Current-distribution estimation device, current-distribution estimation method, current-distribution estimation program, and current-distribution estimation system

Abstract

This current-distribution estimation device has: an acquisition portion that acquires a magnetic-field signal indicating each of a plurality of magnetic fields detected at a plurality of detection positions on a surface of a conductor through which a current flows; an optimization portion that optimizes, on the basis of a predetermined optimization algorithm, a current that generates a magnetic field detected at each of the plurality of detection positions; a calculation portion that calculates a plurality of magnetic fields generated at the plurality of detection positions by the current optimized by the optimization unit; a determination portion that determines whether or not the plurality of magnetic fields acquired by the acquisition portion match the plurality of magnetic fields calculated by the calculation portion; and an output portion that, when the determination portion determines that the plurality of magnetic fields acquired by the acquisition portion match the plurality of magnetic fields calculated by the calculation portion, outputs a current signal indicating the current optimized by the optimization portion.

Description

Current distribution estimation device, current distribution estimation method, current distribution estimation program, and current distribution estimation system

[0001] The present disclosure relates to a current distribution estimation device, a current distribution estimation method, a current distribution estimation program, and a current distribution estimation system.

[0002] In recent years, semiconductor materials forming switching elements used in power converters such as inverters and converters have been replaced from Si semiconductors to SiC semiconductors and GaN semiconductors, which are capable of withstanding higher voltages, in order to improve the withstand voltage of the switching elements.

[0003] Furthermore, in order to miniaturize power converters, switching elements used in power converters are beginning to use leadless packages (surface mount type) instead of conventional TO packages (e.g., TO-247) in order to reduce inductance. When a TO package is used as a switching element, the current flowing through the switching element can be measured via a lead, but in the case of a leadless package, the current flowing through the switching element must be measured using flat wiring arranged on a circuit board. It is not easy to measure the current flowing through flat wiring arranged on a circuit board using general ammeters such as Rogowski coils, current probes, and current transformers.

[0004] Furthermore, in flat conductors such as flat wiring arranged on circuit boards and bus bars connecting devices, current may not flow uniformly within the conductor, but may flow unevenly within the conductor. Measuring the distribution of unevenly flowing current within the conductor can optimize the design of power converters equipped with flat conductors. However, because conventional ammeters such as Rogowski coils measure current by integrating the magnetic field around the conductor, they cannot measure localized current unevenness and current distribution within the flat conductor.

[0005] On the other hand, Japanese Patent No. 7222480 (hereinafter referred to as Patent Document 1) describes a magnetic field sensor device having a Faraday rotator formed of a granular film and capable of being miniaturized. The magnetic field sensor device described in Patent Document 1 has a Faraday rotator that can be miniaturized, so it can detect the magnetic field distribution near the surface of a flat conductor with high accuracy.

[0006] However, although the magnetic field sensor device described in Patent Document 1 can detect the magnetic field distribution on the surface of a flat conductor with high accuracy, it is not easy to estimate the current distribution flowing through the conductor from the magnetic field distribution detected by the magnetic field sensor device described in Patent Document 1.

[0007] The present disclosure is intended to solve such problems, and aims to provide a current distribution estimation device that can estimate the current distribution in a flat conductor and a conductor in which current is unevenly distributed due to the skin effect and the proximity effect.

[0008] The current distribution estimation device according to the present disclosure includes an acquisition unit that acquires magnetic field signals indicating each of a plurality of magnetic fields detected at a plurality of detection positions on the surface of a conductor through which a current flows; an optimization unit that optimizes the current that generates the magnetic field detected at each of the plurality of detection positions based on a predetermined optimization algorithm; a calculation unit that calculates a plurality of magnetic fields generated at the plurality of detection positions by the current optimized by the optimization unit; a determination unit that determines whether the plurality of magnetic fields acquired by the acquisition unit and the plurality of magnetic fields calculated by the calculation unit match; and an output unit that outputs a current signal indicating the current optimized by the optimization unit when the determination unit determines that the plurality of magnetic fields acquired by the acquisition unit and the plurality of magnetic fields calculated by the calculation unit match.

[0009] Furthermore, in the current distribution estimation device according to the present invention, it is preferable that the calculation unit determines an initial value of a current that generates a magnetic field detected at each of the plurality of detection positions, and the optimization unit optimizes the initial value.

[0010] Furthermore, in the current distribution estimation device according to the present invention, it is preferable that the calculation unit calculates each of the multiple magnetic fields by considering the current generating the magnetic field detected at each of the multiple detection positions as a current flowing in a conductor having a circular cross section.

[0011] Furthermore, in the current distribution estimation device according to the present invention, it is preferable that the optimization unit optimizes the current based on one of a genetic algorithm, a gradient descent method, and a local search method.

[0012] Furthermore, in the current distribution estimation device according to the present invention, it is preferable that the optimization unit optimizes the current based on a gradient descent method.

[0013] Furthermore, in the current distribution estimation device according to the present invention, it is preferable that the optimization unit calculates the error between each of the multiple magnetic fields acquired by the acquisition unit and each of the multiple magnetic fields calculated by the calculation unit using the least squares method.

[0014] Furthermore, the current distribution estimation method according to the present invention includes acquiring magnetic field signals indicating each of a plurality of magnetic fields detected at a plurality of detection positions on the surface of a conductor through which a current flows, optimizing the current generating the magnetic field detected at each of the plurality of detection positions based on a predetermined optimization algorithm, calculating a plurality of magnetic fields generated at the plurality of detection positions by the optimized current, determining whether the acquired plurality of magnetic fields and the calculated plurality of magnetic fields match, and outputting a current signal indicating the optimized current when it is determined that the acquired plurality of magnetic fields and the calculated plurality of magnetic fields match.

[0015] In addition, the current distribution estimation program of the present invention causes a computer to execute the following process: acquire magnetic field signals indicating each of multiple magnetic fields detected at multiple detection positions on the surface of a conductor through which a current flows; optimize the current generating the magnetic field detected at each of the multiple detection positions based on a predetermined optimization algorithm; calculate multiple magnetic fields generated at the multiple detection positions by the optimized current; determine whether the acquired multiple magnetic fields and the calculated multiple magnetic fields match; and output a current signal indicating the optimized current when it is determined that the acquired multiple magnetic fields and the calculated multiple magnetic fields match.

[0016] Furthermore, a current distribution estimation system according to the present invention includes a magnetic field sensor device that detects multiple magnetic fields at multiple detection positions on the surface of a conductor through which a current flows and outputs magnetic field signals indicating each of the detected multiple magnetic fields, and a current distribution estimation device that estimates a current distribution that generates a magnetic field corresponding to the magnetic field signals. The current distribution estimation device includes an acquisition unit that acquires the magnetic field signals, an optimization unit that optimizes, based on a predetermined optimization algorithm, the current that generates the magnetic field detected at each of the multiple detection positions, a calculation unit that calculates multiple magnetic fields that are generated at the multiple detection positions by the current optimized by the optimization unit, a determination unit that determines whether the multiple magnetic fields acquired by the acquisition unit and the multiple magnetic fields calculated by the calculation unit match, and an output unit that outputs a current signal that indicates the current optimized by the optimization unit when the determination unit determines that the multiple magnetic fields acquired by the acquisition unit and the multiple magnetic fields calculated by the calculation unit match.

[0017] The current distribution estimation device according to the present invention can estimate the current distribution in a conductor and in a conductor in which current is unevenly distributed due to the skin effect and the proximity effect.

