Computing system, learning model, computing method, and computing program

Abstract

A processor of this computing system selects one of a plurality of cells and inputs the position of the selected cell to a cell model, changes the rotation angle of the cell in increments of a prescribed angle and sequentially inputs the rotation angle to thereby compute an analysis value of the magnetic flux density at each rotation angle of the cell, stores the analysis value in a memory for each rotation angle of the cell, changes the position of the cell inputted to the cell model to compute an analysis value of the magnetic flux density at each rotation angle of the plurality of cells, stores the analysis value in the memory for each rotation angle of the cells, acquires a designation condition for the plurality of cells from an input unit, acquires an analysis value of the magnetic flux density that corresponds to the designation condition from the memory, and adds the analysis values of the plurality of cells together to thereby compute an analysis value of the magnetic flux density at any coordinates of a battery.

Description

Computation system, learning model, computation method, and computation program 【0001】 The present disclosure relates to a computation system, a learning model, a computation method, and a computation program. 【0002】 There is a service that rents out batteries, which are rechargeable secondary batteries such as lithium-ion batteries. In this service, a battery is rented out to a user and returned by inserting it into a dedicated charging device at the time of return. The returned battery is charged, and a charged battery is rented out to the user in place of the returned battery. 【0003】 However, in such a service, there is a risk that a user may bring a counterfeit product into the charging device and the genuine product that was rented out may be stolen. In order to deter such a risk, it is conceivable to perform authenticity determination as to whether the rented battery is a genuine product, and for that purpose, it is required to identify the individual lithium-ion battery. 【0004】 A lithium-ion battery is composed of a plurality of battery cells (hereinafter also referred to as "cells"), which are the basic units of the battery. Since the magnetic flux density of the lithium-ion battery is generated by the synthesis of the magnetic flux densities generated from the cells, individual differences occur in the magnetic flux density generated by the lithium-ion battery due to the orientation and individual differences of the cells contained inside the battery. Therefore, it is considered that the individual lithium-ion battery can be identified using the magnetic flux density generated by the lithium-ion battery. 【0005】 Also, as a problem setting similar to the method of identifying an individual lithium-ion battery, face authentication is taken up. In face authentication, it is determined whether an image in which a face is reflected is the same person as the person to be identified. The outline of the configuration of face authentication using deep learning is disclosed in Non-Patent Document 1. 【0006】In this configuration, after preprocessing of facial images, features that distinguish individual differences are extracted from the facial images (feature extraction), and the extracted features are used to determine whether or not they are the same person. This feature extraction is similar in problem setting to the individual identification of lithium-ion batteries. Non-patent document 1 discloses a method for learning features that distinguish individual differences from facial images using machine learning, and it is expected that if features that distinguish individuals can be extracted from the measured magnetic flux density of lithium-ion batteries, individual identification will also be possible. 【0007】 Mei Wang, Weihong Deng, “Deep face recognition: A survey”, Neurocomputing, Volume 429, 2021,Pages 215-244. 【0008】 However, learning features to distinguish individual differences from facial images using machine learning generally requires a large amount of training data. While many datasets of facial images are available for research purposes, no dataset exists that is suitable for identifying lithium-ion batteries. 【0009】 Generally, developing identification technologies, not just for lithium-ion batteries, requires a large amount of measurement data. For example, when building a classifier using machine learning, measurement data is required for training data and performance evaluation. To identify individual lithium-ion batteries, it is necessary to collect magnetic flux density distributions measured multiple times for many batteries. Therefore, a large amount of measurement data is required, both in terms of the number of batteries and the number of measurements per battery. For this reason, manually collecting magnetic flux density generated from lithium-ion batteries is not easy. 【0010】 This disclosure aims to facilitate the generation of measurement data for magnetic flux density generated from lithium-ion batteries. 【0011】The calculation system relating to one aspect is a calculation system for calculating the magnetic flux density of a battery having multiple cells inside, the calculation system includes an input unit, a memory, and a processor, the processor generates a cell model using information of the actual cells, information of a detector that detects the magnetic flux density of the cells, and the magnetic flux density of the cells actually detected by the detector, the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the position and rotation angle of the cell by inputting the position and rotation angle of the cell, and the cell model is configured to select one of the multiple cells and input the position of that cell, The process includes sequentially inputting the rotation angle of a cell in predetermined angle increments to calculate the analytical value of the magnetic flux density at each rotation angle of the cell, storing it in the memory for each rotation angle of the cell, changing the position of the cell to be input into the cell model to calculate the analytical value of the magnetic flux density at each rotation angle of multiple cells, storing it in the memory for each rotation angle of the cell, obtaining specified conditions for multiple cells from the input unit, obtaining the analytical value of the magnetic flux density from the memory according to the specified conditions, and summing the analytical values ​​of multiple cells to calculate the analytical value of the magnetic flux density at an arbitrary coordinate of the battery. 