JP2021129403A - Abnormality detection device and abnormality detection method - Google Patents

Abstract

To provide an abnormality detection device and an abnormality detection method capable of detecting battery abnormalities with high accuracy.SOLUTION: An abnormality detection device in an embodiment includes an estimation part, a calculation part, and a detection unit. The estimation part estimates a piece of abnormal information indicating the degree of battery abnormality based on battery information including multiple variables related to the battery status. The calculation part generates a probability distribution showing the probability of occurrence of latent variables that occur in the process of estimating abnormal information by the estimation part, and calculates the accuracy information regarding the estimation accuracy of abnormal information based on the probability distribution. The detection unit detects a battery error based on the abnormal information whose accuracy information calculated by the calculation part satisfies a predetermined condition.SELECTED DRAWING: Figure 2

Description

The present invention relates to an abnormality detection device and an abnormality detection method.

Conventionally, for example, an abnormality detection device that detects a deteriorated state of a battery by calculating a resistance value of the battery from a current value and a voltage value of the battery that supplies electric power to a motor that is a driving source of a vehicle is known. In this type of abnormality detection device, there is a technique for calculating the calculation accuracy of the resistance value based on the current value and the voltage value of the battery (see, for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2015-126594

However, in the prior art, the calculation accuracy is calculated based on two variables of the current value and the voltage value of the battery. Therefore, for example, when the number of variables is three or more, the calculation accuracy cannot be calculated, and the abnormality of the battery is highly accurate. There was a risk that it could not be detected.

The present invention has been made in view of the above, and an object of the present invention is to provide an abnormality detecting device and an abnormality detecting method capable of detecting an abnormality of a battery with high accuracy.

In order to solve the above-mentioned problems and achieve the object, the abnormality detection device according to the present invention includes an estimation unit, a calculation unit, and a detection unit. The estimation unit estimates abnormality information indicating the degree of abnormality of the battery based on battery information including a plurality of variables related to the state of the battery. The calculation unit generates a probability distribution indicating the probability of occurrence of a latent variable generated in the process of estimating the abnormality information by the estimation unit, and calculates accuracy information regarding the estimation accuracy of the abnormality information based on the probability distribution. The detection unit detects an abnormality in the battery based on the abnormality information in which the accuracy information calculated by the calculation unit satisfies a predetermined condition.

According to the present invention, abnormalities in the battery can be detected with high accuracy.

FIG. 1A is an explanatory diagram for explaining an outline of the abnormality detection method according to the embodiment. FIG. 1B is an explanatory diagram for explaining an outline of the abnormality detection method according to the embodiment. FIG. 1C is an explanatory diagram for explaining an outline of the abnormality detection method according to the embodiment. FIG. 2 is a block diagram showing a configuration of an abnormality detection device according to an embodiment. FIG. 3 is a diagram showing the processing contents of the estimation unit. FIG. 4 is a diagram showing the processing contents of the calculation unit. FIG. 5 is a diagram showing the processing contents of the threshold value determination unit. FIG. 6 is a diagram showing the processing contents of the threshold value determination unit. FIG. 7 is a flowchart showing a processing procedure of processing executed by the abnormality detection device according to the embodiment.

Hereinafter, embodiments of the abnormality detection device and the abnormality detection method disclosed in the present application will be described in detail with reference to the accompanying drawings. The present invention is not limited to the embodiments shown below.

First, the outline of the abnormality detection method according to the embodiment will be described with reference to FIGS. 1A to 1C. 1A to 1C are explanatory views for explaining the outline of the abnormality detection method according to the embodiment. FIG. 1A shows the abnormality detection system S according to the embodiment. The abnormality detection system S according to the embodiment is mounted on a vehicle, for example.

As shown in FIG. 1A, the abnormality detection system S according to the embodiment includes an abnormality detection device 1 and a battery 101. The battery 101 is, for example, a lithium-ion battery that supplies electric power to a motor that is a drive source of a vehicle. As shown in FIG. 1A, the battery 101 has a plurality of cells C, and each cell C is electrically connected.

Further, the battery 101 has various sensors (not shown), the state of each cell C is detected by such sensors, and the battery information is output to the abnormality detection device 1. Here, the battery information is information including a plurality of variables related to the state of the battery 101 (each cell C).

Specifically, the variables are, for example, the current value of each cell C, the voltage value, the cell temperature, the SOC (State Of Charge), the integrated current capacity, the amount of continuous discharge electricity (current × time), and the like.