[0018] 1A is a schematic diagram of a current distribution estimation system including a current distribution estimation device according to a first embodiment, and FIG. 1B is a cross-sectional view taken along line A-A in FIG. 1A. It is a block diagram of a first magnetic field sensor device shown in FIG. 1. It is a block diagram of a second magnetic field sensor device shown in FIG. 1. It is a block diagram of a current distribution estimation device shown in FIG. 1. It is a flowchart of a current distribution estimation process executed by the current distribution estimation device shown in FIG. 4. It is a flowchart showing a more detailed process of the process indicated by S108 in FIG. 5. (a) is a diagram showing a time-dependent change in the current supplied to the conductor from the current source shown in FIG. 1, (b) is a diagram showing the magnetic field distribution in the width direction of the conductor at the time indicated by arrow D in (a), and (c) is a diagram showing the current distribution estimation result by the current distribution estimation device corresponding to (b). It is a block diagram of a current distribution estimation device according to a second embodiment. It is a flowchart of a current distribution estimation process executed by the current distribution estimation device shown in FIG. 8. It is a diagram for explaining the gradient descent process indicated by S308 in FIG. 9. 9 is a diagram (part 1) showing the estimation result of the current distribution by the current distribution estimation device shown in Fig. 8, where (a) shows the relationship between the initial value of the current and the target current, (b) shows the relationship between the magnetic field corresponding to the initial value of the current and the target magnetic field, (c) shows the relationship between the estimated value of the current and the target current, and (d) shows the relationship between the magnetic field corresponding to the estimated value of the current and the target magnetic field. 9 is a diagram (part 2) showing the estimation result of the current distribution by the current distribution estimation device shown in Fig. 8, where (a) shows the relationship between the initial value of the current and the target current, (b) shows the relationship between the magnetic field corresponding to the initial value of the current and the target magnetic field, (c) shows the relationship between the estimated value of the current and the target current, and (d) shows the relationship between the magnetic field corresponding to the estimated value of the current and the target magnetic field. 9A and 9B are diagrams (part 3) showing the results of current distribution estimation by the current distribution estimation device shown in Fig. 8, in which (a) shows the change over time of the current supplied to the conductor from the current source shown in Fig. 1, (b) shows the change over time of the estimated value of the current supplied from the current source to the conductor, (c) shows the change over time of the estimated value of the current distribution supplied from the current source to the conductor (part 1), and (d) shows the change over time of the estimated value of the current distribution supplied from the current source to the conductor (part 2). Fig. 9 is a flowchart of a current distribution estimation process according to a modified example, which is executed by the current distribution estimation device shown in Fig. 8.(a) is a diagram showing a conductor according to a first modified example, (b) is a diagram showing a conductor according to a second modified example, (c) is a diagram showing a conductor according to a third modified example, and (d) is a diagram showing a conductor according to a fourth modified example.

[0019] Hereinafter, a current distribution estimation device, a current distribution estimation method, a current distribution estimation program, and a current distribution estimation system according to embodiments will be described with reference to the drawings. However, it should be noted that the technical scope of the present invention is not limited to these embodiments, but extends to the inventions set forth in the claims and their equivalents.

[0020] FIG. 1(a) is a schematic diagram of a current distribution estimation system having a current distribution estimation device according to the first embodiment, and FIG. 1(b) is a cross-sectional view taken along line AA shown in FIG. 1(a).

[0021] The current distribution estimation system 100 includes a first magnetic field sensor device 101, a second magnetic field sensor device 102, a current source 108, and a current distribution estimation device 1, and estimates a current distribution when a current supplied from the current source 108 flows in a first direction indicated by an arrow in a flat conductor 200. The flat conductor 200 is, for example, a bus bar disposed on a distribution board of a power receiving device, or a wiring pattern that is a thin metal film disposed on a circuit board.

[0022] The first magnetic field sensor device 101 has a first main device 103 and a first sensor head 105, and detects the X direction of the magnetic field generated by the current flowing through the conductor 200, i.e., the magnetic field in the normal direction to the surface of the conductor 200, and outputs a first magnetic field signal indicating the detected magnetic field to the current distribution estimation device 1. The second magnetic field sensor device 102 has a second main device 104 and a second sensor head 106, and detects the Y direction of the magnetic field generated by the current flowing through the conductor 200, i.e., the magnetic field in a direction perpendicular to the X direction, and outputs a second magnetic field signal indicating the detected magnetic field to the current distribution estimation device 1. The first magnetic field sensor device 101 and the second magnetic field sensor device 102 detect the distribution of the magnetic field generated by the current flowing through the conductor 200 by scanning the first sensor head 105 and the second sensor head 106 in a second direction perpendicular to the first direction in which the current flows through the conductor 200. The second direction in which the first sensor head 105 and the second sensor head 106 scan is parallel to the Y direction in which the second sensor head 106 detects the magnetic field.

[0023] The first sensor head 105 and the second sensor head 106 may be supported by a moving mechanism (not shown) that can move the first sensor head 105 and the second sensor head 106 in the second direction. The moving mechanism (not shown) moves the first sensor head 105 and the second sensor head 106 along the second direction while maintaining the first sensor head 105 and the second sensor head 106 at positions spaced apart by a height h from the center of the conductor 200 in the normal direction to the surface of the conductor 20 ... N The first sensor head 105 and the second sensor head 106 are sequentially moved to N detection positions along the second direction from the first detection position D1 to the Nth detection position D N The distance between the first sensor head 105 and the second sensor head 106 is the detection pitch P. The first sensor head 105 and the second sensor head 106 are located at the first detection position D1 to the Nth detection position D N The first magnetic field H1 to the Nth magnetic field H2 are generated by the current flowing through the conductor 200 in each of the N Detect.

[0024] FIG. 2 is a block diagram of the first magnetic field sensor device 101.

[0025] The first magnetic field sensor device 101 includes a light-emitting unit 110, a circulator 120, a first optical element 130, an optical path unit 140, a first sensor head 105, a beam splitter 150, and a magnetic field signal generating circuit 160. The light-emitting unit 110, the circulator 120, the first optical element 130, the optical path unit 140, the beam splitter 150, and the magnetic field signal generating circuit 160 are housed in a first main body device 103. The optical paths between the light-emitting unit 110, the circulator 120, the first optical element 130, the optical path unit 140, the first sensor head 105, the beam splitter 150, and the magnetic field signal generating circuit 160 are formed by PANDA fiber. The optical path between the light-emitting unit 110 and the magnetic field signal generating circuit 160 may be formed by a polarization-maintaining optical fiber such as a bow-tie fiber or an elliptical jacket fiber.

[0026] The light-emitting unit 110 has a light-emitting element 111, an isolator 112, and a polarizer 113. The light-emitting element 111 is, for example, a semiconductor laser or a light-emitting diode. Specifically, a Fabry-Perot laser, a superluminescence diode, or the like can be preferably used as the light-emitting element 111. The light-emitting unit 110 emits laser light in response to an instruction from an external device (not shown). The laser light emitted from the light-emitting unit 110 is output as an AC waveform, such as a sine waveform, which oscillates at a predetermined period.

[0027] The isolator 112 transmits the light incident from the light emitting element 111 to the circulator 120 side, and does not transmit the light incident from the circulator 120 to the light emitting element 111 side, thereby protecting the light emitting element 111. The isolator 112 is, for example, a polarization-dependent optical isolator, but may also be a polarization-independent optical isolator.

[0028] The polarizer 113 is an optical element for converting the light emitted by the light emitting element 111 into linearly polarized light, and the type thereof is not particularly limited. The first linearly polarized light having a sine waveform obtained by the polarizer 113 is incident on the first optical element 130 via the circulator 120. The first linearly polarized light is also referred to as incident light.

[0029] The circulator 120 is an optical branching unit that transmits the first linearly polarized light emitted from the light emitting unit 110 to the first optical element 130 and branches the second linearly polarized light having a sine waveform emitted from the first optical element 130 to the magnetic field signal generating circuit 160. The circulator 120 is formed by, for example, a Faraday rotator, a half-wave plate, a polarizing beam splitter, and multiple reflecting mirrors. The second linearly polarized light is also called returned light.

[0030] First optical element 130 is, for example, a half-wave plate arranged so that its azimuth angle is 22.5 degrees with respect to the plane of polarization of the first linearly polarized light incident from circulator 120. First optical element 130 rotates the plane of polarization of the first linearly polarized light incident from circulator 120 by 45 degrees and emits the first linearly polarized light to optical path unit 140. The first linearly polarized light whose plane of polarization has been rotated 45 degrees by first optical element 130 has first linearly polarized light CW1 which is P-polarized light and second linearly polarized light CCW1 which is S-polarized light orthogonal to first linearly polarized light CW1.

[0031] Furthermore, the first optical element 130 rotates the polarization plane of the second linearly polarized light, which is linearly polarized light incident from the optical path section 140 , by 45 degrees and outputs the second linearly polarized light to the circulator 120 .

[0032] The optical path section 140 includes a first beam splitter 141 , a second beam splitter 142 , a first optical path 143 , a second optical path 144 , and a second optical element 145 .