【0012】One aspect of the calculation method is a calculation system for calculating the magnetic flux density of a battery having multiple cells inside, the calculation system includes an input unit, a memory, and a processor, the calculation method executed by the processor generates a cell model using information of the actual cell, information of a detector that detects the magnetic flux density of the cell, and the magnetic flux density of the cell actually detected by the detector, the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the position and rotation angle of the cell by inputting the position and rotation angle of the cell, and selects one of the multiple cells and assigns the position of that cell to the cell model. The process includes inputting data, sequentially inputting the rotation angle of the cell in predetermined angle increments to calculate the analytical value of the magnetic flux density at each rotation angle of the cell, storing it in the memory for each rotation angle of the cell, changing the position of the cell to be input into the cell model to calculate the analytical value of the magnetic flux density at each rotation angle of multiple cells, storing it in the memory for each rotation angle of the cell, obtaining specified conditions for multiple cells from the input unit, obtaining the analytical value of the magnetic flux density according to the specified conditions from the memory, and summing the analytical values ​​of multiple cells to calculate the analytical value of the magnetic flux density at an arbitrary coordinate of the battery. 【0013】The calculation program relating to one aspect is a calculation system for calculating the magnetic flux density of a battery having multiple cells inside, the calculation system includes an input unit, a memory, and a processor, the processor generates a cell model using information of the actual cells, information of a detector that detects the magnetic flux density of the cells, and the magnetic flux density of the cells actually detected by the detector, the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the position and rotation angle of the cell by inputting the position and rotation angle of the cell, and the cell model selects one of the multiple cells and inputs the position of the cell, and the cell The process involves sequentially inputting the rotation angle of the cell in predetermined angle increments to calculate the analytical value of the magnetic flux density at each rotation angle of the cell, storing it in the memory for each rotation angle of the cell, changing the position of the cell input to the cell model to calculate the analytical value of the magnetic flux density at each rotation angle of multiple cells, storing it in the memory for each rotation angle of the cell, obtaining specified conditions for multiple cells from the input unit, obtaining the analytical value of the magnetic flux density according to the specified conditions from the memory, and summing the analytical values ​​of multiple cells to calculate the analytical value of the magnetic flux density at an arbitrary coordinate of the battery. 【0014】 According to one embodiment, measurement data of the magnetic flux density generated from a lithium-ion battery can be generated more easily. 【0015】 This figure shows an example of the configuration of the calculation system according to this embodiment. This figure shows an example of the method for generating current density according to this embodiment. This figure shows an example of generating analysis data for magnetic flux density according to this embodiment. This figure shows an example of generating presentation data for magnetic flux density according to this embodiment. This figure shows an example of visualizing magnetic flux density according to this embodiment. This figure shows an example of the configuration of the information processing device according to this embodiment. This figure shows an example of the hardware configuration of the information processing device according to this embodiment. 【0016】The embodiments of the computation system and learning model disclosed herein will be described in detail below with reference to the drawings. However, the present invention is not limited by these embodiments. Furthermore, the same elements are denoted by the same reference numerals, redundant descriptions are omitted as appropriate, and each embodiment can be combined as appropriate within the bounds of consistency. 【0017】 (Overall Configuration) In this embodiment, for example, as a method for collecting measurement data of magnetic flux density generated from a lithium-ion battery, a mathematical model is created to artificially reproduce the magnetic flux density, thereby generating the magnetic flux density artificially. Furthermore, for example, in order to create a mathematical model that reflects the characteristics of a cell, the current density flowing inside is estimated from the magnetic flux density of the cell, and a model is constructed that predicts the magnetic flux density at an arbitrary point in the cell using the Biot-Savart law from the estimated current density. Here, the mathematical model of the cell does not necessarily need to estimate the current density; if the magnetic flux density of the cell is measured at multiple points, it is also possible to directly calculate the magnetic flux density at an arbitrary point. Furthermore, for example, by using multiple mathematical models of cells, the magnetic flux density generated by a lithium-ion battery can be reproduced. In addition, since a lithium-ion battery is composed of multiple cells, and the magnetic flux density generated by the battery changes depending on the rotation angle of the cells contained within, this embodiment realizes a function that calculates the expected magnetic flux density and presents it to the user when the rotation angle of each cell is specified by the user. 【0018】 A computing system for implementing this embodiment will now be described. Figure 1 is a diagram showing an example configuration of the computing system according to this embodiment. As shown in Figure 1, the computing system 1 according to this embodiment includes, for example, a cell model 100, a magnetic sensor 200, a module model 300, and a database 400. The computing system 1 may also be an information processing device such as a desktop PC (Personal Computer), a notebook PC, a smartphone, or a server computer. 【0019】The cell model 100 is, for example, a model that predicts the magnetic flux density generated by a single cell. The calculation system 1 inputs the "dimensions, position, and orientation" of the cell model 100, the "current density," and the "position and orientation" of the magnetic sensor 200 into a "physical model" that calculates the magnetic flux density generated by a single cell, as shown in Figure 1, and numerically calculates the magnetic flux density observed by the magnetic sensor 200. Here, the "physical model" is, for example, a mathematical model that converts the "current density" into the "magnetic flux density" observed at the position of the magnetic sensor 200. The "dimensions, position, and orientation" of the cell model 100 are, for example, the size, coordinates, and rotation angle of the actual cell for which the magnetic flux density is calculated. The "current density" is, for example, the current density generated inside the cell, which is determined in advance by solving an inverse problem from measured values. The "position and orientation" of the magnetic sensor 200 are, for example, the coordinates of the actual sensor and the orientation of the magnetic sensing axis. 