Battery diagnosis device and operation method thereof

Abstract

A battery diagnosis device according to one embodiment disclosed in the present document may comprise: an encoder for converting time series data related to a battery state into a latent variable; a decoder for converting the latent variable into first reconstruction data of a time domain; a first discriminator for discriminating between the time series data and the first reconstruction data and outputting a discrimination result; a first calculator for comparing the time series data with the first reconstruction data and calculating a reconstruction error; a second calculator for calculating an anomaly score on the basis of the determination result and the reconstruction error; and a diagnostic unit for diagnosing an anomaly of the battery on the basis of the anomaly score and a threshold value.

Description

Battery diagnostic device and its operating method

[0001] Cross-citation with related applications

[0002] This invention claims the benefit of priority from Korean Patent Application No. 10-2023-0080445, filed on June 22, 2023, the entire disclosure of which is incorporated herein by reference.

[0003] Technology field

[0004] The embodiments disclosed in this document relate to a battery diagnostic device and an operating method thereof.

[0005] Recently, research and development on secondary batteries has been actively conducted. Here, secondary batteries are rechargeable and include both conventional Ni / Cd and Ni / MH batteries, as well as recent lithium-ion batteries. Among secondary batteries, lithium-ion batteries have the advantage of having a much higher energy density than conventional Ni / Cd and Ni / MH batteries. Furthermore, lithium-ion batteries can be manufactured in a compact and lightweight form, making them suitable for use as power sources for mobile devices. Recently, their use has expanded to include power sources for electric vehicles, attracting attention as a next-generation energy storage medium.

[0006] Additionally, secondary batteries can be utilized as battery packs, which typically include battery modules in which multiple battery cells are connected in series and / or parallel. Furthermore, secondary batteries can be utilized as battery racks, which include multiple battery modules and a rack frame that accommodates these battery modules.

[0007] Battery cells, battery modules, battery packs, or battery racks like these can be utilized in a variety of devices. For example, batteries can be used in mobile devices such as cell phones, laptops, smartphones, and tablets, as well as in electric vehicles (EVs, HEVs, PHEVs) and large-capacity energy storage systems (ESS).

[0008] These batteries can have their status and operation managed and controlled by a battery management system (BMS). The BMS can be included with the batteries in a single device.

[0009] Additionally, the battery management system can manage and control the battery while being separated from the device containing the battery. For example, the battery management system can be implemented as a separate server device. In this case, the battery management system can collect battery data and vehicle data from vehicles and other devices, and utilize the collected data to manage and control the battery.

[0010] If a short circuit or other type of failure occurs within a battery, the risk of damage to devices containing the battery (e.g., EVs, ESS) may increase. Therefore, a method is needed to detect abnormal battery conditions and reduce the risk of damage to devices containing the battery.

[0011] Long-term cycle testing of existing batteries compared the pre-test capacity with the post-test capacity to diagnose battery degradation or abnormalities. Because these tests were performed over a long period of time—at least 100 cycles—it was difficult for humans to continuously monitor the condition. Furthermore, because time-series data collected included sensor noise, many misdiagnoses occurred.

[0012] The embodiments disclosed in this document can provide a battery diagnosis device and an operating method thereof that can diagnose a battery abnormality by using an artificial intelligence model that has learned a previously unlearned pattern as an abnormal pattern, taking into account instantaneous fluctuations in a battery status value.

[0013] The embodiments disclosed in this document can provide a battery diagnosis device and an operating method thereof that can diagnose a battery abnormality based on the correlation between battery status variables by learning various variables as input factors as well as the voltage of the battery.

[0014] The technical problems of the embodiments disclosed in this document are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the descriptions below.

[0015] A battery diagnosis device according to an embodiment disclosed in the present document may include an encoder that converts time series data related to a state of a battery into a latent variable, a decoder that converts the latent variable into first reconstructed data in a time domain, a first discriminator that discriminates the time series data and the first reconstructed data and outputs a discrimination result, a first calculator that compares the time series data and the first reconstructed data to calculate a reconstruction error, a second calculator that calculates an anomaly score based on the discrimination result and the reconstruction error, and a diagnostic device that diagnoses an anomaly of the battery based on the anomaly score and a threshold value.

[0016] The battery diagnostic device according to one embodiment disclosed in the present document may further include a second discriminator that determines the performance of the encoder based on the latent variable and random noise in the latent space.

[0017] In a battery diagnosis device according to an embodiment disclosed in this document, the decoder may convert random noise in a latent space into second reconstructed data in a time domain, and the first discriminator may output the discrimination result based on a discrimination model learned to discriminate between the time series data and the second reconstructed data.

