Simulation system and data distribution processing method

Abstract

An analog system for performing an experiment on a battery pack including a plurality of battery cells by using at least one experimental device according to the disclosure includes an IoT device that receives input and output of the at least one experimental device and pre-processes the input and output to generate input data and output data, an edge computing device for predicting information for predicting a capacity degradation value of the plurality of battery cells by using an equivalent circuit model module and a thermal model module based on received data among the input data and the output data, and a server that receives the predicted information from the edge computing device, predicts the capacity degradation value of the plurality of battery cells by using an aging model module, and predicts a defect rate based on the predicted capacity degradation value. A data size of the input data and the output data transmitted from the IoT device to the edge computing device can be smaller than a predetermined data size.

Description

[0001] Cross-reference to related applications

[0002] This application claims priority and benefit to Korean Patent Application No. 10-2020-0002216, filed with the Korean Intellectual Property Office on January 7, 2020, the entire contents of which are incorporated herein by reference. Technical Field

[0003] This disclosure relates to a simulation system and a data distribution method. Background Technology

[0004] For battery cycle life analysis, methods are used to confirm the degree of degradation through actual charge and discharge experiments of individual battery cells or to predict the degree of degradation through software simulation based on MBD (model-based design).

[0005] Specifically, degradation prediction using software simulation methods can be performed by pre-calculating scenarios and considering various layers of current, ECM (Equivalent Circuit Model), heat, and aging configuration through fixed virtual input values. Since actual battery degradation requires long-term experimental or measured data, software simulation methods are preferred, and the predicted battery degradation is used to determine the battery warranty period. However, currently, hybrid vehicles, electric vehicles, etc., are still in the development stage, and data regarding input values ​​is insufficient. Therefore, driving data from gasoline vehicles is used to convert power / temperature / driving mode, etc., into input values ​​for use. Summary of the Invention

[0006] [Technical Issues]

[0007] This disclosure aims to provide a simulation system and data distribution method that can improve the accuracy and efficiency of battery cycle life prediction.

[0008] [Technical Solution]

[0009] The simulation system of the present invention for experimenting on a battery pack comprising multiple battery cells using at least one experimental device includes: an IoT device that receives inputs and outputs from at least one experimental device and preprocesses the inputs and outputs to generate input data and output data; an edge computing device that predicts information for predicting capacity degradation values ​​of multiple battery cells based on the received data in the input data and output data by using an equivalent circuit model module and a thermal model module; and a server that receives the predicted information from the edge computing device, predicts the capacity degradation values ​​of multiple battery cells by using an aging model module, and predicts a defect rate based on the predicted capacity degradation values.

[0010] The equivalent circuit model module can receive at least one of a battery pack voltage, an initial SOC, a current flowing to the battery pack, a temperature of the battery cells, a number of battery cells constituting the battery pack, and a resistance of the battery pack from the IoT device, and predict at least one of an SOC of each of the battery cells, the battery pack voltage, the current flowing to the battery pack, a heat value of the battery cells, and a current rate (C-Rate).

[0011] The thermal model module can receive the heat value from the equivalent circuit model module and receive a cooling performance input to predict the temperature of the battery cells.

[0012] The aging model module can receive at least one of data on the SOC of each of the battery cells, the battery pack voltage, the current flowing into the battery pack, the current rate (C-Rate), and the temperature of the battery cells to predict a capacity retention rate and an internal resistance of each of the plurality of battery cells.

[0013] The server can further include a capacity degradation value calculation unit that receives the capacity retention rate and the internal resistance of each of the battery cells to predict a capacity degradation value indicating a degree of capacity degradation of each battery cell, and a defect rate prediction unit that stores information on the capacity degradation value of each of the plurality of battery cells for each time unit and predicts a defect rate by using the capacity degradation value.

[0014] For each of n driving curves, the defect rate prediction unit can multiply a corresponding weight value over time by a corresponding capacity degradation value to select a plurality of capacity degradation weight values as a population, generate m probability distributions by using a bootstrap technique on the population, generate a cumulative distribution function for each of the m probability distributions, and calculate a defect rate for each of the m cumulative distribution functions using a number of capacity degradation weight values deviating from a normal range in a normal distribution.

[0015] The defect rate prediction unit can detect weight values corresponding to j probabilities in order from a highest probability of a cumulative distribution function corresponding to a normal distribution having the highest probability among the m cumulative distribution functions.