【0020】 Furthermore, the magnetic flux density calculated using the physical model is stored, for example, in the calculation results of the magnetic sensor 200. In actual lithium-ion batteries, multiple cells are arranged inside. Therefore, to match actual lithium-ion batteries, for example, as shown in Figure 1, there are multiple cell models 100 and magnetic sensors 200, and the magnetic flux density is calculated for all combinations of these and stored in the database 400. In the example shown in the lower left of Figure 1, there are N cell models 100, from cell models 100-1 to 100-N, and M magnetic sensors 200, from magnetic sensors 200-1 to 200-M, where N and M can be any natural numbers. 【0021】 The module model 300 is a model that reproduces the magnetic flux density generated by a lithium-ion battery by assembling multiple cell models 100, and also identifies the battery. For this reason, the calculation system 1 pre-calculates the magnetic flux density for each rotation angle of each cell model 100 and stores these magnetic flux densities in a database as "Cell 1 Analysis Value" to "Cell N Analysis Value". 【0022】Then, the calculation system 1, for example, when the user specifies the rotation angle of each cell as a "specified condition," retrieves the magnetic flux density generated at the magnetic sensor 200 when each cell is at the specified rotation angle from the database. The "specified condition" may be, for example, a vector summarizing the rotation angles of each cell. 【0023】 Furthermore, the calculation system 1, for example, uses a "combination process" to superimpose the magnetic flux densities of N cells obtained from the database, calculates the sum of the magnetic flux densities generated by the N cells, adds noise that simulates measurement errors to generate pseudo-data, and presents it to the user. The noise is generated using a "noise generator." Also, for example, "specified conditions" from the user are input via an input device such as a keyboard, and the presentation of pseudo-data to the user is output via an output device such as a display. 【0024】 Furthermore, the "noise generator" generates noise that mimics measurement errors, for example. For instance, it estimates a noise generation distribution from measurement errors occurring in a magnetic sensor beforehand, and generates random numbers as noise from this distribution. Alternatively, for example, the magnetic flux density is measured multiple times by a magnetic sensor while surrounding magnetic field sources are eliminated, and these measurements are fitted to a normal distribution. It is also possible to generate data that simulates multiple measurements of the same battery by adding multiple patterns of noise to a single battery. It should be noted that the noise is not limited to that originating from the magnetic sensor; it is possible to generate simulated data with various patterns of noise. 【0025】Next, using Figure 2, we will explain how to generate the "current density" input to the "physical model" shown in Figure 1. Figure 2 is a diagram showing an example of the current density generation method according to this embodiment. The calculation system 1 generates the magnetic flux density generated by the cell based on measured data, for example, as shown in Figure 2. In this method, for example, the calculation system 1 first measures the magnetic flux density generated around the cell and estimates the current density inside the cell by solving an inverse problem. Then, the calculation system 1 calculates the magnetic flux density generated by the cell using the estimated current density and the Biot-Savart law. Since the Biot-Savart law can calculate the magnetic flux density at any point, the calculation system 1 can, for example, determine the magnetic flux density at a point different from the point where the magnetic flux density was measured when calculating the inverse problem. As shown in Figure 2, the current density is optimized so that the measured value of the magnetic flux density and the analyzed value (calculation result) are equal. 【0026】 Next, using Figure 3, we will explain in more detail the algorithm for generating magnetic flux density (analysis data) by inputting "current density" and other values ​​into the "physical model" of the cell model 100 shown in Figures 1 and 2. Figure 3 is a diagram showing an example of magnetic flux density analysis data generation according to this embodiment. 【0027】 First, as step (1), the cell models 100 are arranged to match the actual battery, for example, as shown on the left side of Figure 3. In the example in Figure 3, six cell models 100, from Cell00 to Cell05, are arranged. On the "physical model," the arrangement of the cell models 100 is specified by inputting values ​​such as the dimensions, position, and orientation of the actual cells into the "dimensions, position, and orientation" fields of the cell models 100, for example, as shown in Figures 1 and 2. 【0028】 Next, in step (2), the magnetic sensor 200 is positioned to match the actual measuring instrument, for example, as shown on the left side of Figure 3. In the "physical model," the position of the magnetic sensor 200 is specified by inputting the actual position and orientation of the magnetic sensor 200 into the "position and orientation" field, for example, as shown in Figures 1 and 2. 【0029】Next, as step (3), for example, in order to calculate the magnetic flux density for each cell model 100, the calculation system 1 substitutes "1" for the parameter n, which specifies which cell model 100 to calculate, in order to perform the calculation for the first cell model 100. 【0030】 Next, as step (4), for example, in order to calculate the magnetic flux density for each cell model 100, the calculation system 1 controls the system to enable only the nth cell model 100 and disable (i.e., treat as if it does not exist) the other cell models 100. 【0031】Next, as step (5), for example, the calculation system 1 rotates the enabled cell model 100 in 1-degree increments for 360 degrees, calculates the magnetic flux density at all magnetic sensors 200 at each rotation angle, and registers it in the database 400. The calculation system 1, for example, substitutes the rotation angle into the "dimensions, position, and orientation" of the cell model 100 and calculates the magnetic flux density for all magnetic sensors 200. Then, as shown on the right side of Figure 3, for example, the calculation system 1 registers these calculation results (analysis values) in the database 400 for each cell model 100 and for each rotation angle. Since these analysis values ​​have up to three components, X, Y, and Z, all analyzed components are registered in the database 400, as shown on the right side of Figure 3. Furthermore, the X, Y, and Z component data are registered in the database 400 as two-dimensional array data, corresponding to the arrangement of the magnetic sensors 200, for example, as shown on the left side of Figure 3, when the magnetic sensors 200 are arranged in a grid (for example, in the example on the left side of Figure 3, the two-dimensional array data is 5 vertically and 13 horizontally). In addition, to reduce the computational load, the calculation of magnetic flux density may be omitted for some of the magnetic sensors 200. For example, since the magnetic flux density generated by a cell decreases as the distance between the cell and the magnetic sensor increases, the calculation of magnetic flux density for magnetic sensors that are relatively far from the cell can be omitted. This reduces the amount of calculation required for magnetic flux density in step (5). More specifically, for example, the calculation system 1 measures the shortest distance between the battery and each magnetic sensor and omits the calculation of magnetic flux density for magnetic sensors located beyond the shortest distance threshold. 