[0018] In a battery diagnostic device according to an embodiment disclosed in this document, the first calculator can calculate the reconstruction error according to a combined comparison method that compares the time series data and the first reconstruction data over the entire time period included in the time series data.

[0019] In a battery diagnosis device according to an embodiment disclosed in this document, the first calculator can calculate the reconstruction error according to a segmented comparison method that compares the time series data with the first reconstruction data in a designated portion of the entire time interval included in the time series data.

[0020] In the battery diagnostic device according to one embodiment disclosed in this document, the second calculator can calculate the abnormality score based on the following mathematical expression 1.

[0021] [Mathematical Formula 1]

[0022]

[0023] (In Equation 1, α is the weight, Z reconstruction is the above reconstruction error, Z Cx is the result of the above determination.)

[0024] In a battery diagnostic device according to an embodiment disclosed in this document, the second calculator can calculate the abnormality score based on the following mathematical expression 2.

[0025] [Equation 2]

[0026]

[0027] (In Equation 2, α is the weight, Z reconstruction is the above reconstruction error, Z Cx is the result of the above determination.)

[0028] In a battery diagnosis device according to an embodiment disclosed in this document, the diagnosis device increases the threshold value and compares the abnormality score and the threshold value to diagnose an abnormality of the battery, and if the threshold value is equal to or greater than a specified value, the diagnosis device can determine an optimal threshold value based on all threshold values ​​used in diagnosing an abnormality of the battery.

[0029] In a battery diagnostic device according to an embodiment disclosed in this document, the diagnostic device can calculate AUROC (Area Under ROC curve) values ​​corresponding to each of the above threshold values, and determine a threshold value corresponding to the maximum AUROC value among the AUROC values ​​as the optimal threshold value.

[0030] A battery diagnosis method according to an embodiment disclosed in the present document may include an operation of converting time series data related to a state of a battery into a latent variable, an operation of converting the latent variable into first reconstructed data in a time domain, an operation of determining the time series data and the first reconstructed data to produce a determination result, an operation of comparing the time series data and the first reconstructed data to produce a reconstruction error, an operation of calculating an anomaly score based on the determination result and the reconstruction error, and an operation of diagnosing an anomaly of the battery based on the anomaly score and a threshold value.

[0031] A battery diagnosis method according to an embodiment disclosed in this document may further include an operation of converting random noise in a latent space into second reconstructed data in a time domain, and the operation of calculating the determination result may include an operation of calculating the determination result based on a determination model learned to determine the time series data and the second reconstructed data.

[0032] In a battery diagnosis method according to an embodiment disclosed in this document, the operation of calculating the reconstruction error may include an operation of calculating the reconstruction error according to a combined comparison method of comparing the time series data and the first reconstruction data in the entire time interval included in the time series data.

[0033] In a battery diagnosis method according to an embodiment disclosed in this document, the operation of calculating the reconstruction error may include an operation of calculating the reconstruction error according to a segmented comparison method of comparing the time series data with the first reconstruction data in a designated portion of the entire time interval included in the time series data.

[0034] A battery diagnosis method according to an embodiment disclosed in this document may further include an operation of diagnosing an abnormality of the battery by increasing the threshold value and comparing the abnormality score and the threshold value, and an operation of determining an optimal threshold value based on all threshold values ​​used in diagnosing an abnormality of the battery if the threshold value is equal to or greater than a specified value.

[0035] In a battery diagnosis method according to an embodiment disclosed in this document, the operation of determining the optimal threshold value comprises:

[0036] It may include an operation of calculating AUROC (Area Under ROC curve) values ​​corresponding to each of the above threshold values, and an operation of determining a threshold value corresponding to the maximum AUROC value among the above AUROC values ​​as the optimal threshold value.

[0037] According to the embodiments disclosed in this document, the manpower and time consumed in diagnosing battery abnormalities can be reduced.

[0038] According to the embodiments disclosed in this document, compensation costs incurred in the field can be reduced by detecting abnormal patterns of batteries that were previously undetected.

[0039] In addition, various effects may be provided, either directly or indirectly, through this document.

[0040] FIG. 1 is a block diagram of a battery diagnostic device according to one embodiment.

[0041] FIG. 2 is a diagram illustrating a battery diagnosis system according to one embodiment.

[0042] FIG. 3 is a diagram illustrating various examples of how a battery diagnosis system according to one embodiment calculates a reconstruction error.

[0043] Figure 4 is a flowchart illustrating the operation of a battery diagnostic device according to one embodiment.