The abnormality detection device 1 detects an abnormality in the battery 101 by executing the abnormality detection method according to the embodiment. As shown in FIG. 1A, the abnormality detection device 1 according to the embodiment executes an arithmetic process for calculating the abnormality information based on the battery information and a detection process for detecting the abnormality of the battery 101 based on the abnormality information.

Hereinafter, the outline of the arithmetic processing will be described with reference to FIG. 1B, and the outline of the detection processing will be described with reference to FIG. 1C.

First, the outline of the arithmetic processing will be described with reference to FIG. 1B. As shown in FIG. 1B, in the arithmetic processing, the abnormality information is calculated using the generative model generated by machine learning.

The generative model is a deep learning model, and for example, GAN (Generative Adversarial Networks), VAE (Variational Auto-Encoder), or the like can be used.

In the arithmetic processing, a plurality of variables included in the battery information are used as the input layer of the generative model, and a score (abnormality information) indicating the degree of abnormality is estimated based on the comparison result between the input layer and the output layer of this generative model. ..

Further, in the arithmetic processing, the estimation accuracy of the estimated abnormality information is calculated. Specifically, in the generative model, a latent variable is generated in the process of estimating anomalous information, and the estimation accuracy is calculated using this latent variable.

The latent variable is a variable generated between the input layer and the output layer, and is a variable obtained by compressing an m-dimensional input layer composed of a plurality of variables included in the battery information to n dimensions (m> n).

In the abnormality detection method according to the embodiment, a probability distribution indicating the probability of occurrence of this latent variable is generated, and accuracy information regarding the estimation accuracy of the abnormality information is calculated for each abnormality information based on the probability distribution.

The accuracy information is calculated based on the comparison result between the generated probability distribution and the reference probability distribution (reference distribution) prepared in advance, and the details of the accuracy information will be described later in FIG.

Then, in the abnormality detection method according to the embodiment, only the abnormality information selected based on the calculated accuracy information is output to the detection process as the calculation result. Specifically, in the abnormality detection method according to the embodiment, the abnormality of the battery 101 is detected based on the abnormality information in which the calculated accuracy information satisfies a predetermined condition.

That is, in the abnormality detection method according to the embodiment, only the abnormality information having a relatively high estimation accuracy is targeted for the detection process. As a result, the accuracy of abnormality detection in the detection process can be improved. That is, according to the abnormality detection method according to the embodiment, the abnormality of the battery 101 can be detected with high accuracy.

Next, the outline of the detection process will be described with reference to FIG. 1C. In the detection process, a probability distribution indicating the probability of occurrence of abnormality information for each cell C output from the arithmetic processing is generated for each cell C, and the abnormality of the battery 101 is detected based on the comparison result of the probability distribution for each cell C. .. Note that FIG. 1C shows the probability distributions of the two cells C (cell 1 and cell 2) as an example.

As described above, the values and probabilities of the abnormal information generated as the calculation result are different between the two cells 1 and 2 contained in the one battery 101. This is caused by the difference in characteristics of each cell C, the difference in aging deterioration, the difference in cell temperature, and the like.

Conventionally, a temperature sensor for detecting the cell temperature is provided in each cell, and the threshold value of each cell is determined based on the cell temperature. Therefore, the man-hours for determining the threshold value are increased, which increases the design cost. In addition, the cost of parts for providing temperature sensors in all cells increases.

If the number of temperature sensors is reduced, the accuracy of the cell threshold at which the cell temperature cannot be detected decreases, so that the detection accuracy may also decrease. As described above, conventionally, it has been difficult to achieve both detection accuracy and cost.

Therefore, in the abnormality detection method according to the embodiment, the threshold value of the abnormality information is determined (corrected) for each cell C based on the comparison result of the probability distributions of the abnormality information among the plurality of cells C. Then, in the abnormality detection method according to the embodiment, the presence or absence of an abnormality is detected for each cell C based on the determined threshold value.

That is, in the abnormality detection method according to the embodiment, the relative threshold value between the cells C is determined based on the comparison result of the probability distributions between the cells C, so that the determined threshold value is the state between the cells C. The effect of the difference is excluded. Therefore, for example, by comparing the probability distribution of the cell C provided with the temperature sensor, the threshold value of the cell C not provided with the temperature sensor is excluded from the influence of the difference in the cell temperature.