【0032】 Next, in step (6), for example, if parameter n reaches the total number of cell models 100, the magnetic flux density analysis data generation process is terminated; otherwise, the process continues to step (7). 【0033】 Next, as step (7), for example, the calculation system 1 substitutes "n+1" for parameter n in order to perform the calculation of the next cell model 100, returns to step (4), and repeats the process. 【0034】Next, we will explain in more detail the magnetic flux density (presented data) generated by inputting specified conditions into the module model 300 shown in Figure 1, using Figure 4. Figure 4 is a diagram showing an example of the generation of presented data for magnetic flux density according to this embodiment. Note that the magnetic flux density analysis data explained using Figure 3 is data that is pre-registered in the database 400, assuming the development stage, whereas the presented data for magnetic flux density explained using Figure 4 is data that is presented when the user inputs the rotation angle of each cell as a specified condition, assuming the time of use. Furthermore, since the magnetic flux density generated by the battery changes depending on, for example, the rotation angle of the cells contained within it, the module model 300 is a function that takes the rotation angle of each cell input as a specified condition as an argument and returns the magnetic flux density for that rotation angle as its return value. 【0035】 First, as procedure (1), for example, the calculation system 1 stores the rotation angles of all cell models 100 input by the user via the input device in the "specified conditions" of the module model 300. In the example in Figure 4, assuming there are six cell models 100, Cell00 to Cell05, the user specifies the rotation angle for each, as indicated by the callout from the user. 【0036】 Next, as step (2), for example, the calculation system 1 searches the database 400 for analysis values ​​at a specified rotation angle for the six cell models 100, Cell00 to Cell05, as shown in Figure 4, and stores each acquired analysis value in the module model 300, from "Cell 1 Analysis Value" to "Cell N Analysis Value". 【0037】 Next, as step (3), for example, the calculation system 1 takes the sum of the magnetic flux densities stored in "Cell 1 Analysis Value" to "Cell N Analysis Value" for each component, as shown in Figure 4. 【0038】 Next, as step (4), for example, as shown in Figure 4, noise simulating measurement error is added to the sum of the analysis values ​​obtained in step (3). 【0039】 Next, as step (5), for example, as shown in Figure 4, the pseudo-data to which noise was added in step (4) is presented to the user via the output device. 【0040】 When presenting the pseudo data of the magnetic flux density to the user, since the position and dimensions of the cell model 100 and the position and orientation of the magnetic sensor 200 are given, it is also possible to visualize and present the magnetic flux density at the position of the magnetic sensor 200. FIG. 5 is a diagram showing an example of the visualization of the magnetic flux density according to the present embodiment. In the example of FIG. 5, the magnetic flux density as the calculation result is visualized by arrows at the set position of the magnetic sensor 200. This visualization method can be similarly performed for the module model 300. 【0041】 As shown in FIG. 5, the magnetic flux density (analytical value) is visualized by arrows at the set position of the magnetic sensor 200 and is shown together with a single cell model 100. Note that the display of the single cell model 100 may be a battery composed of a plurality of aggregated cell models 100. When the battery is displayed, the magnetic flux density visualized by arrows at the positions of the magnetic sensors 200 set for the plurality of cell models 100 constituting the battery may be displayed. 【0042】 Further, for example, it is also possible to reproduce the magnetic flux density from the deteriorated cell by constructing the cell model 100 from the deteriorated cell at the time of constructing the cell model 100. Also, by creating a plurality of cell models 100 from cells of various individuals and mixing them to make a module, it is possible to give variations to the pseudo data. 【0043】Also, for example, as a hardware design issue, reducing the spacing between magnetic sensors and arranging more magnetic sensors increases the cost. Therefore, it is an issue to design an arrangement that can perform authenticity determination while suppressing the number of magnetic sensors. By using a module for such an issue, it becomes possible to determine the optimal number of magnetic sensors. For example, when generating analysis data of magnetic flux density registered in the database 400, by making the arrangement interval of the magnetic sensors 200 as fine as possible and arranging many magnetic sensors 200, it is possible to generate more detailed pseudo data. Also, when generating analysis data of magnetic flux density, thinning out these magnetic sensors 200 to generate pseudo data with a coarse granularity and evaluating the performance of the battery authenticity determination device makes it possible to take a trade-off between the number of magnetic sensors and the authenticity determination performance. 【0044】 (Functional Configuration of Information Processing Device 10) Next, the functional configuration of the information processing device 10, which is an example of the operation system 1 that is the execution entity of the present embodiment, will be described. FIG. 6 is a diagram showing a configuration example of the information processing device 10 according to the present embodiment. As shown in FIG. 6, the information processing device 10 includes a communication unit 20, a storage unit 30, and a control unit 40. 【0045】 The communication unit 20 is a processing unit that controls communication with other devices, and is, for example, a communication interface such as a network interface card. Note that the network for performing communication with other devices may be various communication networks such as the Internet or an intranet, regardless of whether it is wired or wireless. Also, the network is not a single network, and for example, the Internet and an intranet may be configured via a network device such as a gateway or other devices. 【0046】 The storage unit 30 has a function of storing various data and programs executed by the control unit 40, and is realized by a storage device such as a memory or a hard disk, for example. The storage unit 30 stores cell model information 31, magnetic sensor information 32, magnetic flux density analysis information 33, module model information 34, and the like. 