[0044] Figure 5 is a flowchart of the operation of a battery diagnostic device according to one embodiment.

[0045] Figure 6 is a flowchart illustrating the operation of a battery diagnostic device according to one embodiment.

[0046] Hereinafter, various embodiments of the present invention will be described with reference to the attached drawings. However, this is not intended to limit the present invention to specific embodiments, and it should be understood that the present invention encompasses various modifications, equivalents, and / or alternatives of the embodiments.

[0047] The various embodiments and terminology used in this document are not intended to limit the technical features described in this document to specific embodiments, but should be understood to encompass various modifications, equivalents, or alternatives of the embodiments. In connection with the description of the drawings, similar reference numerals may be used to refer to similar or related components. The singular form of a noun corresponding to an item may include one or more of the items, unless the context clearly indicates otherwise.

[0048] In this document, the phrases "A or B", "at least one of A and B", "at least one of A or B", "A, B, or C", "at least one of A, B, and C", and "at least one of A, B, or C" can each include any one of the items listed together in that phrase, or all possible combinations thereof. Terms such as "first", "second", "first", "second", "A", "B", "(a)", or "(b)" may be used merely to distinguish the corresponding element from other corresponding elements, and do not limit the corresponding elements in any other respect (e.g., importance or order) unless specifically stated otherwise.

[0049] In this document, whenever a component (e.g., a first component) is referred to as being “connected,” “coupled,” or “connected,” with or without the terms “functionally” or “communicatively,” or “coupled” or “connected,” it means that the component can be connected to the other component directly (e.g., wired), wirelessly, or via a third component.

[0050] According to various embodiments, each component (e.g., a module or a program) of the above-described components may include one or more entities, and some of the entities may be separated and placed in other components. According to various embodiments, one or more components or operations of the aforementioned components may be omitted, or one or more other components or operations may be added. Alternatively or additionally, a plurality of components (e.g., a module or a program) may be integrated into a single component. In such a case, the integrated component may perform one or more functions of each of the plurality of components identically or similarly to those performed by the corresponding component among the plurality of components prior to the integration. According to various embodiments, the operations performed by a module, program, or other component may be executed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be executed in a different order, omitted, or one or more other operations may be added.

[0051] FIG. 1 is a block diagram of a battery diagnostic device according to one embodiment.

[0052] Referring to FIG. 1, a battery diagnostic device (110) can be connected to an electronic device (120) and a user terminal (130) via wires and / or wirelessly.

[0053] The operation of the battery diagnostic device (110) to be described below can be performed by a BMS (Battery Management System) in a vehicle, as well as by various devices such as a server, cloud, charger, or charger / discharger.

[0054] According to one embodiment, the connection (101) between the battery diagnostic device (110) and the electronic device (120) may be a communication connection via a wired and / or wireless network. In one embodiment, the wired network may be based on a local area network (LAN) communication or a power line communication. In one embodiment, the wireless network may be based on a short-range communication network (e.g., Bluetooth, wireless fidelity (WiFi), or infrared data association (IrDA)) or a long-range communication network (cellular network, 4G network, 5G network).

[0055] According to another embodiment, the connection (101) between the battery diagnostic device (110) and the electronic device (120) may be a connection via a device-to-device communication method (e.g., a bus, a general purpose input and output (GPIO), a serial peripheral interface (SPI), or a mobile industry processor interface (MIPI)).

[0056] According to one embodiment, the connection (130) between the battery diagnostic device (110) and the user terminal (130) may be a communication connection via a wired and / or wireless network.

[0057] According to one embodiment, the electronic device (120) may be a mobile device (e.g., a mobile phone, a laptop computer, a smart phone, a smart pad), an electric vehicle (e.g., an electric vehicle (EV), a hybrid EV (HEV), a plug-in HEV (PHEV), a fuel cell EV (FCEV)), an energy storage system (ESS), or a battery swapping system (BSS).

[0058] According to one embodiment, the electronic device (120) may include one or more battery units (121, 122, 123). Each of the one or more battery units (121, 122, 123) may be a battery cell, a battery module, a battery pack, or a battery rack.

[0059] According to one embodiment, the user terminal (130) may be a mobile device (e.g., a mobile phone, a laptop computer, a smart phone, a smart pad), or a personal computer (PC). According to one embodiment, the battery diagnostic device (110) may provide information related to the estimation result to the user terminal.