[0016] The defect rate prediction unit can determine validity of a corresponding cumulative distribution function according to whether a probability distribution function corresponding to each of the m cumulative distribution functions can be designated.

[0017] When the defect rate exceeds a predetermined threshold value, the defect rate prediction unit can control the server to stop a prediction operation of the edge computing device.

[0018] The edge computing device can stop the prediction operation by control of the server, store the input data and the output data received from the IoT device during a stop period, and resume the prediction operation reflecting the stored input data and output data when a restart instruction is received from the server.

[0019] The defect rate prediction unit can notify the aging model module that the defect rate exceeds the threshold, and the aging model module can include a plurality of sets consisting of a plurality of formulas, and select one of the plurality of sets to change the aging model.

[0020] The data distribution method for performing an experiment on a battery pack including a plurality of battery cells using at least one experimental device according to the present invention includes: starting a distribution processing operation for an experimental result by an edge computing device; receiving input and output from at least one experimental device by an IoT device and starting the IoT device to enter a ready state for distribution processing; generating input data and output data by the IoT device by receiving and preprocessing the input and output from at least one experimental device; receiving an initial value of the input data and the input data and the output data by the edge computing device to predict information required for predicting a capacity degradation value by using an equivalent circuit model module and a thermal model module; and receiving the predicted information from the edge computing device and estimating a capacity retention rate and an internal resistance of each of the plurality of battery cells based on the predicted information by a server.

[0021] The data distribution method can further include predicting a capacity degradation value based on the capacity retention rate and the internal resistance, and predicting a defect rate based on the predicted capacity degradation value by the server.

[0022]

Advantageous Effects

[0023] Provided are a simulation system and a data distribution method capable of improving the accuracy and efficiency of battery cycle life prediction. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 FIG. 1 is a view illustrating a simulation system according to an embodiment.

[0025] Figure 2 FIG. 2 is a view illustrating a configuration of an edge computing device according to an embodiment.

[0026] Figure 3 FIG. 3 is a view illustrating some configurations of a server according to an embodiment.

[0027] Figure 4 FIG. 4 is a flowchart illustrating a data distribution processing operation of a simulation system according to an embodiment. DETAILED DESCRIPTION

[0028] Hereinafter, embodiments disclosed in the present specification will be described in detail with reference to the accompanying drawings. In the present specification, the same or similar components will be denoted by the same or similar reference numerals, and repetitive description thereof will be omitted. The terms "module" and "unit" used in the following description are used only to conveniently express the present specification. Accordingly, these terms have no meaning that distinguishes them from each other by themselves. Also, in describing the embodiments of the present specification, when it is determined that a detailed description of related art that is related to the present invention can obscure the gist of the present invention, it will be omitted. Also, the accompanying drawings are provided only to facilitate understanding of the embodiments disclosed in the present specification, and should not be interpreted as limiting the spirit disclosed in the present specification, and it should be understood that the present invention includes all modifications, equivalent forms, and substitutions without departing from the scope and spirit of the present invention.

[0029] The terms including ordinal numbers such as first, second, and the like will be used only to describe various components, and should not be construed as limiting the components. The terms are used only to distinguish one component from other components.

[0030] It should be understood that when one component is referred to as being "connected" or "coupled" to another component, it can be directly connected or coupled to the other component, or connected or coupled to the other component with another component interposed therebetween. On the other hand, it should be understood that when one component is referred to as being "directly connected or coupled" to another component, it can be connected or coupled to the other component without another component interposed therebetween.

[0031] It will also be understood that the terms "include" or "have" used in the present specification specify the presence of stated features, numbers, steps, operations, components, parts, or combinations thereof, but do not exclude the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0032] Embodiments perform a battery cycle life deterioration simulation by distributing and processing an edge computing function with an IoT (Internet of Things) device.

[0033] For example, in a simulation of predicting cycle life deterioration of a battery pack for an electric vehicle, by combining real-time input data / output data with IoT edge computing, a cycle life deterioration prediction model with high accuracy can be created at low cost.

[0034] A simulation system according to an embodiment includes an IoT device, an edge computing device, and a server as a distributed layer.

[0035] Hereinafter, embodiments will be described with reference to the accompanying drawings.

[0036] Figure 1 is a view illustrating a simulation system according to an embodiment.

[0037] As Figure 1 shown in FIG. 1, the simulation system 1 includes experimental devices 300 and 400, an IoT device 200, an edge computing device 100, and a server 10.

