Device for determining component state according to component type and its operating method

Abstract

A component condition judgment device according to one embodiment disclosed in this document may include a data acquisition unit that acquires performance data for one or more components included in an electronic device; a data extraction unit that extracts first performance data for a first component whose replacement timing has not been predetermined from the performance data; and a judgment unit that judges the condition of the first component using a pre-trained first artificial intelligence model based on the first performance data, wherein the first artificial intelligence model may be trained based on a plurality of reference data sets according to the degree of deterioration of the first component, and may be trained to reduce the difference between the actual performance of the first component and the predicted performance output from the first artificial intelligence model.

Description

[Technical Field]

[0001] [CROSS-REFERENCE TO RELATED APPLICATIONS] The present invention claims the benefit of priority based on Korean Patent Application No. 10-2023-0029521, filed on March 6, 2023, and all contents disclosed in the documents of said Korean patent application are incorporated herein by reference.

[0002] SUMMARY OF THE INVENTION The embodiments disclosed herein relate to a device and method for determining a part state by part type. [Background technology]

[0003] Recently, research and development into secondary batteries has been actively conducted. Here, secondary batteries are batteries that can be charged and discharged, and include conventional Ni / Cd batteries, Ni / MH batteries, and the latest lithium-ion batteries. Among secondary batteries, lithium-ion batteries have the advantage of having a much higher energy density than conventional Ni / Cd batteries, Ni / MH batteries, etc. Furthermore, lithium-ion batteries can be manufactured in a compact and lightweight form, and are used as power sources for mobile devices. Recently, their use has expanded to include power sources for electric vehicles, and they are attracting attention as a next-generation energy storage medium.

[0004] There are multiple facilities in the process of producing such secondary batteries. As the secondary battery production process progresses, the components that make up each facility may deteriorate. The deterioration of a component may lead to component failure, which may make it impossible to use the facility that contains that component. Therefore, there is a need for technology that can determine the condition of multiple components and predict the replacement time in advance. Summary of the Invention [Problem to be solved by the invention]

[0005] There is a predictive maintenance technology that diagnoses whether multiple pieces of equipment are likely to fail in advance and performs maintenance. Here, predictive maintenance technology can grasp the condition of the equipment and predict its future condition. Furthermore, by utilizing AI-based predictive maintenance technology that combines artificial intelligence (AI) with predictive maintenance, it has become possible to predict the condition of parts even more accurately.

[0006] However, conventional AI-based predictive maintenance technology has the problem that it cannot accurately predict the condition of parts whose failure criteria are unknown, because it trains artificial intelligence models using training datasets that do not take into account the type of part.

[0007] In addition, conventional AI-based predictive maintenance technology trains artificial intelligence models using training data sets that do not take into account component deterioration, which means that it has the problem of being unable to accurately predict the condition of components whose replacement cycles are affected by deterioration.

[0008] 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 following description. [Means for solving the problem]

[0009] A component condition judgment device according to one embodiment disclosed in this document may include a data acquisition unit that acquires performance data for one or more components included in an electronic device; a data extraction unit that extracts first performance data for a first component whose replacement timing has not been predetermined from the performance data; and a judgment unit that judges the condition of the first component using a pre-trained first artificial intelligence model based on the first performance data, wherein the first artificial intelligence model may be trained based on a plurality of reference data sets according to the degree of deterioration of the first component, and may be trained to reduce the difference between the actual performance of the first component and the predicted performance output from the first artificial intelligence model.

[0010] In one embodiment, the first artificial intelligence model may be trained using the plurality of reference data sets sequentially based on a rolling window method.

[0011] In one embodiment, the degree of deterioration may be set based on at least one of the time during which the first component is operated and the operating environment.

[0012] In one embodiment, the data extraction unit may further extract second performance data for a second part whose replacement cycle is predetermined from the performance data, and the judgment unit may judge the condition of the second part using a second artificial intelligence model based on the second performance data.

[0013] In one embodiment, the second artificial intelligence model may be trained based on performance evaluation using k-fold cross validation so that the average performance of the solution is equal to or greater than the required performance.

[0014] An operating method of a component condition judgment device according to one embodiment disclosed in this document may include an operation of acquiring performance data for one or more components included in an electronic device; an operation of extracting first performance data for a first component for which a replacement time has not been predetermined from the performance data; and an operation of judging the condition of the first component using a pre-trained first artificial intelligence model based on the first performance data, wherein the first artificial intelligence model may be trained based on a plurality of reference data sets according to the degree of deterioration of the first component, and may be trained to reduce the difference between the actual performance of the first component and the predicted performance output from the first artificial intelligence model.

