KR20250029508A - Method for predicting state of health of battery of electric vehicle in the future based on analysis of connected car vehicle data - Google Patents

Abstract

A method for predicting the residual value of an electric vehicle battery is disclosed. The method for predicting the residual value of an electric vehicle battery according to the present invention comprises an IoT terminal which is connected to an OBD-II terminal and a CAN module of an electric vehicle and provides vehicle data including time information and data for calculating the number of accumulated charges at the time of data transmission and data for remaining battery value (SOH) to a platform via the Internet, and a platform which is connected to the Internet and stores data received from the IoT, and the step of receiving and storing vehicle data including data for calculating the number of accumulated charges at the time of data transmission and data for remaining battery value (SOH) from various types of electric vehicles equipped with various types of batteries on the platform, the step of dividing the received vehicle data by vehicle type and battery type and configuring a corresponding data set D/B, the step of calculating the number of accumulated charges from the data for calculating the number of accumulated charges of the data set D/B configured by vehicle type and battery type, the step of mapping the number of accumulated charges calculated by vehicle type and battery type and the battery remaining value (SOH) classified by vehicle type and battery type, and the step of mapping the mapped number of accumulated charges by vehicle type and battery type and the battery remaining value (SOH) of the mapped vehicle type and battery type. A step of generating a baseline prediction model by vehicle type and battery type using a regression model in which the cumulative number of charging times at the time of data transmission is an independent variable and the battery residual value (SOH) at the time of data transmission is a dependent variable, and a step of predicting the battery residual value (SOH in the future) at a future point in time according to the cumulative number of charging times is performed using the baseline prediction model by vehicle type and battery type.

Description

{METHOD FOR PREDICTING STATE OF HEALTH OF BATTERY OF ELECTRIC VEHICLE IN THE FUTURE BASED ON ANALYSIS OF CONNECTED CAR VEHICLE DATA}

The present invention relates to a method for predicting the residual value of an electric vehicle battery, and more specifically, to a method for predicting the future residual value of an electric vehicle battery through analysis of connected car vehicle data, which can provide a predicted value of the battery residual value at a future point in time and a trend of changes in the residual value at a future point in time through analysis of connected car vehicle data.

The current method for predicting the state of health (SOH) of an electric vehicle battery is known through patent publication No. 10-2021-0134137, patent publication No. 10-2021-0116801, and registered patent No. 10-0836408, and the BMS (Battery Management System) of electric vehicles currently on the market predicts and provides the current SOH value. The current method for predicting the state of health (SOH) of an electric vehicle battery can be easily predicted by collecting measurement data such as charging voltage, charging current, charging capacity, and temperature per charging cycle of the battery in the past based on the present.

However, the inventions such as Patent Publication No. 10-2021-0134137, Patent Publication No. 10-2021-0116801, and Patent Registration No. 10-0836408, which are methods for predicting the residual value of electric vehicle batteries at the present time, cannot be applied to predicting the residual value of batteries at a future time based on the present. In order to predict the residual value of batteries at a future time based on the present, measurement data such as charge voltage, charge current, charge capacity, and temperature per battery charge cycle from the present to the future time are required. However, since electric vehicle operation or charging has not actually occurred from the present to the future time, measurement data such as charge voltage, charge current, charge capacity, and temperature per battery charge cycle cannot be obtained and used.

However, in the case of electric vehicle batteries, there is a great need to predict not only the SOH at present but also the SOH in the future based on the present. If electric vehicle drivers predict the SOH in the future, electric vehicle users can anticipate the battery replacement period in advance, and by checking the fluctuation status and fluctuation factors of the future SOH through the battery life history and the predicted SOH graph at regular intervals during the electric vehicle ownership period, existing battery management methods and driving habits can be improved.

Korean Patent Publication No. 10-2021-0134137 (Publication Date: 2021.11.09) Korean Patent Publication No. 10-2021-0116801 (Publication Date: 2021.09.28) Korean Patent Registration No. 10-0836408 (Registration Date: 2008.06.02)

The present invention has been made to solve the problems of the conventional electric vehicle battery residual value prediction method described above, and the first task that the present invention seeks to solve is to provide a method for predicting the future residual value (SOH in the future) of an electric vehicle battery by analyzing connected car vehicle data, which can predict the future point-in-time residual value (SOH) of an electric vehicle battery by vehicle type and battery type using data acquired from an OBD-II terminal and a CAN module of an electric vehicle in a connected car platform connected to the electric vehicle by an IoT terminal.