【0047】The cell model information 31 stores, for example, information about the cell model 100. The cell model information 31 may store, for example, a "physical model," which is a mathematical model for calculating the magnetic flux density generated by a single cell, as shown in Figure 1. The cell model information 31 may also store, for example, the "dimensions, position, and orientation" of the cell model 100, which are parameters of the "physical model," as shown in Figure 1, and the "current density," which is obtained by solving an inverse problem from measured values ​​in advance. The cell model information 31 may also store, for example, the "magnetic flux density" observed at the position of the magnetic sensor 200, which is converted from the "current density" by the "physical model," as shown in Figure 1. 【0048】 The magnetic sensor information 32 stores, for example, information about the magnetic sensor 200. The magnetic sensor information 32 may store, for example, the "position and orientation" of the magnetic sensor 200, which are parameters of the "physical model" as shown in Figure 1. The cell model information 31 may store, for example, the "calculation result," which is the magnetic flux density observed at the position of the magnetic sensor 200, converted from the "current density" by the "physical model" as shown in Figure 1. 【0049】 The magnetic flux density analysis information 33 is, for example, a database 400 as shown in Figure 1, in which the magnetic flux density calculated for each rotation angle for each cell model 100 is stored as "Cell 1 Analysis Value" to "Cell N Analysis Value". 【0050】Module model information 34 stores, for example, information about module model 300. Module model information 34 may store, for example, "specified conditions," which are the rotation angles of each cell specified by the user, as shown in Figure 1. Module model information 34 may also store, for example, "cell 1 analysis value" to "cell N analysis value," which are the magnetic flux densities for each specified rotation angle of cell model 100, obtained from database 400, as shown in Figure 1. Module model information 34 may also store, for example, a "noise generator," which generates noise that simulates measurement errors, as shown in Figure 1. Module model information 34 may also store, for example, a "combination process," which is a program that calculates the sum of the magnetic flux densities generated by N cells by superimposing the "cell 1 analysis value" to "cell N analysis value" obtained from database 400, as shown in Figure 1, adds the noise generated by the "noise generator," generates pseudo-data, and presents it to the user. 【0051】 The above information stored in the memory unit 30 is merely an example, and the memory unit 30 can store various other types of information besides the above. 【0052】 The control unit 40 is a processing unit that controls the entire information processing device 10, and is, for example, a processor. The control unit 40 includes an arithmetic unit 41, an input unit 42, and a display unit 43. Each processing unit is an example of an electronic circuit in the processor or an example of a process executed by the processor. 【0053】 The calculation unit 41 calculates the magnetic flux density at an arbitrary coordinate using, for example, a cell model 100 that shows the magnetic flux density of a single battery cell and information on multiple cells that make up the module. The information on multiple cells may include, for example, the number of cells that make up the module, the coordinates of the cells, and the orientation of the cells. Furthermore, calculating the magnetic flux density at an arbitrary coordinate includes using the cell model 100 and the cell information to calculate the magnetic flux density at an arbitrary coordinate where the distance from the cell coordinate is less than or equal to a predetermined threshold. This is a reduction in the amount of calculation required for magnetic flux density in procedure (5) when generating analysis data for magnetic flux density, as explained using Figure 3. 【0054】 Furthermore, calculating the magnetic flux density at an arbitrary coordinate includes using multiple cell models and cell information to calculate the magnetic flux density for each coordinate of multiple cells, and then summing up the magnetic flux densities for each coordinate of multiple cells to calculate the magnetic flux density at an arbitrary coordinate. 【0055】 Furthermore, for example, the calculation unit 41 takes the magnetic flux density at multiple coordinates for one calculated module, that is, the analyzed values ​​of the magnetic flux density of one calculated battery, as a single dataset, and generates a learning model by machine learning using multiple datasets as training data. Then, for example, this learning model is used as a classifier to identify individual batteries. 【0056】 The input unit 42 inputs, for example, the number of cells constituting the module, the coordinates of the cells, and the orientation of the cells as cell information to be used by the calculation unit 41. The input unit 42 may obtain this information from, for example, the cell model information 31. 【0057】 The display unit 43 displays, for example, the magnetic flux density at multiple coordinates obtained by the calculation unit 41, along with a diagram of the module or a diagram of the cells constituting the module. For example, as shown in Figure 5, the display unit 43 can visualize and display the analyzed values ​​of multiple magnetic flux densities together with a single cell model 100, using arrows to indicate the positions of the magnetic sensors 200 set on the single cell model 100. The display unit 43 can also display a battery composed of multiple cell models 100, and visualize and display the analyzed values ​​of multiple magnetic flux densities using arrows to indicate the positions of the magnetic sensors 200 set on the multiple cell models 100 constituting the battery. 【0058】 In Figure 6, the information processing device 10 is shown as a single computer, but it may also be a distributed computing system composed of multiple computers. 【0059】(Effects) As described above, in constructing the cell model 100, the current density is estimated from the magnetic flux density generated around the cell when current is actually applied. Therefore, it is possible to construct a mathematical model that reflects the actual characteristics of the cell. Furthermore, by constructing the cell model 100 from the magnetic flux density of a degraded product, it is possible to reproduce the magnetic flux density from a degraded cell. 【0060】 Furthermore, regarding the generation of pseudo-magnetic flux density data, for example, pseudo-magnetic flux density data can be generated by enabling all cells in the module and calculating the magnetic flux density, without creating a database 400 as described above. However, this method requires numerical calculations to be performed each time data for a single battery is generated, which takes time to generate the pseudo-magnetic flux density. If it takes one second to calculate the magnetic flux density for one battery, it would take nearly three hours to generate pseudo-magnetic flux density data for 10,000 batteries. 