[0060] According to one embodiment, the battery diagnostic device (110) may include a communication circuit (111), a sensor (112), a memory (113), and a processor (114). Depending on the embodiment, the battery diagnostic device (110) illustrated in FIG. 1 may further include at least one component (e.g., a display, an input device, or an output device) other than the components illustrated in FIG. 1, or may omit at least one component (e.g., a sensor (112)) among the components illustrated in FIG. 1. For example, when the battery diagnostic device (110) is implemented as an external electronic device separate from the electronic device (120), such as a server or a cloud, the battery diagnostic device (110) may obtain status information of the battery (121, 122, or 123) using the communication circuit (111), and the battery diagnostic device (110) may not include the sensor (112).

[0061] According to one embodiment, the communication circuit (111) can establish a wired communication channel and / or a wireless communication channel between the battery diagnostic device (110) and the electronic device (120) and / or the user terminal (130), and transmit and receive data with the electronic device (120) and / or the user terminal (130) through the established communication channel.

[0062] According to one embodiment, the sensor (112) can measure information related to the status of the battery units (121, 122, 123) of the electronic device (120).

[0063] According to one embodiment, the communication circuit (111) and / or the sensor (112) may acquire time series data related to the state of the battery (121, 122, and / or 123). In one embodiment, the time series data related to the state of the battery (121, 122, and / or 123) may be data representing voltage, current, resistance, state of charge (SOC), state of health (SOH), and / or temperature of the battery (121, 122, and / or 123) over time. For example, the time series data may be univariate data for one of a plurality of variables (e.g., voltage, current, resistance, SOC, SOH, or temperature) related to the battery state. As another example, the time series data may be multivariate data for a plurality of variables related to the battery state.

[0064] According to one embodiment, the memory (113) may include volatile memory and / or non-volatile memory.

[0065] According to one embodiment, the memory (113) may store data used by at least one component (e.g., processor (114)) of the battery diagnosis device (110). For example, the data may include software (or instructions related thereto), input data, or output data. In one embodiment, the instructions, when executed by the processor (114), may cause the battery abnormality diagnosis device (101) to perform operations defined by the instructions.

[0066] According to one embodiment, the memory (113) may store a battery diagnosis system (e.g., the battery diagnosis system (200) of FIG. 2) that receives time series data related to the state of the battery (111, 112, or 113) and outputs an abnormal diagnosis result of the battery (111, 112, or 113).

[0067] According to one embodiment, the processor (114) may include a central processing unit, an application processor, a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor.

[0068] According to one embodiment, the processor (114) may execute software stored in the memory (113) to control at least one other component (e.g., hardware or software component) of the battery diagnostic device (110) connected to the processor (114) and perform various data processing or calculations. For example, the processor (114) may execute a battery diagnostic system (e.g., the battery diagnostic system (200) of FIG. 2) stored in the memory (113) to train the battery diagnostic system or perform abnormality diagnosis of the batteries (111, 113, and / or 115).

[0069] Hereinafter, the battery diagnosis system stored in the memory (113) will be specifically described with reference to FIGS. 2 and 3. The operations of each component included in the battery diagnosis system to be described below may be performed by various processing or calculations of the processor (114).

[0070] FIG. 2 is a diagram illustrating a battery diagnosis system according to an embodiment. FIG. 3 is a diagram illustrating various examples of how a battery diagnosis system according to an embodiment calculates a reconstruction error.

[0071] Referring to FIG. 2, the battery diagnostic system (200) may include an encoder (210), a latent space (220), a decoder (230), a first discriminator (240), a first generator (250), a second generator (260), a diagnostic device (270), and / or a second discriminator (280).

[0072] According to one embodiment, some components included in the battery diagnosis system (200) (e.g., encoder (210), latent space (220), decoder (230), first discriminator (240), first generator (250), second generator (260), and second discriminator (280)) may be implemented through a Time Series Anomaly Detection Using Generative Adversarial Networks (TadGAN) model.

[0073] According to one embodiment, the encoder (210) may convert time series data related to the state of a battery into latent variables. Here, the latent variables may refer to data stored in a latent space (220) that is potentially shared between the time series data and the latent variables (i.e., between different domains).

[0074] According to one embodiment, the encoder (210) may operate to separate various types of features included in time series data into a latent space (220). For example, if the time series data is multivariate data regarding a plurality of variables related to a battery state, the encoder (210) may separate each of the plurality of variables included in the time series data and generate a plurality of latent variables corresponding to each of the plurality of variables. According to one embodiment, the encoder (210) may be trained to group similar features together in the latent space (220) and to separate similar features when converting input data into latent variables.

[0075] According to one embodiment, the decoder (230) can convert data existing in the latent space (220) into reconstructed data in the time domain.