[0038] In Figure 1 the present embodiment, a cell cycler 300 and a battery pack simulator 400 are shown as two experimental devices 300 and 400, but the present invention is not limited thereto.

[0039] The cell cycler 300 is connected to a battery pack as an experimental target to charge and discharge the battery pack and measure battery information such as cell voltage, pack current, and cell temperature during charging and discharging. The charging and discharging pattern can be determined according to a drive profile.

[0040] The battery pack simulator 400 can set a virtual battery pack having the same electrical and chemical characteristics as the battery pack as an experimental target, charge and discharge the virtual battery pack, and predict battery information such as cell voltage, pack current, and cell temperature during charging and discharging. The charging and discharging pattern can be determined according to an operation profile.

[0041] The IoT device 200 receives input and output from the experimental devices, pre-processes data related to the received input and output, and transmits the pre-processed data to the edge computing device 100 or the server 10 according to the size of the pre-processed data. The IoT device 200 converts the received input and output into data that can be processed by the edge computing device 100 and the server 10, which is called pre-processing. If the size of the pre-processed data is greater than or equal to a predetermined size, the IoT device 200 transmits the corresponding data to the server 10, and if the size is less than or equal to the predetermined size, transmits the data to the edge computing device 100. In this case, the predetermined size as a criterion for classifying the size of data can be determined considering the size of data that can be processed by the edge computing device 100, processing speed, etc.

[0042] The edge computing device 100 can analyze data received from the IoT device 200 to generate data required to predict a capacity degradation value of a battery cell. For example, the IoT device 200 can transmit a pack voltage, an initial SOC, a current flowing through a battery pack, a cell temperature, the number of battery cells constituting a battery pack, a pack resistance, and a cooling performance to the edge computing device 100.

[0043] Figure 2 is a view showing a configuration of an edge computing device according to an embodiment.

[0044] The edge computing device 100 includes an equivalent circuit model (ECM) module 110 and a thermal model (ThM) module 120.

[0045] The equivalent circuit model module 110 receives a battery pack voltage, an initial SOC, a current flowing to the battery pack, a temperature of the battery cell, a number of battery cells constituting the battery pack, a resistance of the battery pack, etc. to predict an SOC of each battery cell, a battery pack voltage, a current flowing to the battery pack, a heat value of the battery cell, a current rate (C-Rate), etc. Various known methods can be applied as a specific method for predicting output data by applying input data to a circuit model equivalent to the equivalent circuit model module 110 of the battery pack.

[0046] For example, an equivalent circuit model (ECM) of the equivalent circuit model module 110 is an equivalent circuit constructed by simply simulating a complex internal caused by chemical characteristics of a battery with electrical elements such as a resistor and a capacitor, and the equivalent circuit model module 110 can predict an output of any input using the ECM. Specifically, the equivalent circuit model module 110 can update an SOC of the battery by performing current integration with respect to a current input of the ECM, and calculate an OCV according to the updated SOC. In addition, the equivalent circuit model module 110 can calculate a voltage drop according to the SOC and a temperature, according to each parameter of a resistance and a capacitance.

[0047] A current rate refers to a size of a charging current supplied to the battery pack or a discharging current supplied from the battery pack per unit time. For example, a 1C-Rate can refer to a current size in which the battery pack is fully charged in 1 hour, or a current size in which the battery pack is fully discharged in 1 hour.

[0048] The thermal model module 120 receives a heat value and a cooling performance as input and predicts a temperature of the battery cell. The heat value is data transferred from the equivalent circuit model module 110, and the cooling performance can vary depending on the battery pack. Various known methods can be applied as a specific method for predicting a temperature of each battery cell of the thermal model module 120 by receiving a heat value and a cooling performance.

[0049] For example, a heat value of a cell input by the equivalent circuit model module 110 can be calculated as Equation 1 below.

[0050] [Equation 1]

[0051] Heat value of battery cell = (correlation coefficient related to battery cell voltage drop) * (current flowing through battery cell) * (OCV - battery cell voltage according to SOC variation)

[0052] The average temperature value of the derived battery cell temperature is calculated by using a relationship equation of the battery cell temperature, the cooling water temperature, and the thermal resistance value, from which a value derived from the heat value of the cell converges to (battery cell mass) * (specific heat capacity) * (rate of change of battery cell temperature (derivative value)), and the average temperature value can be used as an input value of the aging model 11. Among the input data required by the edge computing device 100, the number of battery cells constituting the battery pack, the initial SOC, the cooling performance, etc. can be predetermined values according to experimental conditions.