The second problem to be solved by the present invention is to provide a method for predicting the future state of health (SOH) of an electric vehicle battery through analysis of connected car vehicle data, which can predict the future state of health (SOH) of an individual electric vehicle by using data acquired from an OBD-II terminal and a CAN module of an electric vehicle in a connected car platform connected to an electric vehicle and an IoT terminal.

The third problem that the present invention seeks to solve is to provide a method for predicting the future state of health (SOH) of an electric vehicle battery through analysis of connected car vehicle data that can predict the future state of health (SOH) of the battery by reflecting all existing vehicle data accumulated in real time.

The fourth problem to be solved by the present invention is to provide a method for predicting the future state of health (SOH) of an electric vehicle battery through analysis of connected car vehicle data, which can predict the future state of health (SOH) of the battery over time for each electric vehicle and provide the prediction to electric vehicle users.

The first task of the present invention described above is to provide an IoT terminal which is connected to an OBD-II terminal and a CAN module of an electric vehicle and provides vehicle data including time information and data for calculating the number of accumulated charges at the time of data transmission and battery remaining value (SOH) data to a platform via the Internet, and a platform which is connected to the Internet and stores data received from the IoT, and on the platform, receives and stores vehicle data including data for calculating the number of accumulated charges at the time of data transmission and battery remaining value (SOH) data from various types of electric vehicles equipped with various types of batteries, a step of dividing the received vehicle data by vehicle type and battery type and configuring a data set D/B corresponding to each, a step of calculating the number of accumulated charges from the data for calculating the number of accumulated charges of the data set D/B configured by vehicle type and battery type, a step of mapping the number of accumulated charges calculated by vehicle type and battery type and the SOH of the battery classified by vehicle type and battery type, and a step of mapping the mapped number of accumulated charges by vehicle type and battery type and the battery remaining value (SOH) of the mapped vehicle type and battery type. The problem can be solved by performing a step of generating a baseline prediction model by vehicle type and battery type using a regression model in which the cumulative number of charges at the time of data transmission is an independent variable and the battery SOH at the time of data transmission is a dependent variable, and a step of predicting the battery SOH in the future according to the cumulative number of charges using the baseline prediction model by vehicle type and battery type.

The second task of the present invention described above comprises an IoT terminal which is connected to an OBD-II terminal and a CAN module of an electric vehicle and provides vehicle data including time information and data for calculating the number of accumulated charges at the time of data transmission and data for remaining battery life (SOH) to a platform via the Internet, and a platform which is connected to the Internet and stores data received from the IoT, and in the platform, receives and stores vehicle data including data for calculating the number of accumulated charges at the time of data transmission from individual electric vehicles and data for remaining battery life (SOH), a step of configuring a dataset D/B for vehicle data of individual electric vehicles, a step of calculating the number of accumulated charges of individual electric vehicles from the data for calculating the number of accumulated charges of the individual electric vehicles dataset D/B, a step of mapping the calculated number of accumulated charges of individual electric vehicles and the SOH of the battery of the individual electric vehicles stored in the dataset, and a step of using the number of accumulated charges at the time of data transmission as an independent variable from the mapped number of accumulated charges of individual electric vehicles and the SOH of the battery at the time of data transmission. The problem can be solved by performing a step of creating a baseline prediction model for each electric vehicle by a regression model with the state of health (SOH) as a dependent variable, and a step of predicting the future state of health (SOH) of the battery according to the cumulative number of charging cycles using the baseline prediction model for each electric vehicle.

The third task of the present invention described above can be solved by periodically updating the baseline prediction model for each vehicle type and battery type or the baseline prediction model for each electric vehicle using vehicle data collected in real time.

The fourth object of the present invention described above can be solved by the platform applying the cumulative number of charges of an individual electric vehicle to a baseline prediction model for each vehicle type and battery type and obtaining a function of the cumulative number of charges of an individual electric vehicle and time, thereby predicting a future point-in-time battery remaining value (SOH in the future) of an individual electric vehicle over time, or applying the cumulative number of charges of an individual electric vehicle to a baseline prediction model for the individual electric vehicle and obtaining a function of the cumulative number of charges of the individual electric vehicle and time, thereby predicting a future point-in-time battery remaining value (SOH in the future) of an individual electric vehicle over time, and providing the result of the prediction of the future point-in-time battery remaining value (SOH in the future) over time to a smartphone of an individual electric vehicle user.

