elements in the attribute information be equal to or greater than the number of component types.

[0104] Furthermore, if the number of elements of the attribute information of all nodes in the magnetic coupling circuit is equal, it is desirable because the graph network processing or graph neural network processing can be simply realized without requiring pre-processing or post-processing. Note that, in addition to coils, the types of components may be passive components such as capacitors, resistors, diodes, or antennas, as well as active components that are semiconductors such as power supply ICs or transistors.

[0105] Furthermore, the attribute information may be anything other than the type of part, such as the model number of the part, the manufacturer of the circuit part, etc., as long as it can be classified into a finite number of classes. Furthermore, even for elements with real components, such as drive voltage or drive frequency, which cannot be separated into a finite number of analog data (continuous quantity data), it is possible to make the data separable into a finite number of parts by determining the division width and assigning values, such as 0V to 0.3V as "1" and 0.3V to 0.5V as "2."

[0106] Although not described in the third embodiment, if the direction of current in the magnetic coupling circuit is known, the graph structure may be a directed graph. Furthermore, not only nodes but also edges may have attribute information. For example, by providing edges with current frequency characteristics, it is possible to solve a graph neural network in which signals that differ for each frequency are exchanged between adjacent nodes. Furthermore, if the circuit constants shown in the third embodiment can be assigned to the attribute information of the circuit components, the circuit components and wiring may be processed as nodes, or the wiring may be processed as nodes and the circuit components as edges.

[0107] Furthermore, in the above explanation, an example was described in which attribute information is assigned to each node from the first node to the fourth node in the graph structure shown in FIG. 12 , but in the graph structure shown in FIG. 15 , attribute information may also be assigned to each node from the first node to the fourth node and each terminal node from the first terminal node to the fourth terminal node.

[0108] In this case, the additive average of the attribute information of the first node and the attribute information of the third node may be assigned to the attribute information of the first terminal node, the additive average of the attribute information of the first node and the attribute information of the fourth node may be assigned to the attribute information of the second terminal node, the additive average of the attribute information of the second node and the attribute information of the third node may be assigned to the attribute information of the third terminal node, and the additive average of the attribute information of the second node and the attribute information of the fourth node may be assigned to the attribute information of the fourth terminal node.

[0109] That is, in a graph neural network, one hidden layer corresponds to applying a weight matrix acquired through learning to the attribute information of adjacent nodes and embedding the weight matrix in the attribute information of its own node. In this case, by initially assigning the attribute information of nodes adjacent to the first to fourth terminal nodes to the first to fourth terminal nodes, the weight matrix converges faster, processing divergence is suppressed, and processing results can be prevented from falling into a local minimum. This can reduce calculation time and improve calculation accuracy. Furthermore, if the data set is small or there is no limit to the calculation time, the attribute information of each terminal node from the first to fourth terminal nodes may be set to "0" or an element indicating that the node is a terminal node may be assigned as a one-hot vector. This has the effect of enabling learning with less user preconceptions. Furthermore, because the first terminal node to the fourth terminal node and the first node to the fourth node are different types of nodes, it is desirable for the position of 0 in the matrix (one-hot vector) representing each node attribute to differ by at least one, as this prevents the node attributes from mixing in the graph neural network and allows them to be kept separable. Although the position of 0 is considered to be different because learning can also be performed by inputting circuit constants into elements of 1 in the matrix, when the matrix is ​​composed only of 0s or 1s (general one-hot vectors), a difference in the position of 0 in at least one is synonymous with a difference in the position of 1.

[0110] [Evaluation] Embodiment 3 shows a method for reversibly converting a magnetic coupling circuit into a graph structure and, conversely, converting from the graph structure to a magnetic coupling circuit. Circuit information of the magnetic coupling circuit is preferably provided as text data including a netlist obtained based on an equivalent circuit of the magnetic coupling circuit.