【0061】 Another method for generating pseudo-magnetic flux density data at high speed is to pre-calculate a large amount of pseudo-magnetic flux density data and use it when needed. However, if each cell can rotate up to 360 degrees in 1-degree increments, the total number of possible magnetic flux densities is 360 to the power of N (where N is the number of cells in the battery). Therefore, it is not practical to calculate and database all of these magnetic flux densities. 【0062】 This disclosure uses a database of only the magnetic flux density generated by individual cells, enabling 360 data points with fewer data. Ｎ All patterns of magnetic flux density can be generated. More specifically, since magnetic flux densities can be superimposed, the magnetic flux density generated by each cell is calculated individually and these are added together to form 360. Ｎ All patterns of magnetic flux density can be generated. In this case, the number of data points that need to be calculated in advance is only the magnetic flux density data for 360 degrees of rotation for each cell, so the number of calculations for the physical model is 360 × N. Thus, if 360 × N calculations are performed, 360 ＮThis will allow us to reproduce the magnetic flux density of each street. Creating database 400 requires 360 x N numerical calculations, but these calculations can be parallelized, making it possible to complete them within a realistic timeframe. 【0063】 Furthermore, as described above, the calculation system 1 includes a calculation unit 41 that calculates the magnetic flux density at an arbitrary coordinate using a cell model 100 that shows the magnetic flux density of a single battery cell and information on multiple cells that constitute a module. 【0064】 As a result, the computing system 1 can generate pseudo-magnetic flux density data with fewer data points and fewer calculations, thus more efficiently generating pseudo-magnetic flux density data from lithium-ion batteries. Furthermore, by creating a database of the calculation results of the physical model, the calculation of the physical model can be omitted during data generation, enabling the generation of pseudo-magnetic flux density data quickly and in large quantities. 【0065】 Furthermore, the calculation system 1 includes an input unit 42 that inputs information about the cells, such as the number of cells constituting the module, the coordinates of the cells, and the orientation of the cells. 【0066】 As a result, the calculation system 1 can generate data for various magnetic flux density patterns with fewer data points and fewer calculations, thus enabling more efficient generation of pseudo-data for magnetic flux density generated from lithium-ion batteries. 【0067】 Furthermore, the calculation of magnetic flux density at an arbitrary coordinate, performed by the calculation system 1, includes using the cell model 100 and cell information to calculate the magnetic flux density at an arbitrary coordinate where the distance from the cell's coordinate is less than or equal to a predetermined threshold. 【0068】 As a result, the calculation system 1 can reduce the amount of calculation required for magnetic flux density, thereby speeding up database generation. 【0069】 Furthermore, the calculation of magnetic flux density at an arbitrary coordinate, performed by the calculation system 1, includes calculating the magnetic flux density at each coordinate of multiple cells using multiple cell models 100 and cell information, and then calculating the magnetic flux density at an arbitrary coordinate by adding up the magnetic flux densities at each coordinate of multiple cells. 【0070】 As a result, the calculation system 1 can perform 360 calculations with fewer data points and fewer calculations. Ｎ Because it can generate data for all magnetic flux density patterns, it can more efficiently generate simulated data of magnetic flux density generated from lithium-ion batteries. 【0071】 Furthermore, the learning model uses the magnetic flux density at multiple coordinates for a single module calculated by the computing system 1 as a single dataset, and uses multiple datasets as training data. 【0072】 As a result, since the learning model can be generated using training data that is produced with fewer data points and fewer computations, it is easier to generate a learning model that can be used as a classifier to identify individual batteries compared to when training data is measured in real-world conditions. 【0073】 Furthermore, the calculation system 1 includes a calculation unit 41 that calculates the magnetic flux density at an arbitrary coordinate using a cell model 100 that shows the magnetic flux density of a single battery cell and information on a plurality of cells that constitute a module, and a display unit 43 that displays the magnetic flux densities at the plurality of coordinates obtained by the calculation unit 41 together with a diagram of the module or a diagram of the cells that constitute the module. 【0074】 As a result, the calculation system 1 can generate magnetic flux density data with fewer data points and calculations, making it easier to generate and present pseudo-data of the magnetic flux density generated from a lithium-ion battery to the user. 【0075】 (Other Embodiments) Now, embodiments of the present disclosure have been described, but the present disclosure may be implemented in various other forms besides those described above. 【0076】 (System) Unless otherwise specified, the processing procedures, control procedures, specific names, and various data and parameters shown in the above documents and drawings may be changed at will. 【0077】Furthermore, the components of each illustrated device are functionally conceptual and do not necessarily need to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown. That is, all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions. 【0078】 Furthermore, each processing function performed by each device may be implemented, in whole or in part, by a CPU (Central Processing Unit) and a program executed by that CPU, or by wired logic hardware. 【0079】 (Hardware) Figure 7 shows an example of the hardware configuration of the information processing device 10 according to this embodiment. As shown in Figure 7, the information processing device 10 has a communication terminal 10a, a non-volatile memory 10b, RAM (Random-Access Memory) 10c, and a processor 10d. Furthermore, each of the parts shown in Figure 7 is interconnected by a bus or the like. 【0080】 The communication terminal 10a is an Ethernet® port or the like, and communicates with other information processing devices such as the administrator device 60, the operator device 70, and the DCS control device. The non-volatile memory 10b is a flash memory or the like, and stores programs and data that operate the functions shown in Figure 6. 