[0033] The first beam splitter 141 outputs the first linearly polarized light CW1 to the first optical path 143 and outputs the second linearly polarized light CCW1 to the second optical path 144. Furthermore, the first beam splitter 141 receives the third linearly polarized light CW2 from the second optical path 144 and the fourth linearly polarized light CCW2 from the first optical path 143. The third linearly polarized light CW2 and the fourth linearly polarized light CCW2 are polarization components that are orthogonal to each other of the second linearly polarized light that is output to the first optical element 130.

[0034] The second beam splitter 142 receives the first linearly polarized light CW1 from the first optical path 143 and the second linearly polarized light CCW1 from the second optical path 144, and outputs the input first linearly polarized light CW1 and second linearly polarized light CCW1 to the first sensor head 105. Furthermore, the second beam splitter 142 outputs the third linearly polarized light CW2 input from the first sensor head 105 to the second optical path 144, and outputs the fourth linearly polarized light CCW2 input from the first sensor head 105 to the first optical path 143.

[0035] The first beam splitter 141 and the second beam splitter 142 separate the first linearly polarized light into a P-polarized component and an S-polarized component, and then combine and emit the P-polarized component and the S-polarized component. The first beam splitter 141 and the second beam splitter 142 are, for example, prism-type beam splitters, but may also be planar-type beam splitters or wedge-type beam splitters.

[0036] The first optical path 143 guides the first linearly polarized light CW1 introduced from the first beam splitter 141 to the second beam splitter 142, and also guides the fourth linearly polarized light CCW2 introduced from the second beam splitter 142 to the first beam splitter 141. The second optical path 144 guides the second linearly polarized light CCW1 introduced from the first beam splitter 141 to the second beam splitter 142, and also guides the third linearly polarized light CW2 introduced from the second beam splitter 142 to the first beam splitter 141.

[0037] The first optical path 143 is a PANDA fiber having one end optically connected to the first beam splitter 141 and the other end optically connected to the second beam splitter 142. The second optical path 144 is a PANDA fiber having one end optically connected to the first beam splitter 141 and the other end optically connected to the second beam splitter 142. Note that the first optical path 143 and the second optical path 144 may be polarization-maintaining fibers such as bowtie fibers and elliptical jacket fibers. A second optical element 145 is disposed in the second optical path 144.

[0038] The second optical element 145 includes a first (¼) wave plate 146 , a second (¼) wave plate 147 , and a 45-degree Faraday rotator 148 .

[0039] The first (quarter) wave plate 146 is a quarter-wave plate whose optical axis is inclined at 45 degrees with respect to the slow axis and fast axis of the PANDA fiber that forms the second optical path 144. The first (quarter) wave plate 146 converts linearly polarized light into circularly polarized light and converts circularly polarized light into linearly polarized light.

[0040] The second (1 / 4) wave plate 147 is a 1 / 4 wave plate whose optical axis is inclined by −45 degrees with respect to the slow axis and fast axis of the PANDA fiber forming the second optical path 144. The second (1 / 4) wave plate 147 converts the circularly polarized light from the 45-degree Faraday rotator 148 into linearly polarized light, and also converts the linearly polarized light into circularly polarized light.

[0041] The 45-degree Faraday rotator 148 is a Faraday rotator that changes the phase of the circularly polarized light incident from each of the first (¼) wave plate 146 and the second (¼) wave plate 147 .

[0042] The 45-degree Faraday rotator 148 changes the phase of the circularly polarized light incident from the first (¼) wave plate 146 so as to shift the phase of the second linearly polarized light CCW1 by 45 degrees. The 45-degree Faraday rotator 148 also changes the phase of the circularly polarized light so that the phase of the third linearly polarized light CW2 output from the first (¼) wave plate 146 is shifted by −45 degrees from the phase of the third linearly polarized light CW2 incident on the second (¼) wave plate 147.

[0043] The first sensor head 105 has a quarter-wave plate 151, a Faraday rotator 152, and a mirror element 153, is connected to the second beam splitter 142 via a PANDA fiber, and is an element at least a portion of which can be placed within a predetermined magnetic field. The first sensor head 105 detects the X component of the magnetic field generated by a current flowing through the conductor 200. The first sensor head 105 receives the first linearly polarized light emitted by the light emitting unit 110 as incident light, and emits returning light corresponding to the incident light.

[0044] The quarter-wave plate 151 is a quarter-wave plate whose optical axis is inclined at 45 degrees with respect to the slow axis and fast axis of the PANDA fiber that optically connects to the second beam splitter 142. The quarter-wave plate 151 converts the polarization state of the incident light, which is linearly polarized light, into circularly polarized light, and also converts the polarization state of the return light that is incident as circularly polarized light from the Faraday rotator 152 into linearly polarized light.

[0045] The Faraday rotator 152 is a granular film having a dielectric and nano-sized magnetic particles dispersed in the dielectric in a state of stable phase separation from the dielectric, and is disposed on the end face of the quarter-wave plate 151. The magnetic particles may form oxides in a small portion, such as the outermost layer, but throughout the Faraday rotator 152, the magnetic particles are dispersed independently in the thin film without forming a compound with the dielectric that serves as a binder. The distribution of the magnetic particles in the Faraday rotator 152 does not need to be completely uniform, and may be somewhat biased. If a highly transparent dielectric is used, the magnetic particles will exist in the dielectric at a size smaller than the wavelength of light, making the Faraday rotator 152 optically transparent.

[0046] The Faraday rotator 152 may be a magnetic thin film other than a granular film. For example, the Faraday rotator 152 may be a rare earth iron garnet substituted with Y. Ｘ Y ３－Ｘ Fe ５ O １２ The rare earth iron garnet thin film may be a single crystal having a garnet-type crystal structure represented by the following composition formula: where R is a rare earth metal and is an element that can be substituted for Y.

[0047] The Faraday rotator 152 is not limited to a single layer, but may be a multilayer film in which granular films and dielectric films are alternately stacked. By forming the Faraday rotator 152 using a multilayer granular film, a larger Faraday rotation angle can be obtained due to multiple reflections within the granular film.

[0048] The dielectric is magnesium fluoride (MgF ２ ), aluminum fluoride (AlF ３), yttrium fluoride (YF ３ Fluorides (metal fluorides) such as tantalum oxide (Ta ２ O ５ ), silicon dioxide (SiO ２ ), titanium dioxide (TiO ２ ), niobium pentoxide (Nb ２ O ５ ), zirconium dioxide (ZrO ２ ), hafnium dioxide (HfO ２ ), and aluminum trioxide (Al ２ O ３ For good phase separation from the dielectric magnetic particles, fluorides are more preferable than oxides, and magnesium fluoride, which has high transmittance, is particularly preferable.

[0049] The magnetic particle material is not particularly limited as long as it produces the Faraday effect, but examples of the magnetic particle material include ferromagnetic metals such as iron (Fe), cobalt (Co), and nickel (Ni), as well as alloys thereof. Examples of alloys of Fe, Co, and Ni include FeNi alloys, FeCo alloys, FeNiCo alloys, and NiCo alloys. The Faraday rotation angle per unit length of Fe, Co, and Ni is two to three orders of magnitude larger than that of magnetic garnets used in conventional Faraday rotators.

[0050] The mirror element 153 is formed on the Faraday rotator 152 and reflects light transmitted through the Faraday rotator 152 back toward the Faraday rotator 152. The mirror element 153 can be, for example, a silver (Ag) film, a gold (Au) film, an aluminum (Al) film, or a dielectric multilayer mirror. Ag films with high reflectivity and Au films with high corrosion resistance are particularly preferred for their ease of film formation. The thickness of the mirror element 153 need only be large enough to ensure a sufficient reflectivity of 98% or more. For example, in the case of an Ag film, the thickness is preferably 50 nm or more and 200 nm or less. By using the mirror element 153 to cause light to travel back and forth within the Faraday rotator 152, the Faraday rotation angle can be increased.

[0051] The circularly polarized light incident on the Faraday rotator 152 from the quarter-wave plate 151 passes through the Faraday rotator 152, is reflected by the mirror element 153, and passes through the Faraday rotator 152 again to become returned light. The returned light that has passed through the Faraday rotator 152 is incident on the quarter-wave plate 151 again.

[0052] The circularly polarized light incident on the Faraday rotator 152 from the quarter-wave plate 151 changes its phase in accordance with the magnetic field applied to the Faraday rotator 152. The circularly polarized light reflected by the mirror element 153 further changes its phase in accordance with the magnetic field applied to the Faraday rotator 152.