Therefore, according to the abnormality detection method according to the embodiment, even if the number of temperature sensors is reduced, the accuracy of the threshold value determined in the cell C not provided with the temperature sensor is lower than that of the cell C provided with the temperature sensor. There is nothing to do.

That is, according to the abnormality detection method according to the embodiment, it is possible to reduce the cost while ensuring the abnormality detection accuracy.

Next, the configuration of the abnormality detection device 1 according to the embodiment will be described with reference to FIG. FIG. 2 is a block diagram showing the configuration of the abnormality detection device 1 according to the embodiment.

As shown in FIG. 2, the abnormality detection device 1 according to the embodiment is connected to the battery 101 and various sensors 102.

The various sensors 102 are a group of sensors that measure the state of the battery 101. The various sensors 102 include, for example, a current sensor that measures the current of each cell C, a voltage sensor that measures the voltage of each cell C, a temperature sensor that measures the cell temperature of each cell C, and the like. It should be noted that the various sensors 102 do not have to be provided in all of the plurality of cells C, but may be provided in at least one cell C.

Subsequently, the abnormality detection device 1 according to the embodiment includes a control unit 2 and a storage unit 3. The control unit 2 includes a calculation unit 21 and a detection unit 22. The storage unit 3 stores the reference distribution information 31 and the threshold value information 32.

Here, the abnormality detection device 1 includes, for example, a computer having a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), a flash memory, an input / output port, and various circuits.

The CPU of the computer functions as the calculation unit 21 and the detection unit 22 of the control unit 2, for example, by reading and executing the program stored in the ROM.

Further, at least one or all of the calculation unit 21 and the detection unit 22 of the control unit 2 may be configured by hardware such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

Further, the storage unit 3 corresponds to, for example, a RAM or a flash memory. The RAM or flash memory can store reference distribution information 31, threshold value information 32, information on various programs, and the like. The abnormality detection device 1 may acquire the above-mentioned program and various information via another computer or a portable recording medium connected by a wired or wireless network.

Here, the reference distribution information 31 stored in the storage unit 3 is information on a probability distribution (reference distribution) that serves as a reference indicating the probability of occurrence of a latent variable, and is used by the calculation unit 213 described later. The reference distribution information 31 is information generated in advance, and is generated based on, for example, battery information obtained when the battery 101 is installed (replaced), when the battery 101 is charged from an external power source, or the like.

Further, the threshold value information 32 is information used for determining the threshold value of the abnormal information by the threshold value determination unit 221 described later. The threshold information 32 is generated based on, for example, a result obtained in advance by an experiment or the like.

Next, each function of the control unit 2 (calculation unit 21 and detection unit 22) will be described.

The calculation unit 21 calculates the abnormality information based on the battery information acquired based on the detection values measured by the various sensors 102. As shown in FIG. 2, the calculation unit 21 includes an acquisition unit 211, an estimation unit 212, and a calculation unit 213.

The acquisition unit 211 acquires battery information based on the detected values measured by the various sensors 102. The battery information is information on the state of the battery 101, and specifically, information on the state of each cell C.

The battery information includes a plurality of variables related to the state of the battery 101 (each cell C). Specifically, the variables are, for example, the current value of each cell C, the voltage value, the cell temperature, the SOC (State Of Charge), the integrated current capacity, the amount of continuous discharge electricity (current × time), and the like.

The acquisition unit 211 may acquire the SOC, current integrated capacity, and continuous discharge electricity amount of the battery information by self-calculating from the current value or voltage value of each cell C, or by using an external device or the like. It may be acquired as a calculated result.

The estimation unit 212 estimates the abnormality information based on the battery information acquired by the acquisition unit 211. Here, the processing of the estimation unit 212 will be described with reference to FIG.

FIG. 3 is a diagram showing the processing contents of the estimation unit 212. FIG. 3 describes a case where abnormality information is estimated by inputting three variables included in the battery information into the generative model. Further, in FIG. 3, a case where VAE is used as a generative model is taken as an example.

The generative model as a VAE is composed of five layers, for example, as shown in FIG. The first layer L1 is an input layer into which variables are input. The second layer L2 is an encoder layer and is a layer that compresses the dimensions of the input layer to a predetermined number of dimensions. The third layer L3 is a layer of latent variables, which is a layer of variables compressed to a predetermined number of dimensions by an encoder layer.