[0015] In one embodiment, the first artificial intelligence model may be trained using the plurality of reference data sets sequentially based on a rolling window method.

[0016] In one embodiment, the degree of deterioration may be set based on at least one of the time during which the first component is operated and the operating environment.

[0017] In one embodiment, the method may further include an operation of extracting second performance data from the performance data for a second component having a predetermined replacement cycle; and an operation of determining the condition of the second component using a second artificial intelligence model based on the second performance data.

[0018] In one embodiment, the second artificial intelligence model may be trained based on performance evaluation using k-fold cross validation so that the average performance of the solution is equal to or greater than the required performance. [Effects of the Invention]

[0019] The component condition determination device and its operating method according to various embodiments disclosed in this document classify component types based on whether or not the component has been previously determined to be replaced, and by using an artificial intelligence model optimized for each type, the component condition can be accurately predicted even for components that have not been previously scheduled for replacement.

[0020] The component condition determination device and its operating method according to various embodiments disclosed herein can accurately predict the condition of a component due to deterioration by training an artificial intelligence model using multiple reference data sets that reflect deterioration due to the component's operating time and / or operating environment.

[0021] The effects of the artificial intelligence model learning device based on the degree of deterioration of a component and its operating method disclosed in this document are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the disclosure of this document. [Brief explanation of the drawings]

[0022] [Figure 1] FIG. 1 is a block diagram of an artificial intelligence model learning device according to an embodiment of the present disclosure. [Figure 2] 1 illustrates an example graph of the results of training a first artificial intelligence model using multiple reference data sets that reflect degradation over time, according to one embodiment of the present disclosure. [Figure 3] 10 illustrates an example graph of the results of training a first artificial intelligence model using multiple reference data sets that reflect degradation due to the operating environment according to an embodiment of the present disclosure. [Figure 4] 10 illustrates a graph showing the results of cross-validating a second artificial intelligence model using a k-fold cross-validation method according to one embodiment of the present disclosure. [Figure 5] 4 is a flowchart illustrating a method of operation of a component condition determination device according to one embodiment of the present disclosure.

[0023] With regard to the description of the drawings, the same or similar reference numerals may be used for the same or similar components. DETAILED DESCRIPTION OF THE INVENTION

[0024] Although the present invention has been described below with reference to the accompanying drawings, it should be understood that the present invention is not limited to the specific embodiments and includes various modifications, equivalents, and / or alternatives of the embodiments of the present invention.

[0025] The embodiments and terms used in this document are not intended to limit the technical features described in this document to a specific embodiment, but should be understood to include various modifications, equivalents, or alternatives of the embodiment. In connection with the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of a noun corresponding to an item may include one or more of the said item unless the relevant context clearly dictates otherwise.

[0026] In this document, each of the phrases such as "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" may include any one of the items listed therein or all possible combinations thereof. Terms such as "first," "second," "first," "second," "A," "B," "(a)," or "(b)" may be used merely to distinguish one element from other elements and do not limit the element in other respects (e.g., importance or order) unless specifically stated to the contrary.

[0027] In this document, when a (e.g., first) component is referred to as being "coupled," "coupled," or "connected" to another (e.g., second) component, with or without the terms "functionally" or "communicatively," this means that the component may be directly (e.g., wired or wirelessly) or indirectly (e.g., through a third component) coupled to the other component.

[0028] Methods according to various embodiments disclosed herein may be provided in a computer program product. The computer program product may be traded as a commodity between a seller and a buyer. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., compact disc read-only memory, CD-ROM) or distributed online (e.g., downloaded or uploaded) via an application store or directly between two user devices. In the case of online distribution, at least a portion of the computer program product may be at least temporarily stored or temporarily generated in a machine-readable storage medium, such as the memory of a manufacturer's server, an application store server, or an intermediary server.

[0029] According to the embodiments disclosed herein, each of the aforementioned components (e.g., modules or programs) may include one or more entities, and some of the entities may be located separately in other components. According to the embodiments disclosed herein, one or more of the aforementioned components or operations may be omitted, or one or more other components or operations may be added. Alternatively or additionally, multiple components (e.g., modules or programs) may be integrated into a single component. In this case, the integrated component may perform one or more functions of each of the multiple components in the same or similar manner as those performed by the respective components of the multiple components before the integration. According to the embodiments disclosed herein, the operations performed by a module, program, or other component may be performed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be performed in a different order, omitted, or one or more other operations may be added.