[0053] The beam splitter 150 is a polarizing beam splitter (PBS) such as a prism type, a planar type, a wedge substrate type, or an optical waveguide type, and separates the second linearly polarized light having a sine waveform branched by the circulator 120 into a P-polarized component and an S-polarized component.

[0054] The magnetic field signal generating circuit 160 receives the P-polarized component and the S-polarized component separated from the second linearly polarized light by the beam splitter 150, converts them into electrical signals, and differentially amplifies them to output a first magnetic field signal Ed1 corresponding to the magnetic field applied to the first magnetic field sensor device 101.

[0055] FIG. 3 is a block diagram of the second magnetic field sensor device 102.

[0056] The second magnetic field sensor device 102 differs from the first magnetic field sensor device 101 in that it has a second sensor head 106 instead of the first sensor head 105. The configurations and functions of the components of the second magnetic field sensor device 102 other than the second sensor head 106 are the same as the configurations and functions of the components of the first magnetic field sensor device 101 with the same reference numerals, so detailed description will be omitted here. In the second magnetic field sensor device 102, the light emitting unit 110, the circulator 120, the first optical element 130, the optical path unit 140, the beam splitter 150, and the magnetic field signal generating circuit 160 are housed in a second main body device 104. The second magnetic field sensor device 102 outputs a second magnetic field signal Ed2 corresponding to the magnetic field applied to the second magnetic field sensor device 102.

[0057] The second sensor head 106 has a quarter-wave plate 161, a Faraday rotator 162, a mirror element 163, and an optical element 164. It is connected to the second beam splitter 142 via a PANDA fiber, and at least a portion of it can be placed within a predetermined magnetic field. The second magnetic field sensor element 107 detects the X component of a magnetic field generated by a current flowing through the conductor 200. The configurations and functions of the quarter-wave plate 161, the Faraday rotator 162, and the mirror element 163 are the same as those of the quarter-wave plate 151, the Faraday rotator 152, and the mirror element 153, and therefore will not be described in detail here. The optical element 164 is a light-reflecting element such as a mirror or a prism. The optical element 164 shifts the light that has passed through the quarter-wave plate 161 by 90 degrees in a direction parallel to the Y direction and makes it incident on the Faraday rotator 162, and also shifts the light that has passed through the Faraday rotator 162 by 90 degrees in a direction parallel to the X direction and makes it incident on the quarter-wave plate 161.

[0058] FIG. 4 is a block diagram of the current distribution estimation device 1 according to the first embodiment.

[0059] The current distribution estimation device 1 has a communication unit 11, a storage unit 12, an input unit 13, a display unit 14, and a processing unit 20, and executes various processes based on programs pre-stored in the storage unit 12 and by referring to data stored in the storage unit 12. The current distribution estimation device 1 is an electronic computer that executes various processes in response to instructions input by an operator via the input unit 13 and outputs the execution results to the display unit 14. The current distribution estimation device 1 estimates the distribution of the current supplied to the conductor 200 by replacing the current supplied to the conductor 200 with a current flowing through multiple minute conductors with circular cross sections that are adjacently arranged in a line in a second direction perpendicular to the first direction. The current distribution estimation device 1 estimates multiple currents that generate multiple magnetic fields corresponding to the first magnetic field signal Ed1 and the second magnetic field signal Ed2 based on a genetic algorithm, and outputs current signals indicative of the estimated currents.

[0060] The communication unit 11 has a communication interface circuit for connecting the current distribution estimation device 1 to a network (not shown). The communication unit 11 supplies data received from the first magnetic field sensor device 101, the second magnetic field sensor device 102, and an external device (not shown) via the network to the processing unit 20. The communication unit 11 also transmits data supplied from the processing unit 20 to the first magnetic field sensor device 101, the second magnetic field sensor device 102, and the external device via the network. The communication unit 11 receives a first magnetic field signal Ed1 and a second magnetic field signal Ed2 from the first magnetic field sensor device 101 and the second magnetic field sensor device 102, respectively.

[0061] The storage unit 12 includes, for example, one of a semiconductor memory, a magnetic disk device, and an optical disk device. The storage unit 12 stores an operating system program, a driver program, an application program, data, and the like used for processing by the processing unit 20. For example, the storage unit 12 stores, as driver programs, an input device driver program that controls the input unit 13, an output device driver program that controls the display unit 14, and the like. The storage unit 12 also stores, as application programs, a current distribution estimation program that causes the processing unit 20 to execute a current distribution estimation process that estimates a current distribution when a current flows through the conductor 200. The current distribution estimation program may be installed into the storage unit 12 from a computer-readable portable storage medium such as a CD-ROM or a DVD-ROM using a known setup program or the like. The storage unit 12 also stores various data used in the current distribution estimation.

[0062] The input unit 13 may be any device, such as a keyboard or a touchpad, that can operate the current distribution estimation device 1. An operator can input letters, numbers, and the like via the input unit 13. When operated by the operator, the input unit 13 generates a signal corresponding to the operation. The generated signal is then supplied to the processing unit 20 as an instruction from the operator.

[0063] The display unit 14 may be any device capable of displaying videos, images, text, etc., such as a liquid crystal display, an organic EL display, etc. The display unit 14 displays videos corresponding to video data, images corresponding to image data, text corresponding to text data, etc., supplied from the processing unit 20. The display unit 14 may also display a graphical user interface for operating the current distribution estimation device 1.

[0064] The processing unit 20 has one or more processors and their peripheral circuits. The processing unit 20 is, for example, a CPU, and comprehensively controls the overall operation of the current distribution estimation device 1. The processing unit 20 controls the operations of the communication unit 11, display unit 14, etc. so that various processes of the current distribution estimation device 1 are executed in an appropriate order in accordance with the programs stored in the storage unit 12, operations on the input unit 13, etc. The processing unit 20 executes processes based on the programs stored in the storage unit 12. The processing unit 20 can also execute multiple programs in parallel.

[0065] The processing unit 20 includes an instruction unit 21, an acquisition unit 22, a calculation unit 23, a determination unit 24, a genetic algorithm execution unit 25, and an output unit 26. The genetic algorithm execution unit 25 is also referred to as an optimization unit. Each of these units is a functional module implemented by a program executed by a processor included in the processing unit 20. Alternatively, each of these units may be implemented in the current distribution estimation device 1 as firmware.

[0066] Fig. 5 is a flowchart of the current distribution estimation process executed by the current distribution estimation device 1. The current distribution estimation process shown in Fig. 5 is executed mainly by the processing unit 20 in cooperation with each element of the current distribution estimation system 100, based on a current distribution estimation program stored in advance in the storage unit 12.

[0067] First, the instruction unit 21 outputs a first detection instruction to the first magnetic field sensor device 101, the second magnetic field sensor device 102, the current source 108, and a movement mechanism (not shown) to detect a magnetic field at a first detection position D1 (S101). In response to the input of the first detection instruction, the movement mechanism (not shown) moves the first sensor head 105 and the second sensor head 106 to the first detection position D1. In response to the input of the first detection instruction, the current source 108 supplies a rising current to the conductor 200, which increases the current value from zero to a predetermined rated value. In response to the input of the first detection instruction, the first magnetic field sensor device 101 and the second magnetic field sensor device 102 detect the X component H of the magnetic field at the first detection position D1. X and the Y component H Y The first magnetic field sensor device 101 and the second magnetic field sensor device 102 detect the X component H X and the Y component H Y The current distribution estimation device 1 outputs a first magnetic field signal Ed1 and a second magnetic field signal Ed2 indicating the above.

[0068] Next, the acquisition unit 22 acquires the magnetic field detected at the first detection position (S102). The acquisition unit 22 acquires the X component H of the magnetic field corresponding to the first magnetic field signal Ed1. X and the Y component H of the magnetic field corresponding to the second magnetic field signal Ed2 Y The acquisition unit 22 acquires the X component H of the magnetic field at a predetermined sampling interval. X and the Y component H of the magnetic field. Y is stored in the storage unit 12 in association with identifiers indicating the first detection position D1 and the first sampling time T1 to the Mth sampling time TM, which are the sampling times.