[0053] The edge computing device 100 can transmit the predicted SOC of each battery cell, the battery pack voltage, the current flowing in the battery pack, the C-Rate, and the battery cell temperature to the server 10.

[0054] In Figure 2 , it is shown that the edge computing device 100 includes the equivalent circuit model module 110 and the thermal model module 120, however, one of the equivalent circuit model module 110 and the thermal model module 120 can be provided in the server 10. In addition, the edge computing device 100 can include modules required to predict the capacity degradation value in addition to the equivalent circuit model module 110 and the thermal model module 120.

[0055] The server 10 can receive the output data of the edge computing device 100 to predict the capacity degradation value. The output data of the edge computing device 100 can include data on the SOC of each battery cell, the battery pack voltage, the current flowing through the battery pack, the current rate (C-Rate), and the temperature of the battery cell. If necessary, the server 10 can receive more data required to predict the capacity degradation value from the IoT device 200 as well as the edge computing device 100.

[0056] Figure 3 is a view showing some configurations of a server according to an embodiment.

[0057] As Figure 3 shown, the server 10 includes an aging model (AgM) 11, a capacity degradation value calculation unit 12, and a defect rate calculation unit 12.

[0058] The aging model 11 can predict the capacity retention rate and internal resistance of each battery cell by using the SOC of each battery cell, the battery pack voltage, the current flowing through the battery pack, the current rate (C-Rate), and the battery cell temperature. As a specific method of predicting the increase in the capacity retention rate and internal resistance of each battery cell by using the SOC of each battery cell, the battery pack voltage, the current flowing through the battery pack, the current rate (C-Rate), and the battery cell temperature, various known methods can be applied as the aging model 11.

[0059] For example, based on a theory that explains a process of causing a fracture by accumulating an invisible fatigue damage to a structural material when an irregular load is repeatedly applied to the structural material (for example, Miner's rule), the aging model 11 can define a driving condition as a cycle, define a stopping condition as a calendar, calculate a capacity degradation value using the cycle and the calendar and an occupancy ratio of each thereof, and calculate a resistance increase value by correlating a resistance increase curve derived from a degradation experiment of a battery cell using the capacity degradation value derived in this process. The calculated capacity degradation value and resistance increase value are applied to a capacity retention rate and an internal resistance to be predicted. The capacity retention rate is a result of measuring an ability of a battery cell to retain stored energy during an extended open circuit period, and the internal resistance is a result of measuring a resistance component inside the battery cell. Although the aging model 11 has been described as predicting the internal resistance, the present application is not limited thereto, and can predict the internal resistance and / or an amount of increase in the internal resistance.

[0060] The capacity degradation value calculation unit 12 can receive the capacity retention rate and the internal resistance of each battery cell, and predict a capacity degradation value indicating a degree of capacity degradation of each battery cell. A method of mapping a capacity degradation value corresponding to the capacity retention rate and the internal resistance based on an experiment using a formula, a table, or the like can be applied as a specific method for the capacity degradation value calculation unit 12 to predict a capacity degradation value using the capacity increase amount and the internal resistance of each battery cell. The defect rate prediction unit 13 can store information on the capacity degradation value of each of the plurality of battery cells for each time unit, and predict a defect rate by using the capacity degradation value. A unit of the defect rate can be a parts per million (PPM).

[0061] The defect rate prediction unit 13 multiplies a corresponding capacity degradation value by a corresponding weight value depending on the time lapse of each of the n driving curves. The weight value corresponding to the time lapse can be discretely set in a predetermined time unit, and the predetermined time can change as time elapses or can be a constant time. Hereinafter, the multiplication of the capacity degradation value by the weight value is referred to as a capacity degradation weight value.

[0062] The defect rate prediction unit 13 selects a population by collecting all the plurality of capacity degradation weight values for each of the n driving curves, and generates m probability distributions for the population by using a bootstrap technique. The defect rate prediction unit 13 can select a k value m times by allowing iteration for the population according to the bootstrap technique. The defect rate prediction unit 13 constructs a probability distribution for k capacity degradation weight values, and this operation is performed m times.

[0063] The defect rate prediction unit 13 generates a cumulative distribution function for each of the m probability distributions. The defect rate prediction unit 13 can detect the weight value corresponding to the j probabilities in the order of the highest probability from among the normal distributions corresponding to each of the m cumulative distribution functions. That is, the weight value of the capacity degradation weight value corresponding to each of the j probabilities can be detected.