According to the present invention, vehicle data is received and stored from IoT terminals connected to a plurality of electric vehicles, and the cumulative number of charging cycles calculated by vehicle type and battery type and the battery remaining value (SOH) classified by vehicle type and battery type are mapped, and a regression model is used in which the cumulative number of charging cycles at the time of data transmission is an independent variable and the battery remaining value (SOH) at the time of data transmission is a dependent variable. Therefore, a relationship function between the future cumulative number of charging cycles and the battery remaining value (SOH in the future) can be obtained from vehicle data at a time before the present, and through this, the battery remaining value (SOH in the future) at a future time can be predicted according to the cumulative number of charging cycles by vehicle type and battery type. In addition, by calculating and applying the average cumulative number of charging cycles or the average charging cycle for a unit period of an individual electric vehicle, the battery remaining value (SOH in the future) at a future time can be predicted according to the passage of time of an individual electric vehicle.

In addition, according to the present invention, vehicle data is received and stored from an IoT terminal connected to an individual electric vehicle, and the cumulative number of charging cycles calculated from the vehicle data and the battery remaining value (SOH) included in the vehicle data are mapped, so that a regression model is used in which the cumulative number of charging cycles at the time of data transmission is an independent variable and the battery remaining value (SOH) at the time of data transmission is a dependent variable. Therefore, a relationship function between the future cumulative number of charging cycles and the battery remaining value (SOH in the future) can be obtained from vehicle data at a time before the present, and through this, the battery remaining value (SOH in the future) at a future time point according to the cumulative number of charging cycles of an individual electric vehicle can be predicted. In addition, by calculating and applying the average cumulative number of charging cycles or the average charging cycle for a unit period of an individual electric vehicle, the battery remaining value (SOH in the future) at a future time point according to the passage of time of an individual electric vehicle can be predicted.

In addition, according to the present invention, since the baseline prediction model predicting the battery remaining value (SOH in the future) at a future point in time is periodically updated using vehicle data collected in real time, an accurate prediction can be made by reflecting all existing vehicle data accumulated in real time whenever predicting the battery remaining value (SOH in the future).

In addition, according to the present invention, electric vehicle users can receive prediction results for the battery remaining value (SOH in the future) by vehicle type and battery type and the battery remaining value (SOH in the future) of each electric vehicle, as well as the predicted results for the battery remaining value (SOH in the future) changed by prediction period in the form of graphs or diagrams, and check them on a smartphone app. Therefore, the battery replacement period can be predicted in advance, and this can motivate users to improve existing battery management methods or driving habits.

Figure 1 is a system configuration diagram for implementing a method for predicting the future residual value of an electric vehicle battery through analysis of connected car vehicle data according to the present invention.
FIG. 2 is a flowchart of a method for predicting the future residual value of an electric vehicle battery through analysis of connected car vehicle data according to an embodiment of the present invention.
FIG. 3 is a flowchart of a method for predicting the future residual value of an electric vehicle battery through analysis of connected car vehicle data according to another embodiment of the present invention.
FIG. 4 is a diagram of a baseline prediction model according to one embodiment of the present invention.
FIG. 5 is a diagram of a baseline prediction model according to another embodiment of the present invention.
Figure 6 illustrates a state in which analysis of changes in the predicted future residual value of a battery at each predicted time point according to the present invention is transmitted to a smartphone app and displayed.
FIG. 7 illustrates a state in which the prediction results of the future point-in-time residual value of a battery predicted by vehicle type and battery type according to the present invention and the prediction results of the future point-in-time residual value of a battery predicted for an individual electric vehicle are transmitted and displayed on a smartphone app.

Hereinafter, with reference to the attached drawings, a specific embodiment of a method for predicting the future residual value of an electric vehicle battery through analysis of connected car vehicle data according to the present invention will be described in detail.

As illustrated in FIG. 1, a system for implementing a method for predicting the future residual value of an electric vehicle battery through connected car vehicle data analysis according to the present invention may be configured to include an IoT terminal (1) connected to an OBD-II terminal and a CAN module of an electric vehicle (204), a platform (20) capable of receiving data from the IoT terminal (1) via the Internet (203), and a smartphone (25) capable of receiving data from the platform (20) via the Internet (203).