[0030] FIG. 1 is a block diagram of a component condition determination device 10 according to one embodiment of the present disclosure.

[0031] Referring to FIG. 1, a component condition determination device 10 may be connected to an electronic device 12 and a user terminal 14 via wire and / or wireless.

[0032] In one embodiment, the connection 11 between the component condition determination device 10 and the electronic device 12 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, WiFi (wireless fidelity), or IrDA (infrared data association)) or a long-range communication network (e.g., a cellular network, a 4G network, or a 5G network).

[0033] In another embodiment, the connection 11 between the component condition determination device 10 and the electronic device 12 may be a connection via a communication method between devices (e.g., a bus, a general purpose input and output (GPIO), a serial peripheral interface (SPI), or a mobile industry processor interface (MIPI)).

[0034] In one embodiment, the connection 13 between the component condition determination device 10 and the user terminal 14 may be a communication connection via a wired and / or wireless network.

[0035] In one embodiment, the electronic device 12 may be an electric vehicle (e.g., an electric vehicle (EV), a hybrid EV (HEV), a plug-in HEV (PHEV), or a fuel cell EV (FCEV)), a battery production facility, an energy storage system (ESS), or a battery swapping system (BSS).

[0036] In the following description, it is assumed that the electronic device 12 is an electric vehicle and that the components 121, 123, and 125 are components that constitute the electric vehicle. However, this is merely a premise for the purpose of explanation, and the electronic device 12 of the present invention is not limited to electric vehicles and may be applied to various other devices.

[0037] In one embodiment, the user terminal 14 may be a mobile device (eg, a mobile phone, a laptop computer, a smart phone, a smart pad) or a personal computer (PC).

[0038] In one embodiment, the component condition determination device 10 may include a communication circuit 100, a sensor 120, a memory 140, and a processor 160. Depending on the embodiment, the component condition determination device 10 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.

[0039] In one embodiment, the communication circuit 100 can establish a wired communication channel and / or a wireless communication channel between the component condition determination device 10 and the electronic device 12 and / or the user terminal 14, and can transmit and receive data to and from the electronic device 12 and / or the user terminal 14 via the established communication channel.

[0040] In one embodiment, the sensor 120 can sense the status of one or more components 121, 123, 125 that make up the electronic device 12. In one embodiment, the status data can indicate one or more data related to the type of component, the degree of deterioration, the temperature, or a combination thereof. Here, the component types can be consumable and non-consumable. A consumable component can be a component whose failure criteria are clear and whose replacement interval can be scheduled in advance. For example, a consumable component can be a vehicle filter. A non-consumable component can be a component whose failure criteria are unclear and whose replacement interval cannot be scheduled in advance. For example, a non-consumable component can be a vehicle motor or cylinder.

[0041] Hereinafter, data relating to the performance of the parts may be referred to as performance data, the consumable parts may be referred to as first parts, and the non-consumable parts may be referred to as second parts.

[0042] In one embodiment, memory 140 may include volatile memory and / or non-volatile memory.

[0043] In one embodiment, the memory 140 may store data used by at least one component (e.g., the processor 160) of the component condition determination apparatus 10. For example, the data may include software (or instructions therefor), input data, or output data. In one embodiment, the instructions, when executed by the processor 160, may cause the component condition determination apparatus 10 to perform the operation defined by the instructions.

[0044] In one embodiment, memory 140 may include one or more pieces of software (eg, data acquirer 142, data extractor 144, decision maker 146, model trainer 148).

[0045] In one embodiment, processor 160 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.

[0046] In one embodiment, the processor 160 can execute software (e.g., the data acquisition unit 142, the data extraction unit 144, the judgment unit 146, the model learning unit 148) to control at least one other component (e.g., a hardware or software component) of the component condition judgment device 10 coupled to the processor 160, and can perform various data processing or calculations.

[0047] The following describes a method for determining the condition of a part using the data acquisition unit 142, the data extraction unit 144, the determination unit 146, and the model learning unit 148.

[0048] The data acquisition unit 142 can acquire performance data. The data acquisition unit 142 can acquire the performance data via the communication circuitry 100. The data acquisition unit 142 can acquire performance data regarding one or more components included in the electronic device 12. The data acquisition unit 142 can acquire performance data regarding one or more components included in the electronic device 12 via the communication circuitry 100.