[0076] According to one embodiment, the decoder (230) can convert the latent variable converted by the encoder (210) into first reconstructed data in the time domain. Here, since the latent variable corresponds to one of a plurality of variables included in the time series data, the first reconstructed data can be time series reconstructed data for the one variable.

[0077] According to one embodiment, the decoder (230) may also convert random noise present in the latent space (220) into second reconstruction data in the time domain.

[0078] According to one embodiment, the decoder (230) may be trained to generate reconstructed data (e.g., first reconstructed data) similar to original data (e.g., time series data) based on the determination result output by the first discriminator (240) and / or the reconstruction error output by the first generator (250).

[0079] According to one embodiment, the first discriminator (240) can determine whether the data input to the first discriminator (240) is time series data input to the encoder (210) or whether the original data is first reconstructed data converted by the encoder (210) and the decoder (230). For example, the first discriminator (240) can use data that combines the time series data and the first reconstructed data as reference data. The first discriminator (240) can compare the input data with the reference data to determine whether the input data is time series data or first reconstructed data. For example, the time series data input to the first discriminator (240) can be univariate time series data corresponding to one variable included in the first reconstructed data among the time series data input to the encoder (210). In this case, the first discriminator (240) can perform the determination for each of the plurality of first reconstructed data corresponding to each of the plurality of variables.

[0080] According to one embodiment, the first discriminator (240) can perform the discrimination based on a discriminant model learned to discriminate between time series data and second reconstructed data converted from random noise by the decoder (230).

[0081] According to one embodiment, the first discriminator (240) can discriminate between time series data and first reconstructed data and output a discrimination result. For example, the first discriminator (240) can output the similarity between the time series data and the first reconstructed data as a discrimination result. According to one embodiment, the first discriminator (240) can discriminate between the time series data and the first reconstructed data based on the discrimination model and output a discrimination result.

[0082] In this way, the battery diagnosis system (200) can generate first reconstructed data similar to time series data by repeating the first reconstructed data generation process by the encoder (210) and decoder (230) and the determination process by the first discriminator (240), and the first discriminator (240) can more precisely determine the time series data and the first reconstructed data. In other words, adversarial learning can be performed within the model, and the quality of reconstructed data generation can be improved.

[0083] According to one embodiment, the first calculator (250) may compare time series data with first reconstructed data to calculate a reconstruction error. Here, the reconstruction error may represent a difference between the time series data and the first reconstructed data. For example, the time series data input to the first calculator (250) may be univariate time series data corresponding to one variable included in the first reconstructed data among the time series data input to the encoder (210). In this case, the first calculator (250) may calculate a reconstruction error for each of the plurality of first reconstructed data corresponding to each of the plurality of variables.

[0084] According to one embodiment, the first calculator (250) may calculate the reconstruction error by comparing the time series data and the first reconstructed data in the same time interval. For example, the first calculator (250) may calculate the reconstruction error according to a combined comparison method that compares the time series data and the first reconstructed data in the entire time interval included in the time series data. As another example, the first calculator (250) may calculate the reconstruction error according to a divided comparison method that compares the time series data and the first reconstructed data in a designated portion of the entire time interval included in the time series data. The combined comparison method and divided comparison method used by the first calculator (250) will be described in detail below with reference to FIG. 3.

[0085] In FIG. 3, original data (310), reconstructed data (320), some original data (330, 350) and some reconstructed data (340, 360) are illustrated.

[0086] According to one embodiment, the first calculator (250) can calculate the reconstruction error according to a combined comparison method that compares the original data (310) and the reconstructed data (320) over the entire time interval (T) of the original data (310) and the reconstructed data (320). According to the combined comparison method, the first calculator (250) can calculate the reconstruction error by calculating the difference between the original data (310) and the reconstructed data (320) over the entire time interval (T).

[0087] According to one embodiment, the first generator (250) can calculate the reconstruction error according to a split comparison method that compares the original data (330 or 350) and the reconstructed data (340 or 360) in some time interval (T1 or T2) of the original data (310) and the reconstructed data (320).

[0088] For example, the first calculator (250) can extract a first portion of original data (330) corresponding to a first time section (T1) among the entire time sections (T) from the original data (310), and extract a first portion of reconstructed data (340) corresponding to the first time section (T1) from the reconstructed data (320). The first calculator (250) can compare the first portion of original data (330) and the first portion of reconstructed data (340) to calculate a first reconstruction error. As another example, the first calculator (250) can extract a second portion of original data (350) corresponding to a second time section (T2) among the entire time sections (T) from the original data (310), and extract a second portion of reconstructed data (360) corresponding to the second time section (T2) from the reconstructed data (320). The first calculator (250) can compare the second partial original data (350) with the second partial reconstructed data (360) to calculate the second reconstruction error. According to one embodiment, the first calculator (250) can calculate the first reconstruction error or the second reconstruction error as the reconstruction error. According to another embodiment, the first calculator (250) can calculate the reconstruction error based on the first reconstruction error and the second reconstruction error. For example, the first calculator (250) can calculate the maximum value, minimum value, or average value of the first reconstruction error and the second reconstruction error as the reconstruction error.