[0069] Next, the indicator 21 moves to the Nth detection position D N The instruction unit 21 determines whether or not the magnetic field has been detected at all detection positions up to the N-th detection position D (S103). NThe instruction unit 21 does not determine that the magnetic field has been detected at all detection positions up to the Nth detection position (S103-NO), and outputs a second detection instruction to instruct the magnetic field to be detected at the second detection position D2 (S101). Next, the acquisition unit 22 acquires the magnetic field detected at the second detection position (S102). Thereafter, the instruction unit 21 acquires the magnetic field detected at the Nth detection position D N The processes shown in S101 to S103 are repeated until it is determined that the magnetic field has been detected at all detection positions (S103-YES).

[0070] Nth detection position D N When it is determined that the magnetic field has been detected at all the detection positions up to the first sampling time (S103-YES), the calculation unit 23 calculates the magnetic field at the first detection position D1 to the Nth detection position D2 at the first sampling time. N The first current I1 to the Nth current I N The initial values ​​of the first current I1 to the Nth current I are determined (S104). N The initial values ​​of the first current I1 to the Nth current I2 may be any value, but are preferably determined according to the current value of the current supplied to the conductor 200. For example, the first current I1 to the Nth current I2 N The initial value of may be a current value obtained by dividing the total current supplied to the conductor 200 equally.

[0071] Next, the calculation unit 23 calculates the first current I1 to the Nth current I2 determined in the process shown in S104. N First, the magnetic field generated from each of the initial values ​​of the first detection position D1 to the Nth detection position D is calculated (S105). N The first current I1 to the Nth current I N The calculation unit 23 calculates the magnetic fields detected by the first sensor and the second sensor head 106 disposed at the first detection position D1 by the first current I1 to the Nth current I2. N The X component H of the first magnetic field resulting from each of the initial values ​​of 1x and Y component H 1Y ~ X component H of the Nth magnetic field NX and Y component H NY The X component of the first magnetic field, H 1x and Y component H 1Y is the first current I1 to the Nth current I N It is calculated by combining the magnetic fields generated from each of the above.

[0072] First detection position D1 and second detection position D2 to Nth detection position D N When the nth current I(n) flows through each of the first and second sensor heads 105 and 106, the nth magnetic field vector H(n) detected by the first and second sensor heads 105 and 106 located at the first detection position D1 is expressed by equation (1).

[0073]

[0074] Here, H X (n) denotes the X component of the magnetic field H(n), and H Y (n) indicates the Y component of the magnetic field H(n). H(n) is a magnetic field generated by current flowing through multiple microconductors with circular cross sections, and is therefore expressed by equation (2) based on Ampere's law, and θ' is an angle in the range of ±180° as shown in equation (3). In equations (2) and (3), the height h indicates the distance in the X direction from the center of the conductor 200 to the first sensor head 105 and the second sensor head 106, and the width w(n) indicates the distance between the first detection position D1 and the second detection position D2 to the Nth detection position D N I(n) is the distance in the Y direction between the second detection position D2 to the Nth detection position D N is the initial value of the current.

[0075]

[0076]

[0077] The magnetic field vector H(m) obtained by combining the n-th magnetic field H(n) shown in equation (1) is expressed by equation (4).

[0078]

[0079] The magnitude of the combined magnetic field vector H(m) is given by equation (5), and the angle of the combined magnetic field vector H(m) is given by equation (6).

[0080]

[0081]

[0082] The calculation unit 23 stores the calculated magnitude and angle of the magnetic field vector H(m) in the storage unit 12 in association with an identifier indicating the first detection position D1.

[0083] Next, the calculation unit 23 calculates the first detection position D1 to the Nth detection position D N When the nth current I(n) flows through each of the first and second sensor heads 105, 106, the calculation unit 23 calculates the magnetic fields detected by the first and second sensor heads 105, 106, which are located at the second detection position D2. The calculation unit 23 calculates the magnitude and angle of a combined magnetic field vector H(m) using equations (1) to (6) for the magnetic fields detected by the first and second sensor heads 105, 106, which are located at the second detection position D2. The calculation unit 23 stores the calculated magnitude and angle of the magnetic field vector H(m) in the storage unit 12 in association with an identifier indicating the second detection position D2.

[0084] Thereafter, the calculation unit 23 calculates the third detection position D3 to the Nth detection position D N The magnetic field detected by each of the first sensor head 105 and the second sensor head 106 arranged in the same direction is sequentially calculated, and the calculated magnetic field is stored in the storage unit 12.

[0085] Next, the calculation unit 23 calculates the error ε between the first magnetic field Hx,m(1) to the Nth magnetic field Hx,m(N) measured in the processes shown in S101 to S103 and the first magnetic field Hx(1) to the Nth magnetic field Hx(N) calculated in the process shown in S105 (S106). The calculation unit 23 calculates the error ε between the measured first magnetic field Hx,m(1) to the Nth magnetic field Hx,m(N) and the first magnetic field Hx(1) to the Nth magnetic field Hx(N) calculated in the process shown in S105 by the least squares method using equation (7). The calculation unit 23 stores error information indicating the calculated error ε in the memory unit 12.

[0086]

[0087] Next, the determination unit 24 determines whether the error ε calculated in the process shown in S106 is equal to or less than a predetermined determination threshold value (S107). If it is determined that the error ε is not equal to or less than the determination threshold value (S107-NO), the genetic algorithm execution unit 25 executes the genetic algorithm to calculate the first current I to the Nth current I used in the calculation of the magnetic field in the process shown in S105. N is optimized (S108).

[0088] Next, the calculation unit 23 calculates the first current I1 to the Nth current I2 optimized in the process shown in S108. N The magnetic fields generated from each of these are calculated (S105). Thereafter, the processes shown in S105 to S108 are repeated until it is determined that the error ε is equal to or smaller than the determination threshold value (S107-NO).

[0089] If it is determined that the error ε is equal to or smaller than the determination threshold value (S107-YES), the genetic algorithm executing unit 25 calculates the first current I1 to the Nth current I2 optimized in the process shown in S108. N is determined as the current flowing through the conductor 200 at the first sampling time (S109). N The current information indicating each of the first detection position D1 to the Nth detection position D N and the first sampling time T1, and are stored in the storage unit 12 in association with the identifier indicating the first sampling time T1.

[0090] Next, the calculation unit 23 calculates the first current I1 to the Nth current I2 at all sampling times. N The calculation unit 23 determines whether the first current I1 to the Nth current I N If it is determined that the first current I1 to the Nth current I2 at the second sampling time have not been determined (S110-NO), N The calculation unit 23 determines the initial values ​​of the first current I1 to the N-th current I2 at the second sampling time in the same manner as in the first sampling time (S104). N After that, the calculation unit 23 calculates the first current I1 to the Nth current I at all sampling times. N The processes shown in S104 to S110 are repeated until it is determined that the desired value has been determined (YES in S110).

[0091] At all sampling times, the first current I1 to the Nth current I N is determined to be optimized (S110-YES), the output unit 26 outputs the first current I1 to the Nth current I2 determined in the process shown in S109. NThe output unit 26 outputs a current signal indicating the first current I1 to the Nth current I for the first sampling time T1 to the Mth sampling time TM, respectively (S111). N and outputs a current signal corresponding to the acquired current information to the display unit 14. The display unit 14 displays the first current I1 to the Nth current I2 corresponding to the input current signal. N Display.

[0092] FIG. 6 is a flowchart showing the process of S108 in more detail.

[0093] First, the genetic algorithm executing unit 25 calculates the first current I1 to the Nth current I N The genetic algorithm executing unit 25 calculates the degree of conformance of the first magnetic field H(1) to the Nth magnetic field H(N) calculated in the process shown in S105 and the first magnetic field Hm(1) to the Nth magnetic field Hm(N) measured in the processes shown in S101 to S103. Next, the genetic algorithm executing unit 25 calculates the average value of the absolute values ​​of the errors between the calculated first magnetic field H(1) to the Nth magnetic field H(N) and the first magnetic field Hm(1) to the Nth magnetic field Hm(N).

[0094] Next, the genetic algorithm running unit 25 executes a selection and new individual replenishment process (S202). The genetic algorithm running unit 25 randomly selects 50% of the nth current I(n) corresponding to the nth magnetic field H(n) whose average absolute value of the error is greater than the average absolute value of the error. Next, the genetic algorithm running unit 25 randomly selects the nth current I(n) corresponding to the nth magnetic field H(n) whose absolute value of the error is equal to or less than the average absolute value, and replenishes the selected nth current I(n) with a current obtained by varying the selected nth current I(n) in the range of 80% to 120%.