[0049] The data extraction unit 144 can extract specific data from the performance data. To extract the specific data, the data extraction unit 144 can classify the performance data according to preset items. To extract data for the specific items, the data extraction unit 144 can classify the performance data into first performance data or second performance data according to the type of part. The data extraction unit 144 can extract the first performance data from the performance data. The data extraction unit 144 can extract the second performance data from the performance data. Here, the first performance data may be performance data for a first part, and the second performance data may be performance data for a second part. However, the items classified by the data extraction unit 144 are not limited to the first performance data or the second performance data, and the performance data can be classified based on other classification items.

[0050] The determination unit 146 may determine the state of the component. The determination unit 146 may determine the state of the component using a pre-trained artificial intelligence model. The determination unit 146 may determine the state of the component using at least one of a pre-trained first artificial intelligence model and a pre-trained second artificial intelligence model. The determination unit 146 may determine the state of the first component using the first artificial intelligence model, and may determine the state of the second component using the second artificial intelligence model.

[0051] The model learning unit 148 can learn an artificial intelligence model. The model learning unit 148 can learn at least one of a first artificial intelligence model and a second artificial intelligence model.

[0052] First, artificial intelligence model learning method The model training unit 148 may generate a training dataset. The model training unit 148 may generate a plurality of reference datasets for training the first artificial intelligence model. The model training unit 148 may generate a plurality of reference datasets classified according to the degree of deterioration of the first component. Here, the degree of deterioration may be determined based on at least one of the time the first component is operated and the operating environment in which the first component is operated. The operating environment may be the environment in which equipment including the component is operated. The degree of deterioration of components included in the equipment may differ depending on the type of battery to be produced. For example, the degree of deterioration when producing the same number of button-type batteries during the same period may be different from the degree of deterioration when producing cell-type batteries.

[0053] The model training unit 148 can train a first artificial intelligence model. The model training unit 148 can train the first artificial intelligence model using a plurality of reference data sets. The model training unit 148 can train the first artificial intelligence model based on a rolling window method. The model training unit 148 can train the first artificial intelligence model by sequentially using a plurality of reference data sets based on the rolling window method. Here, the rolling window method may refer to a technique used to train an artificial intelligence model according to the order in which a plurality of reference data sets are acquired for a predetermined period of time.

[0054] When the model learning unit 148 learns the first artificial intelligence model based on the rolling window method, the first artificial intelligence model may be learned sequentially according to the order in which the data sets were acquired, thereby fine-tuning the parameters of the first artificial intelligence model.

[0055] In one embodiment, the model training unit 148 may train the first artificial intelligence model by preferentially using a reference data set with a low degree of degradation. For example, the model training unit 148 may train the first artificial intelligence model by using a reference data set with a low degree of degradation first and a reference data set with a high degree of degradation last for a predetermined period of time.

[0056] The model learning unit 148 may train the first artificial intelligence model so that the difference between pre-stored actual performance and predicted performance decreases. The model learning unit 148 may train the first artificial intelligence model so that the difference between actual performance and predicted performance decreases for each preset period based on a rolling window method. Here, the actual performance may be the actual performance of the first part depending on the degree of deterioration. The predicted performance may be data predicting the performance of the first part output through the first artificial intelligence model. For example, if the first part is a motor that has deteriorated 50% and the predicted performance of the motor predicted by the first artificial intelligence model is 30%, the model learning unit 148 may train the first artificial intelligence model so that the predicted performance converges to 50% through learning for the remaining period of the preset period.

[0057] In one embodiment, the model training unit 148 can train the first artificial intelligence model to output a probability value representing a fault or normality based on a reference data set. For example, if a fault is 0 and a normality is 1, the first artificial intelligence model may be trained to output 0.5 for a motor that is 50% deteriorated.

[0058] Second AI model learning method The model learning unit 148 can train the second artificial intelligence model. The model learning unit 148 can train the second artificial intelligence model using a training dataset. Here, the training dataset may be generated based on performance data of a part for which failure criteria or replacement timing criteria are clear.