[0064] Also, the defect rate prediction unit 13 can determine the validity of the corresponding cumulative distribution function according to whether a probability distribution function corresponding to each of the m cumulative distribution functions can be designated. If there is a probability distribution function matching the cumulative distribution function, the corresponding cumulative distribution function is determined to be valid. In this case, the type of the probability distribution function is "beta", "Birnbaum Saunders", "exponential", "extreme value", "gamma", "generalized extreme value", "generalized Pareto", "inverse Gaussian", "logistic", "loglogistic", "lognormal", "Nakagami", "normal", "Rayleigh", "Rician", "t location-scale", "Weibull", "piecewise Pareto tail", etc.

[0065] The defect rate prediction unit 13 can calculate the defect rate with respect to each of the m cumulative distribution functions using the number of capacity degradation weight values outside the normal range in the normal distribution. For example, if the capacity degradation weight value is 30% or more, it can be determined to be outside the normal range. When the defect rate exceeds a predetermined threshold value, the server 10 can control the prediction operation of the edge computing device 100 to stop. The server 10 can instruct the edge computing device 100 to stop the prediction operation, and the edge computing device 100 can stop the prediction operation. During the stop period, the edge computing device 100 can store the input data and output data of the experimental devices 300 and 400 received from the IoT device 200. After that, when a restart instruction of the edge computing device 100 is received from the server 10, the edge computing device 100 can resume the prediction operation reflecting the stored input data and output data.

[0066] When the defect rate exceeds the threshold value, the defect rate prediction unit 13 notifies the aging model 11, and the aging model 11 changes the aging model used to predict the capacity retention rate and the internal resistance so that the defect rate does not exceed the threshold value. A plurality of sets consisting of various formulas applied to the aging model are provided, and the aging model 11 changes the aging model by selecting one of the plurality of sets. The aging model 11 can consider the trend of the capacity degradation value when selecting one of the plurality of sets. For example, the aging model 11 can select a corresponding set from the plurality of sets according to the change in the capacity degradation value over time.

[0067] Figure 4 is a flowchart illustrating a data distribution processing operation of a simulation system according to an embodiment.

[0068] The edge computing device 100 starts a distribution processing operation on the experimental results (S1). An experimenter's instruction for the distribution processing operation is input to the edge computing device 100, and then the edge computing device 100 can transmit an instruction for starting the distribution processing operation to the IoT device 200.

[0069] The IoT device 200 receives inputs and outputs from the experimental devices 300 and 400 and starts to the ready state for distribution processing (S2).

[0070] The IoT device 200 can set conditions based on a drive curve (S3). The drive curve can include conditions such as power (watt / hour), drive mode (idle / plugged / drive), and temperature (°C).

[0071] The IoT device 200 receives inputs and outputs from the experimental devices 300 and 400 (S4). As inputs and outputs received from the experimental devices 300 and 400, it can include a battery operation mode (rest / charge / discharge), a battery pack voltage, a current flowing through the battery pack, a temperature, etc. The IoT device 200 pre-processes the received inputs and outputs to generate input data and output data.

[0072] The edge computing device 100 receives an initial value of the input data and the pre-processed input data and output data from the IoT device 200 (S5). In this case, the input data and output data received by the edge computing device 100 can be data classified as suitable for the size of the edge computing device 100 by the IoT device 200.

[0073] The edge computing device 100 can predict information required for predicting a capacity degradation value based on the input data by using the equivalent circuit model module 110 and the thermal model module 120 (S6). The required information can include a battery pack voltage, an SOC, a current flowing through the battery pack, a C-rate, a thermal value, etc.

[0074] The information predicted by the edge computing device 100 is transmitted to the server 10 (S7).

[0075] The server 10 predicts the capacity retention rate and internal resistance based on the transmitted predicted information (S8).

[0076] The server 10 predicts the capacity degradation value based on the capacity retention rate and internal resistance (S9).

[0077] The server 10 predicts the defect rate based on the predicted capacity degradation value (S10).