[0111] The accuracy of the conversion from a magnetically coupled circuit to a graph structure can be determined by converting a netlist to a graph structure, then converting the converted graph structure back to a netlist, and restoring the original magnetically coupled circuit. However, because the conversion from a netlist to a graph structure and the conversion from a graph structure to a netlist cannot be performed reversibly, a separate algorithm must be developed. Therefore, it is not possible to evaluate the information degradation of the conversion alone. Another possible method for evaluating the generated netlist is to input the netlist into a circuit simulation and use the results of the circuit simulation. However, this is not desirable because there is no guarantee that the graph structure can be converted back to a netlist that can be used for circuit simulation without error, even with a small amount of information loss.

[0112] For this reason, the inventors of the present application have devised a method in which, instead of performing a reverse conversion from the graph structure to a netlist, graph information obtained from the generated graph structure is input into a graph neural network, and the accuracy of the conversion from the magnetic coupling circuit to the graph structure is confirmed based on the inference accuracy of the graph neural network. The graph information input to the graph neural network is, for example, information about the nodes and edges included in the graph structure, and this information may include the connection information and attribute information described above.

[0113] If there is little information degradation during the conversion from the magnetic coupling circuit to a graph structure, the information necessary for inference by the graph neural network is more likely to remain, resulting in higher inference accuracy. Conversely, if there is significant information degradation during the conversion, it is expected that the inference accuracy of the graph neural network will be lower. To compare the third embodiment with the conventional example under fair conditions, the structure of the graph neural network (the number of hidden layers or the number of channels in each hidden layer), the number of epochs, the number of mini-batches, the optimization method, etc. were not changed, except for the addition of nodes to represent the third and fourth impedances in the equivalent circuit of the magnetic coupling circuit. Furthermore, the training data and test data were fixed. Furthermore, since the graph neural network has variability in training depending on the initial value of the random number, the random number was fixed, and training was performed 10 times using 10 pieces of training data and test data divided in advance, and the arithmetic average of the inference accuracy for the test data was taken to reduce the variance in training. The data set used was 3,308 circuits included with LTspice from Analog Devices, and of the 3,308 circuits, 2,315, or 70%, were used as training data and 993 were used as test data, with the training data and test data being fixed.

[0114] Then, as shown in FIG. 18, a classification problem was generated using seven types of circuit components. FIG. 18 shows the number of circuits, the average number of nodes, the average number of edges, and the average value of node types included in the netlist used as the dataset. In FIG. 18, power products are the most common type of circuit components, accounting for 70% of the total. The training data and test data were randomly divided to ensure equal numbers to prevent bias.

[0115] [Graph Neural Network] The graph neural network was constructed as a neural network with three hidden layers. GraphSage, which achieved the highest inference accuracy under all conditions among existing algorithms, was used as the algorithm for reducing information between nodes to extract the features of the magnetic coupling circuit. A seven-value classification was performed using a three-layer neural network consisting of GraphSage and the ReLU function as an activation function, and the softmax function as an activation function before the output layer.

[0116] [Experimental Results] For comparison, Fig. 19 shows the results of learning and inference as a conventional example, assuming that mutual inductance M is not taken into account and that there is no mutual inductance M. Learning was repeated 4,000 times (epochs), and the average value of 10 trials at the epoch at which inference accuracy was highest is shown. In Fig. 19, the maximum inference accuracy using epochs as a parameter was 97.20%.

[0117] 20 shows the inference accuracy when training data and test data are created using the graph structure conversion method described in embodiment 3, and the test data is used to infer the types of circuit components. As a result, the maximum inference accuracy was 97.41%, which is 0.21% higher than the conventional example shown in FIG.

[0118] Of the 3,308 sample circuits included in LTspice, 105 were single-phase magnetically coupled circuits, accounting for 3.17% of the total. The coupling coefficients in the sample circuits included a value of "1." However, because the formulas described in the first and second embodiments cannot account for "1," the value for "1" was set to "0.999." Thus, the sample circuits described above include a relatively small number of target circuits, and when inferring circuits without mutual inductance, circuits without mutual inductance are also used during training. Therefore, the effect of considering mutual inductance in the equivalent circuit appears small. However, inference accuracy improved even when changing the combination of training data and test data, changing the graph neural network algorithm, changing the activation function or batch normalization function, or training with reduced input data. This suggests that information degradation was reduced when converting to a graph neural network.