【0081】The processor 10d can be a CPU, MPU (Micro Processing Unit), GPU (Graphics Processing Unit), etc. Alternatively, the processor 10d may be implemented using an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or FPGA (Field Programmable Gate Array). The processor 10d also operates threads that perform the functions described in Figure 6 by reading programs that execute the same processing as the processing units shown in Figure 6 from non-volatile memory 10b and loading them into RAM 10c. For example, these threads perform the same functions as the processing units of the information processing device 10. Specifically, the processor 10d reads programs with the same functions as the arithmetic unit 41, input unit 42, and display unit 43 from non-volatile memory 10b. Then, the processor 10d executes threads that perform the same processing as the arithmetic unit 41, input unit 42, and display unit 43. 【0082】 Thus, the information processing device 10 operates as an information processing device that executes various processing methods by reading and executing a program. The program may also be distributed via a network such as the Internet. Furthermore, the program may be recorded on a computer-readable storage medium such as a hard disk, flexible disk (FD), CD-ROM, MO (Magneto-Optical disk), or DVD (Digital Versatile Disc). The program may then be executed by being read from the computer-readable storage medium by the information processing device 10 or the like. 【0083】 (Other) Some examples of combinations of disclosed technical features are listed below. 【0084】(1) A calculation system for calculating the magnetic flux density of a battery having multiple cells inside, wherein the calculation system includes an input unit, a memory, and a processor, wherein the processor generates a cell model using actual cell information, information from a detector that detects the magnetic flux density of the cell, and the magnetic flux density of the cell actually detected by the detector, the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the position and rotation angle of the cell by inputting the position and rotation angle of the cell, the cell model calculates an analytical value of the magnetic flux density at each rotation angle of the cell by selecting one of the multiple cells and inputting the position of that cell, and inputting the rotation angle of the cell sequentially in predetermined angle increments, and stores in the memory for each rotation angle of the cell, the cell model calculates an analytical value of the magnetic flux density at each rotation angle of the multiple cells by changing the position of the cell input to the cell model, and stores in the memory for each rotation angle of the cell, and obtains specified conditions for the multiple cells from the input unit. A calculation system that performs the following: obtaining the analyzed magnetic flux density value from the memory according to the specified conditions, and calculating the analyzed magnetic flux density value at an arbitrary coordinate of the battery by adding up the analyzed values ​​of multiple cells. 【0085】 (2) The specified conditions include the rotation angle, as described in (1). 【0086】 (3) The calculation system according to (1), wherein the information of the cell includes the dimensions, position, and orientation of the cell constituting the battery. 【0087】 (4) A learning model in which the analysis value of the magnetic flux density of one battery calculated by the calculation system described in any one of (1) to (3) is used as one dataset, and multiple datasets are used as training data. 【0088】 (5) A calculation system according to any one of (1) to (3), comprising displaying the obtained plurality of magnetic flux density analysis values ​​together with a drawing of the battery or a drawing of the cells constituting the battery. 【0089】(6) A calculation system for calculating the magnetic flux density of a battery having multiple cells inside, wherein the calculation system includes an input unit, a memory, and a processor, and the calculation method executed by the processor is: to generate a cell model using actual cell information, information from a detector that detects the magnetic flux density of the cell, and the magnetic flux density of the cell actually detected by the detector; the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the position and rotation angle of the cell by inputting the position and rotation angle of the cell; to select one of the multiple cells and input the position of the cell, and sequentially input the rotation angle of the cell by changing it in predetermined angle increments, calculate the analytical value of the magnetic flux density at each rotation angle of the cell and store it in the memory for each rotation angle of the cell; to change the position of the cell input to the cell model and calculate the analytical value of the magnetic flux density at each rotation angle of the multiple cells and store it in the memory for each rotation angle of the cell; and to obtain specified conditions for the multiple cells from the input unit. A calculation method comprising: obtaining an analyzed value of the magnetic flux density according to the specified conditions from the memory, and calculating an analyzed value of the magnetic flux density at an arbitrary coordinate of the battery by adding up the analyzed values ​​of a plurality of cells. 【0090】(7) A calculation system for calculating the magnetic flux density of a battery having multiple cells inside, wherein the calculation system includes an input unit, a memory, and a processor, wherein the processor generates a cell model using actual cell information, information from a detector that detects the magnetic flux density of the cell, and the magnetic flux density of the cell actually detected by the detector; the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the position and rotation angle of the cell by inputting the position and rotation angle of the cell; the cell model calculates an analytical value of the magnetic flux density at each rotation angle of the cell by selecting one of the multiple cells and inputting the position of that cell, and sequentially inputting the rotation angle of the cell in predetermined angle increments, and stores the analytical value of the magnetic flux density at each rotation angle of the cell in the memory for each rotation angle of the cell; and the cell model obtains specified conditions for the multiple cells from the input unit. A calculation program that performs the following steps: obtain the analyzed magnetic flux density value from the memory according to the specified conditions, and calculate the analyzed magnetic flux density value at an arbitrary coordinate of the battery by adding up the analyzed values ​​of multiple cells. 【0091】 (8) A calculation system for calculating the magnetic flux density of a battery having multiple cells inside, wherein the calculation system includes a memory and a processor, and the processor performs the calculation of the magnetic flux density at an arbitrary coordinate using a cell model showing the magnetic flux density of a single cell of the battery and information of the multiple cells constituting a module. 【0092】 (9) The calculation system according to (8), further comprising an input unit for inputting the number of cells constituting the module, the coordinates of the cells, and the orientation of the cells as information of the cells. 【0093】(10) The arithmetic system according to (8) or (9), wherein the processor calculates the magnetic flux density at the arbitrary coordinate, using the cell model and the cell information, the calculation of the magnetic flux density at the arbitrary coordinate where the distance to the coordinate of the cell is less than or equal to a predetermined threshold. 