[0089] Referring again to FIG. 2, the second generator (260) can calculate an anomaly score based on the judgment result output by the first discriminator (240) and the reconstruction error output by the first generator (250). According to one embodiment, the second generator (260) can calculate an anomaly score corresponding to each of a plurality of variables included in the time series data.

[0090] According to one embodiment, the second generator (260) can calculate the above-described ideal score based on the following mathematical expression 1 or mathematical expression 2.

[0091]

[0092]

[0093] (In Equations 1 and 2, α is the weight, Z reconstruction is the above reconstruction error, Z Cx is the result of the above determination.)

[0094] According to one embodiment, the diagnostic device (270) can diagnose an abnormality in the battery (121, 122, and / or 123) based on the abnormality score and threshold value output by the second generator (260). For example, the diagnostic device (270) can diagnose an abnormality in the battery (121, 122, and / or 123) by comparing the abnormality score and threshold value.

[0095] According to one embodiment, the diagnostic device (270) may increase a threshold value and compare the abnormality score with the threshold value to diagnose an abnormality in the battery (121, 122, and / or 123). For example, the diagnostic device (270) may repeat the process of increasing the threshold value by a certain value (e.g., 0.5) and comparing the increased threshold value with the abnormality score to perform an abnormality diagnosis.

[0096] According to one embodiment, the diagnostic device (270) can identify whether the threshold value is greater than or equal to a specified value. According to one embodiment, if the threshold value is greater than or equal to the specified value, the diagnostic device (270) can determine an optimal threshold value based on all threshold values ​​used in the abnormality diagnosis of the battery (121, 122, and / or 123). For example, the diagnostic device (270) can calculate AUROC (Area Under ROC curve) values ​​corresponding to each of the above threshold values, and determine a threshold value corresponding to the maximum AUROC value among the calculated AUROC values ​​as the optimal threshold value.

[0097] In one embodiment, the second discriminator (280) can determine the performance of the encoder (210). In one embodiment, the second discriminator (280) can determine the mapping performance of the encoder (210) based on the latent variables transformed by the encoder (210) and random noise on the latent space (220).

[0098] Fig. 4 is a flowchart illustrating the operation of a battery diagnostic device according to one embodiment. Fig. 4 may be an explanation of the operation of the battery diagnostic device (110) of Fig. 1, and may be explained using the configuration of Fig. 1.

[0099] The embodiment illustrated in FIG. 4 is merely one embodiment, and the order of steps according to various embodiments of the present invention may differ from that illustrated in FIG. 4, and some of the steps illustrated in FIG. 4 may be omitted, the order between steps may be changed, or steps may be merged. For example, operation 420 in FIG. 4 may be omitted.

[0100] Referring to FIG. 4, in operation 405, the battery diagnostic device (110) may convert time series data related to the state of the battery into latent variables. Here, the latent variables may refer to data stored in a latent space (220) that is potentially shared between the time series data and the latent variables (i.e., between different domains).

[0101] In operation 410, the battery diagnostic device (110) may convert the latent variable converted in operation 405 into first reconstructed data in the time domain. Here, since the latent variable corresponds to one of a plurality of variables included in the time series data, the first reconstructed data may be time series reconstructed data for the one variable.

[0102] In operation 415, the battery diagnosis device (110) may discriminate between the time series data and the first reconstructed data and output a discrimination result. For example, the battery diagnosis device (110) may output the similarity between the time series data and the first reconstructed data as the discrimination result. According to one embodiment, the battery diagnosis device (110) may discriminate between the time series data and the first reconstructed data based on the discrimination model and output the discrimination result. Here, the discrimination model may be specifically described with reference to FIG. 5, which will be described later. For example, the time series data used in operation 415 may be univariate time series data corresponding to one variable included in the first reconstructed data among the time series data used in operation 405. In this case, the battery diagnosis device (110) may perform the discrimination for each of the plurality of first reconstructed data corresponding to each of the plurality of variables.

[0103] In operation 420, the battery diagnostic device (110) can determine the performance of the encoder (210). According to one embodiment, the battery diagnostic device (110) can determine the mapping performance of the encoder (210) based on the latent variables transformed in operation 405 and random noise on the latent space (220).