[0119] 21 shows, for comparison, the results when the circuit constants of the third impedance and the fourth impedance are not divided equally, but the circuit constant of the third impedance is doubled and the circuit constant of the fourth impedance is set to "0" as described in embodiment 1. In this case, the fourth impedance is short-circuited, and therefore a decrease in inference accuracy is expected in a graph neural network in which Kirchhoff's law of conservation of current does not hold. However, when the training data and test data are interchanged, the maximum value of inference accuracy is 97.20%, which is certainly lower than the result shown in FIG. 20 using the proposed method according to embodiment 3.

[0120] Furthermore, this result has the same inference accuracy as the result (FIG. 19) when the proposed method according to the third embodiment is not used, i.e., when the mutual inductance (coupling coefficient) is considered to be "0." From this, it can be seen that the proposed method according to the third embodiment can reduce information degradation when processing a graph structure converted from an equivalent circuit using a graph network or graph neural network.

[0121] As described above, the conversion method according to the third embodiment is a method for converting the equivalent circuit of the above-described magnetic coupling circuit into a graph structure, and is characterized in that the first impedance of the equivalent circuit is defined as the first node, the second impedance of the equivalent circuit is defined as the second node, the third impedance of the equivalent circuit is defined as the third node, and the fourth impedance of the equivalent circuit is defined as the fourth node, and the first and third nodes, the first and fourth nodes, the second and third nodes, and the second and fourth nodes are connected by edges, respectively. In this way, the conversion method according to the third embodiment can convert the equivalent circuit of the above-described magnetic coupling circuit into a graph structure expressed by a combination of nodes and edges.

[0122] Furthermore, the + terminal on the primary side of the equivalent circuit is defined as a first terminal node, the - terminal on the primary side of the equivalent circuit is defined as a second terminal node, the + terminal on the secondary side of the equivalent circuit is defined as a third terminal node, and the - terminal on the secondary side of the equivalent circuit is defined as a fourth terminal node, and edges are respectively connected between the first terminal node and the first node, between the first terminal node and the third node, between the second terminal node and the first node, between the second terminal node and the fourth node, between the third terminal node and the second node, between the third terminal node and the third node, between the fourth terminal node and the second node, and between the fourth terminal node and the fourth node. As a result, the conversion method according to the third embodiment can convert the equivalent circuit of the above-mentioned magnetic coupling circuit into a graph structure expressed by a combination of nodes, edges, and terminal nodes.

[0123] Furthermore, attribute information is assigned to each of the first to fourth nodes, and a circuit constant of a first impedance is assigned to the attribute information of the first node, a circuit constant of a second impedance is assigned to the attribute information of the second node, a circuit constant of a third impedance is assigned to the attribute information of the third node, and a circuit constant of a fourth impedance is assigned to the attribute information of the fourth node. As a result, the conversion method according to the third embodiment can convert the equivalent circuit of the magnetic coupling circuit described above into a graph structure expressed by a combination of nodes, edges, and attribute information of the nodes.

[0124] Furthermore, the attribute information of the first node has a first element representing the positive / negative sign of the circuit constant of the first impedance and a second element representing the absolute value of the circuit constant of the first impedance, and the attribute information of the second node has a first element representing the positive / negative sign of the circuit constant of the second impedance and a second element representing the absolute value of the circuit constant of the second impedance. As a result, in the third embodiment, it is possible to suppress information degradation when the attribute information of the first node and the attribute information of the second node are normalized.

[0125] Furthermore, attribute information is assigned to each terminal node from the first terminal node to the fourth terminal node and each node from the first node to the fourth node, and the arithmetic average of the attribute information of the first node and the attribute information of the third node is assigned to the attribute information of the first terminal node, the arithmetic average of the attribute information of the first node and the attribute information of the fourth node is assigned to the attribute information of the second terminal node, the arithmetic average of the attribute information of the second node and the attribute information of the third node is assigned to the attribute information of the third terminal node, and the arithmetic average of the attribute information of the second node and the attribute information of the fourth node is assigned to the attribute information of the fourth terminal node. As a result, in the third embodiment, when the converted graph structure is processed by the graph neural network, the convergence of the weight matrix is ​​accelerated, divergence of the processing is suppressed, and the processing result can be prevented from falling into a minimum value.