[0078] While the present application has been described in connection with the embodiments currently considered to be the best mode for practicing the application, it is to be understood that the application is not limited to the disclosed embodiments. Rather, it is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A simulation system for performing experiments on a battery pack comprising multiple individual cells using at least one experimental apparatus, comprising: An IoT device receives inputs and outputs from at least one experimental device and preprocesses the inputs and outputs to generate input data and output data. An edge computing device, the edge computing device being used to predict information for predicting capacity degradation values ​​of multiple battery cells based on received data in the input data and the output data by using an equivalent circuit model module and a thermal model module. as well as The server receives the predicted information from the edge computing device, predicts the capacity degradation value of multiple battery cells using an aging model module, and predicts the defect rate based on the predicted capacity degradation value. The equivalent circuit model module receives the temperature of the individual battery cells from the IoT device and predicts the calorific value of the individual battery cells. The thermal model module receives the calorific value from the equivalent circuit model module and receives input of the cooling performance of the battery pack to predict the temperature of the individual battery cells.

2. The simulation system according to claim 1, wherein The equivalent circuit model module further receives at least one of the following from the IoT device: battery pack voltage, initial SOC, current flowing into the battery pack, number of battery cells constituting the battery pack, and resistance of the battery pack, and further predicts at least one of the following: SOC of each battery cell, battery pack voltage, current flowing into the battery pack, and current rate C-Rate.

3. The simulation system according to claim 1, wherein The aging model module receives at least one of the following data regarding the SOC of each of the battery cells, the battery pack voltage, the current flowing into the battery pack, the current rate C-Rate, and the temperature of the battery cells, to predict the capacity retention and internal resistance of each of the plurality of battery cells.

4. The simulation system according to claim 1, wherein The server also includes: A capacity degradation value calculation unit receives the capacity retention rate and internal resistance of each of the battery cells to predict a capacity degradation value indicating the degree of capacity degradation of each battery cell. as well as A defect rate prediction unit stores information about the capacity degradation value of each of the plurality of battery cells for each time unit and predicts the defect rate by using the capacity degradation value.

5. The simulation system according to claim 4, wherein For each of the n driving curves, the defect rate prediction unit multiplies the weight value corresponding to the elapsed time by the corresponding capacity degradation value to select multiple capacity degradation weight values ​​as a group. It generates m probability distributions by using a guiding technique on the group, generates a cumulative distribution function for each of the m probability distributions, and calculates the defect rate for each of the generated m cumulative distribution functions using the number of capacity degradation weight values ​​that deviate from the normal range of the normal distribution.

6. The simulation system according to claim 5, wherein The defect rate prediction unit detects weight values ​​corresponding to j probabilities in the order of the highest probability of the cumulative distribution function corresponding to the normal distribution with the highest probability among the normal distributions of each of the m cumulative distribution functions.

7. The simulation system according to claim 5, wherein The defect rate prediction unit determines the validity of the corresponding cumulative distribution function based on whether a probability distribution function corresponding to each of the m cumulative distribution functions can be specified.

8. The simulation system according to claim 5, wherein When the defect rate exceeds a predetermined threshold, the defect rate prediction unit controls the server to stop the prediction operation of the edge computing device.

9. The simulation system according to claim 8, wherein The edge computing device stops the prediction operation under the control of the server. During the stop period, it stores the input data and output data received from the IoT device. When it receives a restart command from the server, it resumes the prediction operation that reflects the stored input data and output data.

10. The simulation system according to claim 8, wherein The defect rate prediction unit notifies the aging model module that the defect rate exceeds the threshold, and The aging model module includes multiple sets of formulas, and one of the multiple sets is selected to change the aging model.

11. A data distribution processing method for performing experiments on a battery pack comprising multiple individual cells using at least one experimental apparatus, comprising: The experimental results are then distributed and processed using edge computing devices. The IoT device receives input and output from at least one experimental device and initiates the IoT device into a ready state for distributed processing. The IoT device generates input data and output data by receiving and preprocessing inputs and outputs from at least one experimental device; The edge computing device receives the initial value of the input data, as well as the input data and the output data, to predict the information needed to predict capacity degradation values ​​by using an equivalent circuit model module and a thermal model module. as well as The predicted information is received from the edge computing device, and the server estimates the capacity retention and internal resistance of each of the multiple battery cells based on the predicted information. The data distribution processing method further includes: The equivalent circuit model module receives the temperature of the battery cell from the IoT device and predicts the calorific value of the battery cell. The thermal model module receives the calorific value from the equivalent circuit model module and receives input of the cooling performance of the battery pack to predict the temperature of the individual battery cells.

12. The data distribution processing method according to claim 11, further comprising: The capacity degradation value is predicted based on the capacity retention rate and the internal resistance, and the server predicts the defect rate based on the predicted capacity degradation value.

Images