The above IoT terminal (1) is connected to the OBD-II terminal (200) and CAN module (202) of the electric vehicle (204) and is equipped with a wireless modem (7) so as to provide vehicle data including time information, data for calculating the cumulative number of charges at the time of data transmission, and battery remaining health (SOH) data to the platform (20) periodically or whenever a status changes or an event occurs via the Internet. The above IoT terminal (1) can receive and provide vehicle (telematics) information and BMS (Battery Management System) information of the electric vehicle from the OBD-II terminal (200) and CAN module (202).

The above IoT terminal (1) can be configured to include an OBD interface (13) capable of receiving data from an OBD-II terminal, a CAN transmission/reception module (11) capable of transmitting and receiving CAN protocol data, a wireless modem (7) connected to a wireless Internet, a microcontroller unit (9) that controls these components, a terminal battery (15), a terminal charging circuit (19), and a terminal power controller (17).

The above IoT terminal (1) can provide data collection time as time information, can provide GPS longitude data and latitude data as space information, can provide vehicle status (driving, parking, charging), driving distance per trip (km/trip), speed, acceleration, number of rapid accelerations per trip (times/trip), number of rapid decelerations per trip (times/trip), etc. as driving information, and can provide battery charge status (stationary charge - wired charge while parked, driving charge - wireless charge while driving, not charged, fully charged), charge type (slow, rapid, ultra-rapid), battery charge rate (SOC: State of Charge), battery remaining value (SOH: State of Health), maximum charging power (Kw), energy consumption per trip (Kwh/trip), battery module temperature (℃), battery pack current (A), battery pack voltage (V), minimum cell voltage (mV), battery coolant inlet temperature (℃), etc. periodically or when a state changes or an event occurs.

More specifically, the IoT terminal (1) includes data acquisition time, vehicle identification number, indoor temperature, outdoor temperature, battery charge rate (SOC: State of Charge), allowable charge power, allowable discharge power, main relay status, slow charge port connection status, slow charge status, rapid charge port connection status, rapid charge status, ultra-rapid charge port connection status, ultra-rapid charge status, pack current, pack voltage, module maximum temperature, module minimum temperature, module temperature list, module temperature list, battery internal temperature, maximum cell voltage, maximum voltage cell number, minimum cell voltage, minimum voltage cell number, battery fan status (provides speed value), auxiliary battery voltage, cumulative charge current, cumulative discharge current, cumulative charge power, cumulative discharge power, operation time, BMS operation, inverter capacitor voltage, Driver Motor Speed 1, Driver Motor Speed 2, insulation resistance, number of cells, cell voltage List, Airbag H/wire Duty, cell voltage distribution, HVAC list 1, HVAC list 2, battery remaining value (SOH: State of Health), maximum deterioration cell number, minimum deterioration degree, minimum deterioration cell number, SOC display rate, e-mobility speed km/h, module average temperature, battery power, driving distance km, rapid charge relay status, charger connection status, estimated charging time (seconds), trip charging power, trip discharging power, battery coolant inlet temperature, battery PRA busbar temperature, battery LTR rear end temperature, V2L connection status, ignition status, DTC code occurrence list, last ignition ON time (including while driving), ignition battery voltage, door open/close status, trunk open/close status, headlight status, TPMS warning light status, door opening/closing status, electric charging cable connection status, battery charging status (stationary charging - wired charging while parked, driving charging - wireless charging while driving, not charged, charging complete), driving distance, battery charge level, tiger air pressure (front wheel), tire air pressure (front right wheel), tire air pressure (rear left wheel), tire air pressure (rear right wheel), side brake status, Vehicle speed, vehicle gear status, vehicle gear number, TRIP distance, drivable distance, TRIP fuel consumption, TRIP average fuel efficiency, TRIP average speed, TRIP total stopping time, TRIP total stopping time fuel consumption, number of sudden accelerations, number of sudden decelerations, number of sudden starts, number of sudden stops, LTE (or 5G, etc.) reception sensitivity, LTE (or 5G, etc.) reception quality, LTE (or 5G, etc.) transmission quality, GPS data, etc. can be provided periodically or when a status changes or an event occurs.

The above platform (20) is connected to the Internet (203) and stores data received from the IOT (1), and executes a method for predicting the future residual value of an electric vehicle battery through analysis of connected car vehicle data according to the present invention using the stored data. The above platform may include servers such as a web server and a DB server, network equipment, and security equipment.

The method for predicting the future residual value of an electric vehicle battery through analysis of connected car vehicle data according to the present invention can predict the future residual value (SOH in the future) by vehicle type and battery type, or predict the future residual value (SOH in the future) for an individual electric vehicle.