Further, the latent variable in the third layer L3 is a variable indicating the characteristics of the dimensionally compressed battery information. The fourth layer L4 is a layer of a decoder, and is a layer that restores a predetermined number of dimensional latent variables to variables of the same dimension as the input layer. The fifth layer L5 is an output layer, and variables Aa, Bb, and Cc correspond to variables A, B, and C of the input layer, respectively. In FIG. 3, the latent variable is one-dimensional, but it may be two-dimensional or more.

The estimation unit 212 estimates as abnormal information a score according to the comparison result between the variables A, B and C of the input layer and the variables Aa, Bb and Cc of the output layer. Such a score is, for example, a score calculated based on the difference between the variable A and the variable Aa, the difference between the variable B and the variable Bb, and the difference between the variable C and the variable Cc.

Returning to FIG. 2, the calculation unit 213 will be described. The calculation unit 213 generates a probability distribution indicating the probability of occurrence of a latent variable generated in the process of estimating the abnormality information by the estimation unit 212, and calculates accuracy information regarding the estimation accuracy of the abnormality information based on the probability distribution.

Here, the processing of the calculation unit 213 will be described with reference to FIG. FIG. 4 is a diagram showing the processing contents of the calculation unit 213. FIG. 4 shows a probability distribution (solid line) showing the probability of occurrence of a latent variable in a generative model in which battery information is input, and a reference distribution (broken line) in the reference distribution information 31.

The calculation unit 213 calculates the difference information indicating the difference between the reference distribution and the probability distribution as the accuracy information. In the example shown in FIG. 4, the amount of KL (Kullback-Leibler) information is calculated as the difference information. Note that the difference information is not limited to the KL information amount, e.g., Pearson distance and the relative Pearson distance may be L ² distance, or the like.

Further, the calculation unit 213 is not limited to the difference information, and may calculate, for example, a score or the like indicating whether the probability distribution is on the positive side or the negative side with respect to the reference distribution as the accuracy information. That is, the calculation unit 213 may be any accuracy information calculated based on the comparison result between the reference distribution and the probability distribution.

In this way, by calculating the accuracy information based on the comparison result (difference) with the reference distribution, it is possible to calculate the highly accurate accuracy information.

Then, the calculation unit 213 outputs the abnormality information that the calculated accuracy information satisfies a predetermined condition to the detection unit 22. For example, the calculation unit 213 outputs the abnormality information in which the amount of KL information, which is the accuracy information, falls within a predetermined range to the detection unit 22.

That is, the detection unit 22 detects an abnormality based on the abnormality information in which the amount of KL information falls within a predetermined range. As a result, the detection unit 22 can perform abnormality detection based on abnormality information having high estimation accuracy, so that the accuracy of abnormality detection can be improved.

Returning to FIG. 2, the detection unit 22 will be described. The detection unit 22 generates a probability distribution indicating the probability of occurrence of the abnormality information calculated by the calculation unit 21 for each cell C, and detects the abnormality of the battery 101 based on the comparison result of the probability distribution for each cell C. As shown in FIG. 2, the detection unit 22 includes a threshold value determination unit 221, a determination unit 222, and a counter unit 223.

The threshold value determination unit 221 determines the threshold value to be used in the subsequent determination unit 222. Here, the processing of the threshold value determination unit 221 will be described with reference to FIGS. 5 and 6.

5 and 6 are diagrams showing the processing contents of the threshold value determination unit 221.

As shown in FIG. 5, the threshold value determination unit 221 first generates a probability distribution indicating the probability of occurrence of abnormal information for each cell C. FIG. 5 shows the probability distributions of the three cells 1-3.

Subsequently, the threshold value determination unit 221 determines the probability distribution of the reference cell C among the plurality of cells C as the reference distribution. In FIG. 5, it is assumed that the probability distribution of cell 2 is determined as the reference distribution (broken line).

Subsequently, the threshold value determination unit 221 determines the threshold value based on the comparison result between the reference distribution of the cell 2 and the probability distribution of the other cells 1 and 3. For example, the threshold value determination unit 221 calculates the amount of KL information indicating the difference between the reference distribution and the probability distributions of the other cells 1 and 3, and corrects the reference reference threshold value based on the calculated amount of KL information. ..

The reference threshold is, for example, the threshold set in cell 2. Specifically, the larger the amount of KL information, the larger the correction value for correcting the reference threshold value.