[0095] Next, the genetic algorithm execution unit 25 executes a crossover process (S203). The genetic algorithm execution unit 25 selects a pair of n-th currents I(n) that are genes of randomly selected parents from a predetermined number of n-th currents I(n), and then executes the crossover process by uniform crossover.

[0096] Next, the genetic algorithm execution unit 25 executes a mutation process (S204). The genetic algorithm execution unit 25 executes a mutation process by mutating the n-th current I(n), which is 10% of the n-th current I(n) on which the crossover process has been executed, in the range of 50% to 150%.

[0097] The current distribution estimation device 1 estimates the current inside the conductor 200 based on a genetic algorithm, and can estimate the current distribution in the conductor 200 with high accuracy.

[0098] FIG. 7( a) is a diagram showing the change over time in the current supplied from the current source 108 to the conductor 200. FIG. 7( b) is a diagram showing the magnetic field distribution in the width direction of the conductor 200 at the time indicated by the arrow D in FIG. 7( a), and FIG. 7( c) is a diagram showing the current distribution estimation result by the current distribution estimation device 1 corresponding to FIG. 7( b). In FIG. 7( a), the horizontal axis represents time, the vertical axis represents the current supplied to the conductor 200, and a waveform W101 represents the change over time in the total current supplied to the conductor 200. In FIG. 7( b), the horizontal axis represents the position in the width direction of the conductor 200, and the vertical axis represents the magnetic field. The waveform W201 represents the change over time in the magnetic field detected by the first magnetic field sensor device 101 when a current is supplied to the conductor 200, and the waveform W202 represents the change over time in the magnetic field detected by the second magnetic field sensor device 102 when a current is supplied to the conductor 200. 7(c), the horizontal axis indicates the position in the width direction of the conductor 200, the vertical axis indicates the current, and dots indicate the current estimated by the current distribution estimation device 1. In Figures 7(a) to 7(c), a metal flat plate having a width of 12 mm and a thickness of 0.5 mm was used as the conductor 200, the detection pitch was 1 mm, and the current supplied to the conductor 200 was a pulse wave having an amplitude of 100 A and a slew rate of 15 A / μs.

[0099] 7( c), the current distribution estimation device 1 detects that, at the time indicated by arrow D, which is the rising edge of the current supplied to the conductor 200, the current at the end of the conductor 200 is greater than the current at the center of the conductor 200 due to the skin effect. The current distribution estimation device 1 can estimate the current distribution inside the conductor 200 with high accuracy.

[0100] 8 is a block diagram of a current distribution estimation device according to the second embodiment. Similar to the current distribution estimation device 1, the current distribution estimation device 2 is arranged in the current distribution estimation system 100, and estimates the current distribution of a current flowing through a flat conductor 200 based on the gradient descent method.

[0101] The current distribution estimation device 2 differs from the current distribution estimation device 1 in that it has a processing unit 30 instead of the processing unit 20. The processing unit 30 differs from the processing unit 20 in that it has a gradient descent processing execution unit 35, also called an optimization unit, instead of the genetic algorithm execution unit 25. The configurations and functions of the components of the current distribution estimation device 2 other than the gradient descent processing execution unit 35 are the same as the configurations and functions of the components of the current distribution estimation device 1 that are assigned the same reference numerals, and therefore will not be described here.

[0102] Fig. 9 is a flowchart of the current distribution estimation process executed by the current distribution estimation device 2. The current distribution estimation process shown in Fig. 9 is executed mainly by the processing unit 30 in cooperation with each element of the current distribution estimation system 100, based on a current distribution estimation program stored in advance in the storage unit 12. The processes shown in S301 to S307 are similar to the processes shown in S101 to S107, and therefore will not be described in detail here.

[0103] If it is determined that the error ε is not equal to or less than the determination threshold value (S307-NO), the gradient descent process execution unit 35 executes the gradient descent process to calculate the first current I1 to the Nth current I2 used in the calculation of the magnetic field in the process shown in S305. N is optimized (S308).

[0104] 10 is a diagram for explaining the gradient descent process shown in S308. In FIG. 10, the first current I1 to the Nth current I N are the first detection position D1 to the Nth detection position D on the conductor 200. N The first calculated magnetic field H C1 ~Nth calculated magnetic field H CN is the first detection position D1 to the Nth detection position D N The first magnetic field H1 to the N-th magnetic field H2 detected by the first sensor head 105 and the second sensor head 106 are N The first measured magnetic field H M1~ Nth measured magnetic field H MN is the first detection position D1 to the Nth detection position D N The first magnetic field H1 to the N-th magnetic field H2 detected by the first sensor head 105 and the second sensor head 106 are N are the actual measured values.

[0105] jth calculated magnetic field H Cj is the first current I1 to the Nth current I N Each of these has a weighting coefficient α ji and is expressed as the sum of the values ​​multiplied by (8).

[0106]

[0107] Weighting coefficient α ji is the first current I1 to the Nth current I N is a coefficient that indicates the magnetic field generated from each of the height h and horizontal distance w M1 is a variable with parameters and is shown in equation (9).

[0108]

[0109] The height h is the distance between the first sensor head 105 and the second sensor head 106 and the center of the conductor 200. ji is the i-th detection position D ｉ and the first detection position D1 to the Nth detection position D N For example, the horizontal distance w between the first detection position D1 and the second detection position D2 is 12 is twice the length of the detection pitch P, and the first detection position D1 and the Nth detection position D N The horizontal distance between 1N is N times the detection pitch P.

[0110] First measured magnetic field H M1 ~ Nth measured magnetic field H MN and the first calculated magnetic field H C1 ~Nth calculated magnetic field H CN The j-th error ε i is the first measured magnetic field H M1 ~ Nth measured magnetic field H MN and the first calculated magnetic field H C1 ~Nth calculated magnetic field HCN is the mean square error of the difference between the two, and is shown in equation (10).

[0111]

[0112] The loss function, i.e., the objective function, of the gradient descent process is the first measured magnetic field H M1 ~ Nth measured magnetic field H MN and the first calculated magnetic field H C1 ~Nth calculated magnetic field H CN The error ε between i is the sum E of the above, and is expressed by equation (11).

[0113]

[0114] In the gradient descent process, the first current I1 to the Nth current I N are the errors e included in the objective function E shown in equation (11). j The first current I1 to the Nth current I are calculated by partially differentiating N The gradients of the first current I1 to the Nth current I2 are multiplied by the step size η. N The error e is subtracted from each of the j The partial derivative of is expressed by Equation (12). As shown in Equation (12), the j-th error ε i and j calculated magnetic field H Cj The partial derivative δ j and the jth calculated magnetic field H Cj and j measured magnetic field H Mj is the product of partial derivatives of

[0115]

[0116] jth error ε i and j calculated magnetic field H Cj The partial differential of is shown in equation (13), and the jth calculated magnetic field H Cj and j measured magnetic field H Mj The partial derivative of is shown in equation (14).

[0117]

[0118]

[0119] From the above, the i-th gradient (∂ε i / ∂I i) is shown in equation (15).

[0120]

[0121] In the gradient descent process shown in S308, the first current I to the Nth current I N are expressed as the i-th current I i As shown in the figure, the i-th gradient (∂ε i / ∂I i ) and the learning rate, also called the step size, each time the gradient descent process is performed. i ) PRESENT is the i-th current I used in the gradient descent process j (I i ) NEW is the i-th current I used in the next gradient descent process. i Shows.

[0122]

[0123] If it is determined that the error ε is equal to or smaller than the determination threshold value (S307-YES), the gradient descent processing execution unit 35 performs the process shown in S308 to optimize the first current I1 to the Nth current I2. N is determined as the current flowing through the conductor 200 at the first sampling time (S309). The processes shown in S310 to S311 are the same as the processes shown in S110 to S111, and therefore detailed description thereof will be omitted here.

[0124] The current distribution estimation device 2 estimates the current inside the conductor 200 based on the gradient descent method, thereby being able to estimate the current distribution of the conductor 200 with high accuracy.