As illustrated in FIG. 2, in order to predict the state of health in the future by vehicle type and battery type, the platform (20) performs a step (101a) of receiving and storing vehicle data including data for calculating the number of accumulated charges at the time of data transmission and battery state of health (SOH) data from various types of electric vehicles equipped with various types of batteries, a step (102a) of dividing the received vehicle data by vehicle type and battery type and forming a corresponding data set D/B, and a step (103a) of calculating the number of accumulated charges from the data for calculating the number of accumulated charges in the data set D/B formed by vehicle type and battery type.

The above-mentioned data for calculating the cumulative charging count is one or more of the following: vehicle status (driving, parking, charging), battery charging status (stationary charging - wired charging while parked, driving charging - wireless charging while driving, not charged, charging complete) data, charger connection status (slow charging port connection status, rapid charging port connection status, ultra-rapid charging port connection status) data, charging type (slow charging status, rapid charging status, ultra-rapid charging status) data, electric charging cable connection status data, battery charging status (stationary charging - wired charging while parked, driving charging - wireless charging while driving, not charged, charging complete) data. If the number of times these status values are stored is counted, the cumulative charging count can be calculated.

Battery SOH data is the battery SOH at the time of data transmission predicted by collecting measurement data such as charging voltage, charging current, charging capacity, and temperature by charging cycle unit from the BMS of the electric vehicle. Currently, most electric vehicle BMS predict and store the battery SOH by charging cycle unit or at a certain period.

One of the features of the present invention is that, in the platform (20), the step (104a) of mapping the accumulated number of charges calculated for each vehicle type and battery type and the battery remaining value (SOH) classified for each vehicle type and battery type is performed, and the step (105a) of generating a baseline prediction model for each vehicle type and battery type by a regression model in which the accumulated number of charges at the time of data transmission is an independent variable and the battery remaining value (SOH) at the time of data transmission is a dependent variable from the mapped accumulated number of charges and the battery remaining value (SOH) for each vehicle type and battery type is performed. The baseline prediction model for each vehicle type and battery type can be periodically updated (106a) using vehicle data collected in real time. Accordingly, since the vehicle data reflected according to the prediction time is updated, the baseline prediction model for each vehicle type and battery type may also be different, and the predicted future remaining value (SOH in the future) may also be different.

Fig. 4 is an example of a baseline prediction model generated by vehicle type and battery type. Using this baseline prediction model by vehicle type and battery type, the platform (20) can predict the battery remaining value (SOH in the future) at a future point according to the cumulative number of charges of individual electric vehicles using the same vehicle type and battery (108a). Or, it can predict at what cumulative number of charges in the future the remaining value (SOH in the future) reaches the battery replacement point. Normally, the time when the SOH is 80% is considered the battery replacement point.

The above platform (20) can derive a function of the cumulative number of charging cycles and time of each electric vehicle from the accumulated vehicle data of each electric vehicle (108a). Using the function of the cumulative number of charging cycles and time of each electric vehicle derived in this way, the battery remaining value (SOH in the future) of each electric vehicle can be predicted over time (109a).

As shown in FIGS. 6 and 7, the battery remaining value (SOH in the future) for a future point in time can be transmitted to the electric vehicle user's smartphone app (27) and displayed on the smartphone app (27).

As shown in Fig. 6, the future battery residual value (SOH in the future) of each electric vehicle can be predicted over time by changing the prediction time. Fig. 6 shows the future battery residual value (SOH in the future) based on the prediction time of last month and the future battery residual value (SOH in the future) based on the prediction time of this month. In addition, it shows the expected battery performance (residual value) one year later and the expected battery performance (residual value) three years later.

In the embodiment illustrated in FIG. 2, a baseline prediction model for the battery residual value at a future point in time is generated using a dataset by vehicle type and battery type, and the battery residual value (SOH in the future) of an individual electric vehicle at a future point in time is predicted through this. However, when the cumulative number of charging cycles for an individual electric vehicle is sufficiently accumulated, the battery residual value (SOH in the future) that sufficiently reflects the characteristics of an individual electric vehicle can be predicted using a method as illustrated in FIG. 3.

As illustrated in FIG. 3, in order to predict the battery remaining value (SOH) in the future reflecting the characteristics of each electric vehicle, the platform (20) performs a step (101b) of receiving and storing vehicle data including data for calculating the number of cumulative charges at the time of data transmission from each electric vehicle and battery remaining value (SOH) data, a step (102b) of configuring a dataset D/B for vehicle data of each electric vehicle, and a step (103b) of calculating the number of cumulative charges of each electric vehicle from the data for calculating the number of cumulative charges of the dataset D/B of each electric vehicle.