【0094】 (11) The arithmetic system according to any one of (8) to (10), wherein the processor calculates the magnetic flux density at the arbitrary coordinate by calculating the magnetic flux density for each coordinate of the plurality of cells using the plurality of cell models and the cell information, and by adding up the magnetic flux densities for each coordinate of the plurality of cells. 【0095】 (12) A learning model in which the magnetic flux densities at multiple coordinates for one module calculated by the calculation system described in any one of (8) to (11) are used as a single dataset, and the multiple datasets are used as training data. 【0096】 (13) A calculation system for calculating the magnetic flux density of a battery having a plurality of cells inside, the calculation system includes a memory, a processor, and the processor performs the following: calculating the magnetic flux density at an arbitrary coordinate using a cell model showing the magnetic flux density of a single cell of the battery and information of a plurality of the cells constituting a module; and displaying the calculated magnetic flux density at the arbitrary coordinate together with a drawing of the module or a drawing of the cells constituting the module. 【0097】 10 Information processing device 10a Communication terminal 10b Non-volatile memory 10c RAM 10d Processor 20 Communication unit 30 Storage unit 31 Cell model information 32 Magnetic sensor information 33 Magnetic flux density analysis information 34 Module model information 40 Control unit 41 Calculation unit 42 Input unit 43 Display unit 100 Cell model 200 Magnetic sensor 300 Module model 400 Database

Claims

1. A calculation system for calculating the magnetic flux density of a battery having multiple cells inside, wherein the calculation system includes an input unit, a memory, and a processor, the processor generates a cell model using actual cell information, information from a detector that detects the magnetic flux density of the cell, and the magnetic flux density of the cell actually detected by the detector, the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the cell's position and rotation angle by inputting the cell's position and rotation angle, the cell model calculates the analytical value of the magnetic flux density at each rotation angle of the cell by selecting one of the multiple cells and inputting its position, and sequentially changing the cell's rotation angle in predetermined angle increments, and stores it in the memory for each rotation angle of the cell, the cell model calculates the analytical value of the magnetic flux density at each rotation angle of the multiple cells by changing the cell's position input to the cell model, and stores it in the memory for each rotation angle of the cell, and obtains specified conditions for the multiple cells from the input unit. A calculation system that performs the following: obtaining the analyzed magnetic flux density value from the memory according to the specified conditions, and calculating the analyzed magnetic flux density value at an arbitrary coordinate of the battery by adding up the analyzed values ​​of multiple cells.

2. The calculation system according to claim 1, wherein the specified conditions include a rotation angle.

3. The calculation system according to claim 1, wherein the information of the cell includes the dimensions, position, and orientation of the cell constituting the battery.

4. A learning model in which the analysis value of the magnetic flux density of one battery calculated by the calculation system according to any one of claims 1 to 3 is set as one dataset, and multiple datasets are used as training data.

5. The arithmetic system according to any one of claims 1 to 3, further comprising the processor displaying the obtained plurality of magnetic flux density analysis values ​​together with a diagram of the battery or a diagram of the cells constituting the battery.

6. A calculation system for calculating the magnetic flux density of a battery having multiple cells inside, wherein the calculation system includes an input unit, a memory, and a processor, and the calculation method executed by the processor is: to generate a cell model using actual cell information, information from a detector that detects the magnetic flux density of the cell, and the magnetic flux density of the cell actually detected by the detector; the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the position and rotation angle of the cell by inputting the position and rotation angle of the cell; to calculate the analytical value of the magnetic flux density at each rotation angle of the cell by selecting one of the multiple cells and inputting the position of that cell, and sequentially inputting the rotation angle of the cell in predetermined angle increments, and store in the memory for each rotation angle of the cell; to calculate the analytical value of the magnetic flux density at each rotation angle of the multiple cells by changing the position of the cell input to the cell model, and store in the memory for each rotation angle of the cell; and to obtain specified conditions for the multiple cells from the input unit. A calculation method comprising: obtaining an analyzed value of the magnetic flux density according to the specified conditions from the memory, and calculating an analyzed value of the magnetic flux density at an arbitrary coordinate of the battery by adding up the analyzed values ​​of a plurality of cells.

7. A calculation system for calculating the magnetic flux density of a battery having multiple cells inside, wherein the calculation system includes an input unit, a memory, and a processor, and the processor generates a cell model using actual cell information, information from a detector that detects the magnetic flux density of the cell, and the magnetic flux density of the cell actually detected by the detector; the cell model calculates and outputs an analytical value of the magnetic flux density at an arbitrary position corresponding to the cell's position and rotation angle by inputting the cell's position and rotation angle; the cell model calculates an analytical value of the magnetic flux density at each rotation angle of the cell by selecting one of the multiple cells and inputting its position, and sequentially inputting the rotation angle of the cell in predetermined angle increments, and stores it in the memory for each rotation angle of the cell; the cell model calculates an analytical value of the magnetic flux density at each rotation angle of the multiple cells by changing the cell's position input, and stores it in the memory for each rotation angle of the cell; and the input unit obtains specified conditions for the multiple cells. A calculation program that performs the following steps: obtain the analyzed magnetic flux density value from the memory according to the specified conditions, and calculate the analyzed magnetic flux density value at an arbitrary coordinate of the battery by adding up the analyzed values ​​of multiple cells.

Images