[0104] In operation 425, the battery diagnosis device (110) may calculate a reconstruction error. According to one embodiment, the battery diagnosis device (110) may calculate a reconstruction error by comparing the time series data with the first reconstruction data converted in operation 410. Here, the reconstruction error may represent a difference between the time series data and the first reconstruction data. For example, the time series data used in operation 425 may be univariate time series data corresponding to one variable included in the first reconstruction data among the time series data used in operation 405. In this case, in this case, the battery diagnosis device (110) may calculate a reconstruction error for each of the plurality of first reconstruction data corresponding to each of the plurality of variables.

[0105] According to one embodiment, the battery diagnosis device (110) may calculate the reconstruction error by comparing the time series data and the first reconstruction data in the same time interval. For example, the battery diagnosis device (110) may calculate the reconstruction error according to a combined comparison method that compares the time series data and the first reconstruction data in the entire time interval included in the time series data. As another example, the battery diagnosis device (110) may calculate the reconstruction error according to a divided comparison method that compares the time series data and the first reconstruction data in a designated portion of the entire time interval included in the time series data.

[0106] In operation 430, the battery diagnosis device (110) may calculate an anomaly score based on the determination result calculated in operation 415 and the reconstruction error calculated in operation 425. According to one embodiment, the battery diagnosis device (110) may calculate an anomaly score corresponding to each of a plurality of variables included in the time series data.

[0107] According to one embodiment, the battery diagnostic device (110) can calculate the abnormality score based on the above mathematical expression 1 or mathematical expression 2.

[0108] In operation 435, the battery diagnostic device (110) can diagnose an abnormality in the battery (121, 122, and / or 123) based on the abnormality score and threshold value calculated in operation 430. For example, the battery diagnostic device (110) can diagnose an abnormality in the battery (121, 122, and / or 123) by comparing the abnormality score and threshold value.

[0109] Fig. 5 is a flowchart illustrating the operation of a battery diagnostic device according to an embodiment. Fig. 5 may be an explanation of the operation of the battery diagnostic device (110) of Fig. 1, and may be explained using the configuration of Fig. 1.

[0110] The embodiment illustrated in FIG. 5 is only one embodiment, and the order of steps according to various embodiments of the present invention may be different from that illustrated in FIG. 5, and some of the steps illustrated in FIG. 5 may be omitted, the order between steps may be changed, or steps may be merged.

[0111] Referring to FIG. 5, in operation 505, the battery diagnostic device (110) can convert random noise existing in the latent space (220) into second reconstructed data in the time domain.

[0112] In operation 510, the battery diagnostic device (110) can train a discrimination model to discriminate between time series data and second reconstructed data converted in operation 510.

[0113] Fig. 6 is a flowchart of the operation of a battery diagnosis device according to one embodiment. Fig. 6 may be a description of the operation of the battery diagnosis device (110) of Fig. 1, and may be described using the configuration of Fig. 1.

[0114] The embodiment illustrated in FIG. 6 is only one embodiment, and the order of steps according to various embodiments of the present invention may be different from that illustrated in FIG. 6, and some of the steps illustrated in FIG. 6 may be omitted, the order between steps may be changed, or steps may be merged.

[0115] Referring to FIG. 6, in operation 605, the battery diagnostic device (110) may increase the threshold value. For example, the battery diagnostic device (110) may increase the threshold value by a certain value (e.g., 0.5).

[0116] In operation 610, the battery diagnostic device (110) can diagnose an abnormality in the battery (121, 122, and / or 123) by comparing the abnormality score with a threshold value.

[0117] At operation 615, the battery diagnostic device (110) can identify whether the increased threshold value at operation 605 is greater than or equal to a specified value.

[0118] If the threshold value is identified as being less than the specified value in operation 615 ('NO'), the battery diagnostic device (110) may perform operation 605 again.

[0119] If the threshold value is identified as being greater than or equal to the specified value in operation 615 ('YES'), in operation 620, the battery diagnostic device (110) may determine an optimal threshold value based on all threshold values ​​used in the abnormality diagnosis of the battery (121, 122, and / or 123). For example, the battery diagnostic device (110) may calculate AUROC values ​​corresponding to each of the above threshold values, and determine a threshold value corresponding to the maximum AUROC value among the calculated AUROC values ​​as the optimal threshold value.