[0125] 11A and 11B are diagrams (part 1) showing the results of current distribution estimation by the current distribution estimation device 2. FIG. 11A shows the first current I1 to the Nth current I N 11(b) shows the relationship between the initial value of the first current I1 to the Nth current I N The first magnetic field H1 to the Nth magnetic field H2 corresponding to the initial values ​​of N and the target magnetic field. N 11(d) shows the relationship between the estimated value of the first current I1 to the Nth current I NThe first magnetic field H1 to the Nth magnetic field H2 corresponding to the estimated values ​​of N and the target magnetic field.

[0126] 12A and 12B are diagrams (part 2) showing the current distribution estimation results by the current distribution estimation device 2. FIG. 12A shows the first current I1 to the Nth current I N 12(b) shows the relationship between the initial value of the first current I1 to the Nth current I N The first magnetic field H1 to the Nth magnetic field H2 corresponding to the initial values ​​of N and the target magnetic field. N 12(d) shows the relationship between the estimated value of the first current I1 to the Nth current I N The first magnetic field H1 to the Nth magnetic field H2 corresponding to the estimated values ​​of N and the target magnetic field.

[0127] 11(a) to 11(d) and 12(a) to 12(d), the horizontal axis indicates the position in the width direction of the conductor 200. In FIGS. 11(a), 11(c), 12(a), and 12(c), the vertical axis indicates the current, and in FIGS. 11(b), 11(d), 12(b), and 12(d), the vertical axis indicates the magnetic field. In FIGS. 11(a) and 12(a), the dots indicate the first current I1 to the Nth current I2. N 11(a), the waveform W401 indicates the target current, and in FIG. 12(a), the waveform W501 indicates the target current. In FIGS. 11(b) and 12(b), the dots indicate the first current I1 to the Nth current I N The first magnetic field H1 to the Nth magnetic field H2 corresponding to the initial values ​​of N 11(b), the waveform W402 indicates the target magnetic field, and in FIG. 12(b), the waveform W502 indicates the target magnetic field. In FIGS. 11(c) and 12(c), the dots indicate the first current I1 to the Nth current I N 11(c), the waveform W403 indicates the target current, and in FIG. 12(a), the waveform W503 indicates the target current. In FIGS. 11(d) and 12(d), the dots indicate the first current I1 to the Nth current I N The first magnetic field H1 to the Nth magnetic field H2 corresponding to the estimated values ​​of NIn FIG. 11(d), a waveform W404 indicates the target magnetic field, and in FIG. 12(d), a waveform W504 indicates the target magnetic field.

[0128] The current distribution estimation process shown in FIGS. 11 and 12 uses a step size η of 1.0×10 －6 The gradient descent process was repeated 100 times.

[0129] As shown in FIGS. 11 and 12, the current distribution estimation device 2 executes current distribution estimation, and the first magnetic field H1 to the Nth magnetic field H N The first current I1 to the Nth current I N is estimated, and the first current I1 to the Nth current I N can be made to match the target value.

[0130] FIG. 13 shows a third diagram illustrating the results of current distribution estimation by the current distribution estimation device 2. FIG. 13( a ) shows the time-dependent change in the total current supplied from the current source 108 to the conductor 200, and FIG. 13( b ) shows the time-dependent change in the estimated value of the total current supplied from the current source 108 to the conductor 200. FIG. 13( c ) shows the time-dependent change in the estimated value of the current distribution supplied from the current source 108 to the conductor 200 (part 1), and FIG. 13( d ) shows the time-dependent change in the estimated value of the current distribution supplied from the current source 108 to the conductor 200 (part 2). In FIGS. 13( a ) and 13 ( b ), the horizontal axis represents time and the vertical axis represents current. In FIG. 13( a ), waveform W601 shows the time-dependent change in the current supplied from the current source 108 to the conductor 200, and in FIG. 13( b ), waveform W602 shows the time-dependent change in the estimated value of the current supplied from the current source 108 to the conductor 200. Waveform W602 1st current I1 to Nth current I N 13(c) and 13(d), the first horizontal axis represents the position in the width direction of the conductor 200, the second horizontal axis represents the time, and the vertical axis represents the current.

[0131] In the current distribution estimation process shown in FIG. 13, the step size η is set to 1.0×10 －8 The gradient descent process was repeated 1000 times.

[0132] As shown in Fig. 13(b), current distribution estimation performed by the current distribution estimation device 2 makes it possible to reproduce the change over time in the current supplied from the current source 108 to the conductor 200. Furthermore, as shown by arrow E in Figs. 13(c) and 13(d), current distribution estimation performed by the current distribution estimation device 2 detects that the current at the end of the conductor 200 is larger than the current at the center of the conductor 200 due to the skin effect. The current distribution estimation device 2 can estimate the current distribution inside the conductor 200 with high accuracy.

[0133] Furthermore, by optimizing the current supplied to the conductor 200 based on the gradient descent method, the current distribution estimation device 2 can reduce the number of steps in the optimization process compared to the current distribution estimation device 1, which optimizes the current supplied to the conductor 200 based on a genetic algorithm. By reducing the number of steps in the optimization process compared to the current distribution estimation device 1, the current distribution estimation device 2 can perform the optimization process faster than the current distribution estimation device 1.

[0134] Furthermore, the current distribution estimation device 2 calculates the error between each of the plurality of magnetic fields acquired by the acquisition unit 22 and each of the plurality of magnetic fields calculated by the calculation unit 23 as a mean square error. By calculating the error as a mean square error, the current distribution estimation device 2 calculates the j-th error ε i and j calculated magnetic field H Cj The partial derivative δ j The jth calculated magnetic field H Cj and the jth measured magnetic field H Mj The current distribution estimation device 2 can use a coefficient proportional to the difference between the j-th error ε i and j calculated magnetic field H Cj The partial derivative δ j The jth calculated magnetic field H Cj and the jth measured magnetic field H Mj By using a coefficient proportional to the difference between and , the calculation of the optimization process can be simplified and the optimization process can be speeded up.

[0135] The current distribution estimation devices 1 and 2 estimate the first current I1 to the Nth current I2 until it is determined that the error ε is equal to or smaller than the determination threshold. NHowever, in the current distribution estimation current device according to the embodiment, the optimization process may be terminated after being repeated a predetermined number of times.

[0136] Fig. 14 is a flowchart of a current distribution estimation process according to a modified example, which is executed by the current distribution estimation device 2. The current distribution estimation process shown in Fig. 14 is executed mainly by the processing unit 30 in cooperation with each element of the current distribution estimation system 100, based on a current distribution estimation program stored in advance in the storage unit 12. The processes shown in S401 to S405 are similar to the processes shown in S101 to S105, and therefore will not be described in detail here.

[0137] Following the process shown in S405, the determination unit 24 determines whether the gradient descent process has been performed a predetermined number of times (S406), thereby determining whether the multiple magnetic fields acquired by the acquisition unit 21 and the multiple magnetic fields calculated by the calculation unit 23 match. When the determination unit 24 determines that the gradient descent process has not been performed a predetermined number of times (S406-NO), the gradient descent process execution unit 35 executes the gradient descent process (S407). Thereafter, the processes shown in S405 to S407 are repeated until the determination unit 24 determines that the gradient descent process has been performed a predetermined number of times (S406-YES). When the determination unit 24 determines that the gradient descent process has been performed a predetermined number of times (S406-YES), the gradient descent process execution unit 35 executes the first current I1 to the Nth current I2 optimized in the process shown in S307. N is determined as the current flowing through the conductor 200 (S408). The processes shown in S409 to S410 are similar to the processes shown in S110 to S111, and therefore detailed description thereof will be omitted here.

[0138] Furthermore, although the current distribution estimation devices 1 and 2 estimate the distribution of the current supplied to a single, flat conductor 200, the current distribution estimation devices according to the embodiments may estimate the distribution of the current supplied to a plurality of conductors. Furthermore, the current distribution estimation devices 1 and 2 estimate the distribution of the current supplied to the conductor 200 by replacing the current supplied to the conductor 200 with a current flowing through a plurality of minute conductors having circular cross sections that are adjacently arranged in a line in a second direction perpendicular to the first direction. However, the current distribution estimation devices according to the embodiments may estimate the distribution of the current supplied to the conductor 200 by replacing the current supplied to the conductor 200 with a current flowing through a plurality of minute conductors having circular cross sections that are adjacently arranged in a plurality of lines in the second direction.