[0059] The model training unit 148 may validate the second AI model based on a k-fold cross validation method. The model training unit 148 may validate the second AI model using a validation dataset based on the k-fold cross validation method. The model training unit 148 may use the k-fold cross validation to verify whether the average solution performance of the second AI model is equal to or greater than a pre-stored required performance. Here, the k-fold cross validation method may be a technique in which a validation dataset is divided into k folds (k: a natural number greater than or equal to 2), (k-1) training datasets and one validation dataset are selected from the folds, and k cross validations are performed on the AI ​​model. The average solution performance may be the average performance of the k cross validation results obtained by the k-fold cross validation method.

[0060] The model learning unit 148 can train the second AI model so that the solution average performance of the second AI model is equal to or greater than the required performance. When the result of k-fold cross-validation shows that the solution average performance is lower than the pre-stored required performance, the model learning unit 148 can construct a second AI model based on the validation results. When the result of k-fold cross-validation shows that the solution average performance is lower than the pre-stored required performance, the model learning unit 148 can construct a second AI model whose solution average performance satisfies the pre-stored required performance. Here, the required performance may be a target value for the accuracy with which the second AI model predicts the presence or absence of a fault in the second component. The solution average performance may be an average value of k validation results obtained by k-fold cross-validation.

[0061] FIG. 2 illustrates a graph 200 of results from training a first artificial intelligence model using multiple reference data sets that reflect degradation over time, according to one embodiment of the present disclosure.

[0062] 2, the model learning unit 148 may compare predicted performance 202 of the first part predicted by the first AI model with actual performance 204 based on the result graph 200. Here, the predicted performance 202 may be the performance of the first part predicted using the first AI model based on the degree of deterioration of the first part over time. The actual performance 204 may be the actual performance of the first part in a deteriorated state.

[0063] The model learning unit 148 can learn the first artificial intelligence model so that the difference between the predicted performance 202 and the actual performance 204 decreases as a predetermined period of time elapses based on the rolling window method.

[0064] Referring to the results graph 200, it can be seen that the difference between predicted performance 202 and actual performance 204 decreases as the training period progresses.

[0065] FIG. 3 illustrates an example result graph 300 of training a first artificial intelligence model using multiple reference data sets reflecting degradation due to the driving environment, according to one embodiment of the present disclosure.

[0066] 3, predicted performance 302 for a first part predicted by a first artificial intelligence model can be compared with actual performance 304 based on a result graph 300. Here, predicted performance 302 may be the performance of the first part predicted using the first artificial intelligence model due to degradation of the first part's operating environment, and actual performance 304 may be the actual performance of the first part in a degraded state.

[0067] 3 and 4, the results of the model learning unit 148 training the first artificial intelligence model using a plurality of reference data sets reflecting temporal degradation or working environment degradation have been described, but the model learning unit 148 can train the first artificial intelligence model using a plurality of reference data sets reflecting both temporal degradation and working environment degradation. For example, the plurality of reference data sets may be generated by dividing the degradation level into 10 levels and the working environment level into 10 levels, and by arbitrarily combining the degradation level and the working environment level.

[0068] FIG. 4 illustrates a result graph 400 of cross-validating a second artificial intelligence model using k-fold cross-validation according to one embodiment of the present disclosure.

[0069] 4, it can be seen that the solution average performance 402 is less than the required performance 404. Here, the solution average performance 402 may be the average of the performance values ​​when k is from 1 to N (N: a natural number equal to or greater than 2). The required performance 404 may be the prediction accuracy required for the second artificial intelligence model that predicts the replacement time of the second part.

[0070] If the solution average performance 402 is less than the required performance 404, the model learning unit 148 can generate a new learning dataset based on the verification results.

[0071] The model training unit 148 can train the second artificial intelligence model using a new training dataset. If the solution average performance 402 is less than the required performance 404, the model training unit 148 can generate a new training dataset and train the second artificial intelligence model until the solution average performance 402 satisfies the required performance 404.

[0072] FIG. 5 is a flowchart illustrating a method of operation of a component condition determination device according to one embodiment of the present disclosure.

[0073] 5 , in operation 500, the data acquirer 142 may acquire performance data. The data acquirer 142 may acquire the performance data via the communication circuitry 100. The data acquirer 142 may acquire performance data regarding one or more components included in the electronic device 12. The data acquirer 142 may acquire the performance data for one or more components included in the electronic device 12 via the communication circuitry 100.

[0074] In operation 502, the data extraction unit 144 can extract specific data from the performance data. In order to extract the specific data, the data extraction unit 144 can classify the performance data by preset items. In order to extract data for the specific items, the data extraction unit 144 can classify the performance data into first performance data or second performance data depending on the type of part. The data extraction unit 144 can extract first performance data from the performance data. The data extraction unit 144 can extract second performance data from the performance data.

