Big-data-based battery diagnostic system and control method therefor

1 . A method of controlling a big-data-based battery diagnostic system having a battery data server, the method comprising:

receiving, by the battery data server, first state information related to a first battery module from a first vehicle in which the first battery module used as at least one power source is disposed;

learning, by the battery data server, the first state information through an artificial neural network based on at least one factor affecting the voltage of the battery;

deriving, by the battery data server, a reference voltage that serves as a criterion for safety diagnosis of battery cells for the first vehicle based on the learned first state information;

transmitting, by the battery data server, the reference voltage to a second vehicle in which the second battery module used as a power source is disposed;

measuring, by the second vehicle, second state information related to the second battery module;

determining, by the second vehicle, whether at least one abnormal battery cell exists in the second battery module by comparing second state information related to the second battery module with the reference voltage related to the first battery module thereby determining whether the battery state of the second vehicle is in an abnormal state relative to that of the first vehicle;

determining, by the second vehicle, a diagnostic level among a plurality of diagnostic levels based on a number of abnormal cells and a number of cells that remain in an abnormal state for a preset time, when at least one abnormal battery cell of the second battery module is determined to exist; and

inducing, by the second vehicle, a preliminary diagnosis or performing preventive maintenance, in accordance with the determined diagnostic level,

wherein deriving the reference voltage for the first vehicle includes setting, as the reference voltage, at least one of a maximum voltage, a minimum voltage, or an average voltage for each of a first battery cell, the first battery module, and a first battery pack.

2 . The method according to claim 1 , further comprising:

transmitting, by the second vehicle, a result of determination to the battery data server;

learning, by the battery data server, the result through the artificial neural network, and updating, by the battery data server, the reference voltage for the first vehicle to be an updated reference voltage for the first vehicle based on the learned diagnostic data; and

transmitting, by the battery data server, the updated reference voltage for the first vehicle to the first and second vehicles.

3 . The method according to claim 2 , wherein the deriving a reference voltage for the first vehicle comprises:

preprocessing a plurality of first state information in a form suitable for the purpose and method of data analysis;

extracting the at least one factor affecting the voltage of the battery as a feature value of the first vehicle from the preprocessed first state information;

learning based on the extracted feature value of the first vehicle through artificial neural network; and

deriving the reference voltage for the first vehicle based on the learned feature value of the first vehicle.

4 . The method according to claim 3 , wherein when the first vehicle has at least one feature value, the deriving the reference voltage for the first vehicle comprises deriving at least one reference voltage for the first vehicle in response to the at least one feature value of the first vehicle.

5 . The method according to claim 4 , wherein the feature value of the first vehicle comprises at least one of a vehicle identification number (VIN) of the first vehicle, and SOH/SOC/temperature/coolant temperature/current load related to the first battery module.

6 . The method according to claim 3 , wherein the second vehicle is the same model as a plurality of mass-produced first vehicles.

7 . The method according to claim 1 , wherein the determining further comprising:

monitoring, when at least one abnormal battery cell exists, the abnormal battery cell for a predetermined time;

generating and counting a fault code when a voltage of the abnormal battery cell deviates from the reference voltage for a preset time or more;

generating and counting a warning code when a voltage of the abnormal battery cell does not deviate from the reference voltage for a preset time or more;

determining one of preset levels based on a number of the fault code or the warning code; and

generating a battery state data on the current state of the second vehicle in response to the determined preset level.

8 . The method according to claim 7 , wherein:

each of the preset levels is a bit in which a battery state predefined according to the number of fault codes or the number of warning codes.

9 . The method according to claim 8 , comprising notifying a driver of the second vehicle according to the generated battery state data with the fault code or the warning code using a display or a speaker disposed within the second vehicle.

10 . A non-transitory computer-readable medium containing a program for executing the method of controlling a big-data-based battery diagnostic system according to claim 1 .

11 . A battery diagnostic system based on big data, comprising:

a first vehicle in which the first battery module used as at least one power source is disposed;

a battery data server having a processor configured to:

receive first state information related to the first battery module from the first vehicle;

learn the collected first state information through an artificial neural network based on at least one factor affecting the voltage of the battery; and

derive a reference voltage that serves as a criterion for safety diagnosis of battery cells for the first vehicle based on the learned first state information; and

a second vehicle in which the second battery module used as a power source is disposed,

wherein the second vehicle is configured to;

receive the reference voltage from the battery data server; and

measuring second state information related to the second battery module;

determine whether an abnormal battery cell exists in the second battery module by comparing second state information related to the second battery module with the reference voltage related to the first battery module thereby determining whether the battery state of the second vehicle is in an abnormal state relative to that of the first vehicle;

determine a diagnostic level among a plurality of diagnostic levels based on a number of abnormal cells and a number of cells that remain in an abnormal state for a preset time, when at least one abnormal battery cell of the second battery module is determined to exist; and

induce a preliminary diagnosis or perform preventive maintenance in accordance with the determined diagnostic level,

wherein the battery data server sets, as the reference voltage, at least one of the maximum voltage, the minimum voltage, or the average voltage for each of a first battery cell, the first battery module, and a first battery pack.

12 . The battery diagnostic system according to claim 11 , wherein:

the second vehicle receive a result of determination from the battery data server; and

the processor is configured to learn the result through artificial neural network update the reference voltage for the first vehicle to be an updated reference voltage for the first vehicle based on the learned diagnostic data, and transmit the updated reference voltage for the first vehicle to the first and second vehicles.

13 . The battery diagnostic system according to claim 12 , wherein the processor comprises:

a preprocessing unit configured to preprocess a plurality of first state information in a form suitable for the purpose and method of data analysis;

an extraction unit configured to extract the at least one factor affecting the voltage of the battery as a feature value of the first vehicle from the preprocessed first state information;

a learning unit configured to learn based on the extracted feature value of the first vehicle through the artificial neural network; and

a derivation unit configured to derive the reference voltage for the first vehicle based on the learned feature value of the first vehicle.

14 . The battery diagnostic system according to claim 13 , wherein when the first vehicle has at least one feature value, the derivation unit derives at least one reference voltage for the first vehicle in response to the at least one feature value of the first vehicle.

15 . The battery diagnostic system according to claim 14 , wherein the feature value of the first vehicle comprises at least one of a vehicle identification number (VIN) of the first vehicle, and SOH/SOC/temperature/coolant temperature/current load related to the first battery module.

16 . The battery diagnostic system according to claim 13 , wherein the second vehicle is the same model as a plurality of mass-produced first vehicles.

17 . The battery diagnostic system according to claim 11 , wherein when at least one abnormal battery cell exists, the second vehicle further configured to:

monitor the abnormal battery cell for a predetermined time;

generate and count a fault code when a voltage of the abnormal battery cell deviates from the reference voltage for a preset time or more;

generate and count a warning code when a voltage of the abnormal battery cell does not deviate from the reference voltage for a preset time or more;

determine one of preset levels based on a number of the fault code or the warning code; and

generating a battery state data on the current state of the second vehicle in response to the determined preset level.

18 . The battery diagnostic system according to claim 17 , wherein:

each of the preset levels is a bit in which a battery state predefined according to the number of fault codes or the number of warning codes.

19 . The battery diagnostic system according to claim 18 , wherein the second vehicle is configured to notify a driver of the second vehicle according to the generated battery state data.