[0126] Furthermore, attribute information is assigned to each terminal node from the first terminal node to the fourth terminal node and each node from the first node to the fourth node by a one-hot vector, and the positions of 0 in each piece of attribute information from the first terminal node to the fourth terminal node are equal, the positions of 0 in each piece of attribute information from the first node to the fourth node are equal, and the positions of 0 in each piece of attribute information from the first terminal node to the fourth terminal node differ from the positions of 0 in each piece of attribute information from the first node to the fourth node in at least one location. As a result, in the third embodiment, the node attributes of each terminal node and each node are not mixed within the graph neural network, and the two can be kept separable.

[0127] Furthermore, a feature extraction method according to a third embodiment is a method for extracting feature quantities of a single-phase magnetically coupled circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, and extracts feature quantities of the magnetically coupled circuit by using information indicating nodes and information indicating edges obtained from the graph structure obtained by the above-described conversion method as input information for a graph neural network. As a result, the feature extraction method according to the third embodiment can extract feature quantities of the magnetically coupled circuit by using the graph structure converted from the equivalent circuit of the single-phase magnetically coupled circuit.

[0128] Furthermore, a feature extraction method according to a third embodiment is a method for extracting feature quantities of a single-phase magnetically coupled circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, and extracts feature quantities of the magnetically coupled circuit by using, as input information to a graph neural network, information indicating nodes, information indicating edges, and node attribute information obtained from the graph structure obtained by the above-described conversion method. As a result, the feature extraction method according to the third embodiment can extract feature quantities of the magnetically coupled circuit by using the graph structure converted from the equivalent circuit of the single-phase magnetically coupled circuit.

[0129] In addition, the present disclosure allows for free combination of the embodiments, modification of any of the components of the embodiments, or omission of any of the components of the embodiments.

[0130] The present disclosure provides an equivalent circuit of a single-phase magnetically coupled circuit, which makes it possible to obtain an equivalent circuit that enables graph processing based on the characteristics of the magnetically coupled circuit, and is suitable for use in an equivalent circuit, a conversion method, and a feature extraction method for a magnetically coupled circuit.

[0131] L １ , L ２ , L ３ , L ４ , L ５ , L ６ , L ７ , L ８ , L ９ Inductance, M Mutual inductance, N１ , N ２ Conductor, v1 voltage between the terminals of the primary inductance element, v2 voltage between the terminals of the secondary inductance element, X, X １ , X ２ Impedance.

Claims

1. An equivalent circuit of a single-phase magnetically coupled circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, a first impedance provided between the positive terminal of the primary side and the negative terminal of the primary side; a second impedance provided between the positive terminal of the secondary side and the negative terminal of the secondary side; a third impedance provided between the positive terminal of the primary side and the positive terminal of the secondary side; a fourth impedance provided between the negative terminal of the primary side and the negative terminal of the secondary side; An equivalent circuit of a magnetic coupling circuit, wherein the circuit constants of the third impedance and the fourth impedance are equal.

2. Let j be a complex number, Let ω be the angular frequency. The inductance of the inductance element on the primary side of the magnetic coupling circuit is L １ , The inductance of the inductance element on the secondary side of the magnetic coupling circuit is L ２ , When the mutual inductance between the primary inductance element and the secondary inductance element is M, The first impedance Z １ , the second impedance Z ２ , the third impedance Z ３ , and the fourth impedance Z ４ is calculated by the following formulas (4) to (6):

2. An equivalent circuit of the magnetic coupling circuit according to claim 1.

3. The inductance L of the inductance element on the primary side of the magnetic coupling circuit １ , The inductance L of the secondary inductance element of the magnetic coupling circuit ２ , and The mutual inductance M between the primary inductance element and the secondary inductance element can be calculated by the following equations (11) to (13):