[0075] In operation 504, the determination unit 146 can determine the condition of the part. The determination unit 146 can determine the condition of the part using a pre-trained artificial intelligence model. When the data extracted by the data extraction unit 144 is first performance data, the determination unit 146 can determine the condition of the first part using a first artificial intelligence model.

[0076] In operation 506, the determining unit 146 may determine the state of the second part using a second artificial intelligence model if the data extracted by the data extracting unit 144 is second performance data.

[0077] In operation 508 , the determination result of at least one of the first component status and the second component status determined by the determination unit 146 may be transmitted to the user terminal 14 via the communication circuitry 100 .

[0078] As used above, terms such as "comprise," "constitute," or "have," unless otherwise specified, mean that the relevant element can be contained within the term, and should be interpreted as meaning that other elements can be further included, rather than excluding other elements. All terms, including technical and scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art to which the embodiments disclosed herein belong, unless otherwise defined. Commonly used terms, such as predefined terms, should be interpreted in a manner consistent with the context of the relevant art, and should not be interpreted in an ideal or overly formal sense unless expressly defined herein.

[0079] The above description is merely an illustrative example of the technical concept of the present invention, and various modifications and variations may be made by a person skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed herein are intended to illustrate, not limit, the technical concept of the present invention, and the scope of the technical concept of the present invention is not limited by these embodiments. The scope of protection of the technical concept disclosed herein should be interpreted by the following claims, and all technical concepts within the scope equivalent thereto should be interpreted as being included in the scope of the present invention. [Explanation of symbols]

[0080] 10. Parts condition determination device 11 Consolidated 12 Electronic equipment 13 Consolidated 14 User terminal 100 Communication Circuit 120 sensors 121 parts 123 parts 125 parts 140 memory 142 Data Acquisition Unit 144 Data Extraction Unit 146 Judgment Department 148 Model Learning Department 160 processors

Claims

1. a data acquisition unit that acquires performance data for one or more components included in the electronic device; a data extraction unit that extracts first performance data for a first part whose replacement time has not been predetermined from the performance data; and a determination unit that determines a state of the first component using a pre-trained first artificial intelligence model based on the first performance data; The first artificial intelligence model is learning based on a plurality of reference data sets according to the deterioration degree of the first part; The first artificial intelligence model is trained to reduce a difference between an actual performance of the first part and a predicted performance output from the first artificial intelligence model. Parts condition determination device.

2. The first artificial intelligence model is The plurality of reference data sets are sequentially trained based on a rolling window method. The component condition determination device according to claim 1 .

3. The degree of deterioration is The time is set based on at least one of the time when the first component is driven and the driving environment.

3. The component state determination device according to claim 1 or 2.

4. the data extraction unit further extracts second performance data for a second part whose replacement cycle is predetermined from the performance data; the determination unit determines the state of the second part using a second artificial intelligence model based on the second performance data.

3. The component state determination device according to claim 1 or 2.

5. The second artificial intelligence model is trained based on performance evaluation using a k-fold cross validation method so that the average performance of the solution is equal to or greater than the required performance. The component state determination device according to claim 4.

6. obtaining performance data for one or more components included in the electronic device; extracting first performance data for a first part whose replacement time has not been predetermined from the performance data; and determining a state of the first component using a pre-trained first artificial intelligence model based on the first performance data; The first artificial intelligence model is learning based on a plurality of reference data sets according to the deterioration degree of the first part; The first artificial intelligence model is trained to reduce a difference between an actual performance of the first part and a predicted performance output from the first artificial intelligence model. A method of operation of the component condition determination device.

7. The first artificial intelligence model is The plurality of reference data sets are sequentially trained based on a rolling window method.

7. The method of claim 6.

8. The degree of deterioration is The time is set based on at least one of the time when the first component is driven and the driving environment.

8. A method according to claim 6 or 7.

9. extracting second performance data for a second part having a predetermined replacement period from the performance data; and and determining a condition of the second component using a second artificial intelligence model based on the second performance data.

8. A method according to claim 6 or 7.

10. The second artificial intelligence model is trained based on performance evaluation using a k-fold cross validation method so that the average performance of the solution is equal to or greater than the required performance.

10. The method of claim 9.

Images