Threshold determination unit 221 is not limited to the KL information amount, e.g., Pearson distance and the relative Pearson distance may determine the threshold by L ² distance, or the like.

Alternatively, the threshold value determination unit 221 makes a correction of adding a fixed value from the reference threshold value when the probability distribution is located on the positive side with respect to the reference distribution, and sets a fixed value from the reference threshold value when it is located on the negative side. The correction to be subtracted may be performed.

In this way, by correcting the reference threshold value according to the states of the other cells 1 and 3 based on the state of the cell 2, the threshold value excluding the difference between the states of the cell 2 and the cells 1 and 3 can be obtained. Since the determination can be made, the accuracy of the abnormality determination by the determination unit 222 in the subsequent stage can be improved.

The cell 2 having the reference distribution is preferably the cell C provided with the temperature sensor, for example. As a result, since the probability distribution of the cell C provided with the cell temperature is used as a reference, the threshold value excluding the cell temperature having a relatively large influence on the abnormality information can be determined. Therefore, for each cell C by the determination unit 222 in the subsequent stage. Judgment variation due to cell temperature can be suppressed.

Further, the threshold value determination unit 221 may perform a process of eliminating the characteristics of aging change for each cell C and the influence of aging deterioration. This point will be described with reference to FIG.

In FIG. 6, the vertical axis shows the “threshold value” and the horizontal axis shows the threshold information 32 of the two-dimensional graph in which the “ΔKL information amount”. The "threshold value correction value" indicates a reference threshold value or a correction value from the threshold value after the above-mentioned correction.

“ΔKL information” is the difference between the initial value and the latest value in the amount of KL information. The initial value of the KL information amount is, for example, the KL information amount calculated at the end of the previous trip or the KL information amount calculated at the latest start.

The threshold value determination unit 221 corrects the threshold value for each cell C with a threshold value correction value according to the difference between the initial value and the latest value of the KL information amount, so that the characteristics of the secular change for each cell C and the influence of the secular deterioration Can be eliminated.

In other words, in the subsequent threshold value determination process of the threshold value determination unit 221, only the heat receiving variation for each cell C needs to be considered (that is, the cell C in which the temperature sensor is located may be used as the reference distribution). The accuracy of the threshold can be improved.

The determination unit 222 determines the presence or absence of an abnormality in the battery 101 by using the threshold value for each cell C determined by the threshold value determination unit 221. Specifically, the determination unit 222 determines that the cell C corresponding to the score is abnormal when the score indicating the degree of abnormality is equal to or higher than the threshold value, and corresponds to the score when the score is less than the threshold value. It is determined that the cell C to be used is normal.

Then, the determination unit 222 outputs the determination result for each cell C to the counter unit 223.

The counter unit 223 updates the counter value for each cell C based on the determination result of the determination unit 222, and when the counter value of one or more (or a predetermined number or more) of cells C becomes a predetermined value or more. Detects that the battery 101 is abnormal.

Specifically, the counter unit 223 counts up the counter value for the cell C determined to be abnormal by the determination unit 222. Further, the counter unit 223 maintains (or clears) the counter value for the cell C determined to be normal by the determination unit 222.

Next, the processing procedure of the processing executed by the abnormality detection device 1 according to the embodiment will be described with reference to FIG. 7. FIG. 7 is a flowchart showing a processing procedure of processing executed by the abnormality detection device 1 according to the embodiment.

As shown in FIG. 7, first, the calculation unit 21 acquires the battery information of each of the plurality of cells C of the battery 101 (step S101).

Subsequently, the calculation unit 21 estimates the abnormality information for each cell C based on the acquired battery information (step S102).

Subsequently, the calculation unit 21 generates a probability distribution indicating the probability of occurrence of the latent variable generated in the process of estimating the abnormality information, and calculates accuracy information regarding the accuracy of the abnormality information based on the probability distribution (step S103).

Subsequently, the calculation unit 21 determines whether or not the accuracy information of the abnormality information satisfies a predetermined condition (step S104), and when the accuracy information of the abnormality information does not satisfy the predetermined condition (step S104: No). The process ends without performing an abnormality judgment.

On the other hand, when the accuracy information of the abnormality information satisfies a predetermined condition (step S104: Yes), the calculation unit 21 outputs the abnormality information to the detection unit 22, and the detection unit 22 has a probability distribution indicating the probability of occurrence of the abnormality information. Is generated (step S105).