[0120] The terms "include," "comprise," or "have" used herein, unless otherwise specifically stated, imply that the corresponding component may be included, and therefore should be interpreted to include other components rather than to exclude other components. All terms, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art to which the embodiments disclosed in this document pertain, unless otherwise defined. Commonly used terms, such as terms defined in dictionaries, should be interpreted to be consistent with their meaning in the context of the relevant technology, and shall not be interpreted in an idealized or overly formal sense, unless explicitly defined in this document.

Claims

1. An encoder that converts time series data related to the battery status into latent variables; A decoder that converts the latent variable into first reconstructed data in the time domain; A first discriminator that discriminates the above time series data and the first reconstruction data and outputs a discrimination result; A first calculator that compares the time series data with the first reconstruction data to calculate a reconstruction error; A second calculator that calculates an anomaly score based on the above judgment result and the above reconstruction error; and A battery diagnostic device comprising a diagnostic device that diagnoses an abnormality of the battery based on the above-mentioned abnormality score and threshold value.

2. In claim 1, A battery diagnostic device further comprising a second discriminator that determines the performance of the encoder based on the latent variable and random noise in the latent space.

3. In claim 1, The above decoder converts random noise in the latent space into second reconstructed data in the time domain, A battery diagnostic device, wherein the first discriminator outputs the discrimination result based on a discrimination model learned to discriminate between the time series data and the second reconstructed data.

4. In claim 1, A battery diagnostic device, wherein the first generator calculates the reconstruction error according to a combined comparison method that compares the time series data and the first reconstruction data over the entire time period included in the time series data.

5. In claim 1, A battery diagnostic device, wherein the first generator calculates the reconstruction error according to a segmented comparison method that compares the time series data with the first reconstruction data in a designated portion of the entire time interval included in the time series data.

6. In claim 1, The second generator is a battery diagnostic device that calculates the abnormal score based on the following mathematical expression 1. [Mathematical Formula 1] (In Equation 1, α is the weight, Z reconstruction is the above reconstruction error, Z Cx is the result of the above determination.) 7. In claim 1, The second generator is a battery diagnostic device that calculates the abnormal score based on the following mathematical expression 2. [Equation 2] (In Equation 2, α is the weight, Z reconstruction is the above reconstruction error, Z Cx is the result of the above determination.) 8. In claim 1, The above diagnostic device, Increase the above threshold value and compare the above abnormal score and the above threshold value to diagnose an abnormality in the battery, A battery diagnostic device that determines an optimal threshold value based on all threshold values ​​used in diagnosing an abnormality of the battery when the threshold value is greater than or equal to a specified value.

9. In claim 8, The above diagnostic device, Calculate the AUROC (Area Under ROC curve) values ​​corresponding to each of the above threshold values, A battery diagnostic device that determines a threshold value corresponding to the maximum AUROC value among the above AUROC values ​​as the optimal threshold value.

10. The action of converting time series data related to the battery status into latent variables; An operation of converting the above latent variable into first reconstructed data in the time domain; An operation of determining the above time series data and the first reconstruction data and producing a determination result; An operation of calculating a reconstruction error by comparing the above time series data with the first reconstruction data; An operation of calculating an anomaly score based on the above judgment result and the above reconstruction error; and A battery diagnosis method, comprising an operation of diagnosing an abnormality of the battery based on the above abnormality score and threshold value.

11. In claim 10, Further comprising an operation of converting random noise in latent space into second reconstructed data in the time domain, A battery diagnosis method, wherein the operation of producing the above judgment result includes an operation of producing the judgment result based on a judgment model learned to judge the time series data and the second reconstruction data.

12. In claim 10, A battery diagnosis method, wherein the operation of calculating the reconstruction error includes an operation of calculating the reconstruction error according to a combined comparison method of comparing the time series data and the first reconstruction data over the entire time period included in the time series data.

13. In claim 10, A battery diagnosis method, wherein the operation of calculating the reconstruction error includes an operation of calculating the reconstruction error according to a segmented comparison method that compares the time series data and the first reconstruction data in a designated portion of the entire time interval included in the time series data.

14. In claim 10, An operation of increasing the threshold value and comparing the abnormal score and the threshold value to diagnose an abnormality in the battery, and A battery diagnosis method further comprising an operation of determining an optimal threshold value based on all threshold values ​​used in the abnormal diagnosis of the battery, if the threshold value is greater than or equal to a specified value.

15. In claim 14, The operation of determining the above optimal threshold value is: An operation of calculating AUROC (Area Under ROC curve) values ​​corresponding to each of the above threshold values, and A battery diagnosis method, comprising an operation of determining a threshold value corresponding to the maximum AUROC value among the above AUROC values ​​as the optimal threshold value.

Images