[0139] FIG. 15(a) is a diagram showing a conductor according to a first modified example, FIG. 15(b) is a diagram showing a conductor according to a second modified example, FIG. 15(c) is a diagram showing a conductor according to a third modified example, and FIG. 15(d) is a diagram showing a conductor according to a fourth modified example.

[0140] The conductor 201 according to the first modification is formed by three flat conductors 211 to 213 arranged with their surfaces facing each other. The current distribution estimation system according to the embodiment estimates the distribution of currents flowing through each of the conductors 211 to 213 in accordance with the magnetic field detected by scanning the first sensor head 105 and the second sensor head 106 along the surface of the conductor 211.

[0141] In the conductor 202 according to the second modified example, the current distribution estimation system according to the embodiment estimates the distribution of the current supplied to the conductor 202 by replacing the current with a current flowing through multiple micro-conductors with circular cross sections that are arranged in several rows in the conductor 202.

[0142] A conductor 203 according to the third modification has a pair of wiring patterns 231 and 232 formed on a circuit board 230. The current distribution estimation system according to the embodiment detects a magnetic field by scanning the first sensor head 105 and the second sensor head 106 along the surfaces of the wiring patterns 231 and 231 from the outer end of the wiring pattern 231 to the outer end of the wiring pattern 232. The current distribution estimation system according to the embodiment estimates the distribution of the current flowing through each of the wiring patterns 231 and 232 according to the detected magnetic field.

[0143] The conductor 201 according to the fourth modification is formed by a bus bar 241 arranged on one surface of the thin film 240 and a bus bar 242 arranged on the other surface of the thin film 240. The current distribution estimation system according to the embodiment estimates the distribution of currents flowing through each of the bus bars 241 and 242 in accordance with the magnetic field detected by scanning the first sensor head 105 and the second sensor head 106 along the surface of the bus bar 241.

[0144] Furthermore, while the current distribution estimation devices 1 and 2 estimate the distribution of current supplied to the flat conductor 200, the current distribution estimation device according to the embodiment may also estimate the distribution of current supplied to a conductor in which the current is unevenly distributed, for example, due to its unevenness.

[0145] Furthermore, the current distribution estimation device 1 optimizes the current supplied to the conductor 200 based on a genetic algorithm, and the current distribution estimation device 2 optimizes the current supplied to the conductor 200 based on a gradient descent method. However, the current distribution estimation device according to the embodiment may optimize the current supplied to the conductor 200 based on an optimization algorithm other than the genetic algorithm and the gradient descent method. For example, the current distribution estimation device according to the embodiment may optimize the current supplied to the conductor 200 based on a local search method such as a hill-climbing method and a simulated annealing method.

[0146] The current distribution estimation device 2 also calculates the j-th error ε i is calculated as a mean square error, and the j-th error ε i is not limited to the mean square error, but may be calculated by the least squares method.

[0147] Furthermore, the current distribution estimation system 100 detects a plurality of magnetic fields by scanning the first sensor head 105 and the second sensor head 106 in a second direction perpendicular to the first direction in which the current flows on the surface of the conductor 200. However, the current distribution estimation system according to the embodiment only needs to detect the magnetic fields at detection positions in the second direction that do not overlap each other.

[0148] Furthermore, although the current distribution estimation system 100 has a first magnetic field sensor device 101 and a second magnetic field sensor device 102, the detected magnetic fields in the X and Y directions are in a differential relationship and are correlated with each other, so the current distribution estimation system according to the embodiment may have only a single magnetic field sensor device.

[0149] The current distribution estimation method according to the present disclosure may be in the following aspects: (Aspect 1) A method for estimating the current distribution of a current flowing through a conductor, comprising: an initial condition process for setting initial conditions such as the arrangement and dimensions of the conductor and a magnetic field measurement position; a calculation process for calculating a magnetic field generated at an arbitrary location around the conductor, treating the conductor as a collection of circular conductors; an evaluation process for evaluating an error between magnetic field vector information measured from the magnetic field generated at an arbitrary location around the conductor and the magnetic field vector information obtained by the calculation process; and an optimization process for optimizing the current distribution and the measurement position. (Aspect 2) The current distribution estimation method according to Aspect 1, characterized in that a least-squares method averaging the magnetic field vector information in two directions orthogonal to the direction of current flow is used as the evaluation result. (Aspect 3) The current distribution estimation method according to Aspect 1 or 2, characterized in that an optimal solution for the current distribution is obtained by optimization using a local search algorithm or a genetic algorithm.

Claims

1. A current distribution estimation device comprising: an acquisition unit that acquires magnetic field signals indicating each of a plurality of magnetic fields detected at a plurality of detection positions on a surface of a conductor through which a current flows; an optimization unit that optimizes a current generating the magnetic field detected at each of the plurality of detection positions based on a predetermined optimization algorithm; a calculation unit that calculates a plurality of magnetic fields generated at the plurality of detection positions by the current optimized by the optimization unit; a judgment unit that judges whether the plurality of magnetic fields acquired by the acquisition unit and the plurality of magnetic fields calculated by the calculation unit match; and an output unit that outputs a current signal indicating the current optimized by the optimization unit when the judgment unit judges that the plurality of magnetic fields acquired by the acquisition unit and the plurality of magnetic fields calculated by the calculation unit match.

2. The current distribution estimation device according to claim 1, wherein the calculation unit determines an initial value of a current generating a magnetic field detected at each of the plurality of detection positions, and the optimization unit optimizes the initial value.

3. A current distribution estimation device as described in claim 1, wherein the calculation unit calculates each of the multiple magnetic fields by treating the current generating the magnetic field detected at each of the multiple detection positions as a current flowing through a conductor having a circular cross section.

4. A current distribution estimation device according to any one of claims 1 to 3, wherein the optimization unit optimizes the current based on one of a genetic algorithm, a gradient descent method, and a local search method.

5. The current distribution estimation device according to claim 4, wherein the optimization unit optimizes the current based on a gradient descent method.

6. A current distribution estimation device as described in claim 5, wherein the optimization unit calculates the error between each of the multiple magnetic fields acquired by the acquisition unit and each of the multiple magnetic fields calculated by the calculation unit using the least squares method.

7. A method for estimating current distribution comprising: acquiring magnetic field signals indicating each of a plurality of magnetic fields detected at a plurality of detection positions on a surface of a conductor through which a current flows; optimizing a current generating the magnetic field detected at each of the plurality of detection positions based on a predetermined optimization algorithm; calculating a plurality of magnetic fields generated at the plurality of detection positions by the optimized current; determining whether the acquired plurality of magnetic fields and the calculated plurality of magnetic fields match; and outputting a current signal indicating the optimized current when it is determined that the acquired plurality of magnetic fields and the calculated plurality of magnetic fields match.

8. A current distribution estimation program which causes a computer to execute the following processes: acquiring magnetic field signals indicating each of a plurality of magnetic fields detected at a plurality of detection positions on the surface of a conductor through which a current flows; optimizing a current which generates the magnetic field detected at each of the plurality of detection positions based on a predetermined optimization algorithm; calculating a plurality of magnetic fields which are generated at the plurality of detection positions by the optimized current; determining whether the acquired plurality of magnetic fields and the calculated plurality of magnetic fields match; and outputting a current signal indicating the optimized current when it is determined that the acquired plurality of magnetic fields and the calculated plurality of magnetic fields match.

9. A current distribution estimation system comprising: a magnetic field sensor device that detects a plurality of magnetic fields at a plurality of detection positions on a surface of a conductor through which a current flows and outputs a magnetic field signal indicating each of the detected magnetic fields; and a current distribution estimation device that estimates a current distribution that generates a magnetic field corresponding to the magnetic field signal, wherein the current distribution estimation device comprises: an acquisition unit that acquires the magnetic field signal; an optimization unit that optimizes the current that generates the magnetic field detected at each of the detection positions based on a predetermined optimization algorithm; a calculation unit that calculates a plurality of magnetic fields generated at the detection positions by the current optimized by the optimization unit; a judgment unit that judges whether the plurality of magnetic fields acquired by the acquisition unit and the plurality of magnetic fields calculated by the calculation unit match; and an output unit that outputs a current signal indicating the current optimized by the optimization unit when the judgment unit judges that the plurality of magnetic fields acquired by the acquisition unit and the plurality of magnetic fields calculated by the calculation unit match.

Images