Subsequently, the detection unit 22 determines the threshold value of the abnormality information based on the generated probability distribution (step S106).

Subsequently, the detection unit 22 determines whether or not the abnormality information is equal to or greater than the threshold value (step S107).

When the abnormality information is equal to or greater than the threshold value (step S107: Yes), the detection unit 22 counts up the counter value (step S108).

Subsequently, the detection unit 22 determines whether or not the counter value is equal to or greater than the predetermined value (step S109), and if the counter value is equal to or greater than the predetermined value (step S109: Yes), detects an abnormality in the battery 101. (Step S110), the process ends.

On the other hand, when the counter value is less than a predetermined value (step S109: No), the detection unit 22 ends the process.

Further, in step S107, when the abnormality information is less than the threshold value (step S107: No), the detection unit 22 keeps counting the counter value (step S111) and ends the process.

As described above, the abnormality detection device 1 according to the embodiment includes an estimation unit 212, a calculation unit 213, and a detection unit 22. The estimation unit 212 estimates the abnormality information indicating the degree of abnormality of the battery 101 based on the battery information including a plurality of variables related to the state of the battery 101. The calculation unit 213 generates a probability distribution indicating the probability of occurrence of a latent variable generated in the process of estimating the abnormality information by the estimation unit 212, and calculates accuracy information regarding the estimation accuracy of the abnormality information based on the probability distribution. The detection unit 22 detects an abnormality in the battery 101 based on the abnormality information in which the accuracy information calculated by the calculation unit 213 satisfies a predetermined condition. As a result, the abnormality of the battery 101 can be detected with high accuracy.

Further, as described above, the abnormality detection device 1 according to the embodiment includes a calculation unit 21 and a detection unit 22. The calculation unit 21 calculates the abnormality information regarding the degree of abnormality for each cell C based on the battery information which is the information regarding the state of each of the plurality of cells C included in the battery 101. The detection unit 22 generates a probability distribution indicating the probability of occurrence of the abnormality information calculated by the calculation unit 21 for each cell C, and detects the abnormality of the battery 101 based on the comparison result of the probability distribution for each cell C. As a result, the cost can be reduced while ensuring the abnormality detection accuracy.

Further effects and variations can be easily derived by those skilled in the art. For this reason, the broader aspects of the invention are not limited to the particular details and representative embodiments expressed and described above. Therefore, various modifications can be made without departing from the spirit or scope of the general concept of the invention as defined by the appended claims and their equivalents.

1 Anomaly detection device 2 Control unit 3 Storage unit 21 Calculation unit 22 Detection unit 31 Reference distribution information 32 Threshold information 101 Battery 211 Acquisition unit 212 Estimate unit 213 Calculation unit 221 Threshold determination unit 222 Judgment unit 223 Counter unit S Abnormality detection system

Claims (4)

An estimation unit that estimates abnormality information indicating the degree of abnormality of the battery based on battery information including a plurality of variables related to the battery state, and an estimation unit.
A calculation unit that generates a probability distribution indicating the probability of occurrence of a latent variable generated in the process of estimating the abnormality information by the estimation unit, and calculates accuracy information regarding the estimation accuracy of the abnormality information based on the probability distribution.
An abnormality detection device including a detection unit that detects an abnormality in the battery based on the abnormality information in which the accuracy information calculated by the calculation unit satisfies a predetermined condition. The calculation unit
The abnormality according to claim 1, wherein a reference distribution, which is a reference probability distribution, is acquired in advance, and the accuracy information is calculated based on a comparison result between the reference distribution and the generated probability distribution. Detection device. The calculation unit
As the accuracy information, difference information indicating the difference between the reference distribution and the probability distribution is calculated.
The detection unit
The abnormality detection device according to claim 2, wherein the abnormality is detected based on the abnormality information in which the difference information falls within a predetermined range. An estimation step for estimating abnormality information indicating the degree of abnormality of the battery based on battery information including a plurality of variables related to the battery state, and
A calculation step of generating a probability distribution indicating the probability of occurrence of a latent variable generated in the process of estimating the anomaly information by the estimation step, and calculating accuracy information regarding the estimation accuracy of the anomaly information based on the probability distribution.
An abnormality detection method, wherein the accuracy information calculated by the calculation step includes a detection step of detecting an abnormality of the battery based on the abnormality information satisfying a predetermined condition.

Images