3. An equivalent circuit of the magnetic coupling circuit according to claim 2.

4. Let j be a complex number, Let ω be the angular frequency. The inductance of the inductance element on the primary side of the magnetic coupling circuit is L １ , The inductance of the inductance element on the secondary side of the magnetic coupling circuit is L ２ , The mutual inductance between the primary inductance element and the secondary inductance element is M, an impedance connected between the positive terminal of the primary inductance element and the positive terminal of the secondary inductance element; and When the impedance connected between the negative terminal of the primary inductance element and the negative terminal of the secondary inductance element is X, The first impedance Z １ , the second impedance Z ２ , the third impedance Z ３ , and the fourth impedance Z ４ is calculated by the following formulas (14) to (16):

2. An equivalent circuit of the magnetic coupling circuit according to claim 1.

5. The inductance L of the inductance element on the primary side of the magnetic coupling circuit １ , The inductance L of the secondary inductance element of the magnetic coupling circuit ２ , and The mutual inductance M between the primary inductance element and the secondary inductance element can be calculated by the following equations (18) to (20):

5. An equivalent circuit of the magnetic coupling circuit according to claim 4.

6. Let j be a complex number, Let ω be the angular frequency. The inductance of the inductance element on the primary side of the magnetic coupling circuit is L １ , The inductance of the inductance element on the secondary side of the magnetic coupling circuit is L ２ , The mutual inductance between the primary inductance element and the secondary inductance element is M, The impedance connected between the + terminal and - terminal of the primary side inductance element is X １ , The impedance connected between the + and - terminals of the secondary inductance element is X ２ When The first impedance Z １ , the second impedance Z ２ , the third impedance Z ３ , and the fourth impedance Z ４ is calculated by the following equations (21) to (23):

2. An equivalent circuit of the magnetic coupling circuit according to claim 1.

7. The inductance L of the inductance element on the primary side of the magnetic coupling circuit １ , The inductance L of the secondary inductance element of the magnetic coupling circuit ２ , and The mutual inductance M between the primary inductance element and the secondary inductance element can be calculated by the following equations (27) to (29):

7. An equivalent circuit of the magnetic coupling circuit according to claim 6.

8. Let j be a complex number, Let ω be the angular frequency. The inductance of the inductance element on the primary side of the magnetic coupling circuit is L １ , The inductance of the inductance element on the secondary side of the magnetic coupling circuit is L ２ , The mutual inductance between the primary inductance element and the secondary inductance element is M, The impedance connected between the + terminal and - terminal of the primary side inductance element is X １ , The impedance connected between the + and - terminals of the secondary inductance element is X ２ , an impedance connected between the positive terminal of the primary inductance element and the positive terminal of the secondary inductance element; and When the impedance connected between the negative terminal of the primary inductance element and the negative terminal of the secondary inductance element is X, The first impedance Z １ , the second impedance Z ２ , the third impedance Z ３ , and the fourth impedance Z ４ is calculated by the following equations (33) to (35):

2. An equivalent circuit of the magnetic coupling circuit according to claim 1.

9. The inductance L of the inductance element on the primary side of the magnetic coupling circuit １ , The inductance L of the secondary inductance element of the magnetic coupling circuit ２ , and The mutual inductance M between the primary inductance element and the secondary inductance element can be calculated by the following equations (36) to (38):

9. An equivalent circuit of the magnetic coupling circuit according to claim 8.

10. 2. A method for converting an equivalent circuit of a magnetic coupling circuit according to claim 1 into a graph structure, comprising: The first impedance of the equivalent circuit is a first node, The second impedance of the equivalent circuit is a second node, The third impedance of the equivalent circuit is a third node, A fourth impedance of the equivalent circuit is defined as a fourth node, Between the first node and the third node, Between the first node and the fourth node, between the second node and the third node, and The second node and the fourth node are connected by edges. A conversion method characterized by:

11. The positive terminal on the primary side of the equivalent circuit is the first terminal node, The negative terminal on the primary side of the equivalent circuit is the second terminal node, The positive terminal on the secondary side of the equivalent circuit is the third terminal node, The negative terminal on the secondary side of the equivalent circuit is set as a fourth terminal node, Between the first terminal node and the first node, Between the first terminal node and the third node, Between the second terminal node and the first node, Between the second terminal node and the fourth node, Between the third terminal node and the second node, Between the third terminal node and the third node, Between the fourth terminal node and the second node, and The fourth terminal node and the fourth node are connected by edges.

11. The method of claim 10.

12. assigning attribute information to each of the first to fourth nodes; assigning a circuit constant of the first impedance to attribute information of the first node; assigning a circuit constant of the second impedance to attribute information of the second node; assigning a circuit constant of the third impedance to attribute information of the third node; The circuit constant of the fourth impedance is assigned to the attribute information of the fourth node.

11. The method of claim 10.

13. The attribute information of the first node is a first element representing the positive or negative sign of the circuit constant of the first impedance; a second element representing an absolute value of a circuit constant of the first impedance; The attribute information of the second node is a first element representing the positive or negative sign of the circuit constant of the second impedance; a second element representing the absolute value of the circuit constant of the second impedance; 13. The method of claim 12.

14. Attribute information is assigned to each terminal node from the first terminal node to the fourth terminal node and each node from the first node to the fourth node, assigning an arithmetic average of the attribute information of the first node and the attribute information of the third node to the attribute information of the first terminal node; assigning an arithmetic average of the attribute information of the first node and the attribute information of the fourth node to the attribute information of the second terminal node; assigning an arithmetic average of the attribute information of the second node and the attribute information of the third node to the attribute information of the third terminal node; An arithmetic mean of the attribute information of the second node and the attribute information of the fourth node is assigned to the attribute information of the fourth terminal node.

12. The method of claim 11.

15. Attribute information is assigned to each terminal node from the first terminal node to the fourth terminal node and each node from the first node to the fourth node by a one-hot vector, The positions of 0 in the attribute information from the first terminal node to the fourth terminal node are equal, The positions of 0 in the attribute information of each of the first node to the fourth node are equal, The position of 0 in each piece of attribute information from the first terminal node to the fourth terminal node is different from the position of 0 in each piece of attribute information from the first node to the fourth node in at least one position.

12. The method of claim 11.

16. A method for extracting a feature quantity of a single-phase magnetically coupled circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, comprising: The information indicating the nodes and the information indicating the edges obtained from the graph structure obtained by the conversion method according to claim 10 is used as input information for a graph neural network, thereby extracting the feature quantity of the magnetic coupling circuit. A feature extraction method characterized by:

17. A method for extracting a feature quantity of a single-phase magnetically coupled circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, comprising: The information indicating the nodes, the information indicating the edges, and the attribute information of the nodes obtained from the graph structure obtained by the conversion method according to claim 12 are used as input information for a graph neural network, thereby extracting the feature quantity of the magnetic coupling circuit. A feature extraction method characterized by:

18. A method for manufacturing an equivalent circuit of the magnetic coupling circuit of claim 1, comprising: placing circuit elements having the first impedance, the second impedance, the third impedance, and the fourth impedance; a process of arranging the circuit constants of the third impedance and the fourth impedance to be equal; A method for manufacturing an equivalent circuit of a magnetic coupling circuit, comprising:

19. A method for manufacturing a system for performing the conversion method of claim 10, comprising: a process for generating each of the first to fourth nodes; Between the first node and the third node, Between the first node and the fourth node, between the second node and the third node, and generating an edge connecting the second node and the fourth node; A method for manufacturing a system comprising:

20. A method for manufacturing an apparatus for extracting features of a single-phase magnetically coupled circuit in which a primary-side inductance element and a secondary-side inductance element are magnetically coupled by mutual inductance, by using a graph structure of a circuit including an equivalent circuit of the magnetically coupled circuit as input information for a graph neural network, comprising: a process for constructing the graph neural network; inputting the graph structure, including node and edge information, into the graph neural network; a process of configuring an output unit that outputs a feature quantity of the magnetic coupling circuit; Including, The graph structure is The graph structure is obtained by generating nodes corresponding to at least one of each element or each terminal constituting a circuit including the equivalent circuit of the magnetic coupling circuit, generating edges according to the connection relationships between the nodes, and connecting the nodes with the generated edges. A method for manufacturing a feature extraction device, comprising:

Images