The above data for calculating the cumulative charging count is one or more of the following data: vehicle status (driving, parking, charging) data of each electric vehicle, battery charging status (stationary charging - wired charging while parked, driving charging - wireless charging while driving, not charged, charging complete) data, charger connection status (slow charging port connection status, rapid charging port connection status, ultra-rapid charging port connection status) data, charging type (slow charging status, rapid charging status, ultra-rapid charging status) data, electric charging cable connection status data, battery charging status (stationary charging - wired charging while parked, driving charging - wireless charging while driving, not charged, charging complete) data. If the number of times these status values are stored is counted, the cumulative charging count of each electric vehicle can be calculated.

Battery SOH data is the battery SOH at the time of data transmission predicted by collecting measurement data such as charge voltage, charge current, charge capacity, and temperature by each charge cycle from the BMS of each electric vehicle. Currently, most electric vehicle BMS predict and store the battery SOH by charge cycle or at a certain period. The battery SOH of each electric vehicle reflects the driving habits (cumulative number of rapid accelerations, cumulative number of rapid brakings), charging patterns, cumulative driving distance, and vehicle age of each electric vehicle driver.

Another feature of the present invention is a step (104b) of mapping the accumulated number of charging cycles of each electric vehicle and the battery remaining value (SOH) of each electric vehicle stored in a dataset, and a step (105b) of generating a baseline prediction model of each electric vehicle by a regression model in which the accumulated number of charging cycles at the time of data transmission is an independent variable and the battery remaining value (SOH) at the time of data transmission is a dependent variable from the mapped accumulated number of charging cycles of each electric vehicle and the battery remaining value (SOH) of each electric vehicle. The baseline prediction model of each electric vehicle can be periodically updated using vehicle data collected in real time (106b). Accordingly, since the vehicle data reflected according to the prediction time is updated, the baseline prediction model of each electric vehicle may also change and the predicted future remaining value (SOH in the future) may also change.

Fig. 5 illustrates an example of a baseline prediction model for an individual electric vehicle generated in this manner. Using this baseline prediction model for an individual electric vehicle, the platform (20) can predict the battery remaining value (SOH in the future) at a future point according to the cumulative number of charges for an individual electric vehicle (108a). Or, it can predict at what cumulative number of charges in the future the remaining value (SOH in the future) reaches the battery replacement point. Typically, the point in time for battery replacement is considered to be when the SOH is 80%.

The above platform (20) can derive a function of the cumulative number of charging cycles and time of each electric vehicle from the accumulated vehicle data of each electric vehicle (108b). Using the function of the cumulative number of charging cycles and time of each electric vehicle derived in this way, the battery remaining value (SOH in the future) of each electric vehicle can be predicted over time (109b).

As shown in FIGS. 6 and 7, the battery remaining value (SOH in the future) of each electric vehicle for a future point in time can be transmitted to the electric vehicle user's smartphone app (27) and displayed on the smartphone app (27).

As shown in Fig. 6, the future battery residual value (SOH in the future) of each electric vehicle can be predicted over time by changing the prediction time. Fig. 6 shows the future battery residual value (SOH in the future) based on the prediction time of last month and the future battery residual value (SOH in the future) based on the prediction time of this month. In addition, it shows the expected battery performance (residual value) one year later and the expected battery performance (residual value) three years later.

As illustrated in Figure 7, the future point-in-time battery remaining value (SOH in the future) predicted by the baseline prediction model of each electric vehicle and the future point-in-time battery remaining value (SOH in the future) predicted by the baseline prediction model by vehicle type and battery type can be sent together to motivate drivers to improve their existing battery management methods or driving habits.

In addition to the cumulative number of charges as an independent variable of the above regression model, a baseline prediction model is obtained through multiple regression analysis using at least one of the following: cumulative driving distance, cumulative number of sudden accelerations, cumulative number of sudden brakings, and vehicle age, so that the battery remaining value (SOH in the future) can be predicted more accurately.

1: IOT terminal
3: External GPS
5: Built-in GPS
7: Wireless Modem
9: Microcontroller unit
11: CAN Transmit/Receive Module
13: OBD II Interface
15: Battery for terminal
17: Power controller for terminals
19: Terminal charging circuit
20 : Platform
21: Database (D/B)
23 : Server
25 : Smartphone
27: Smartphone App

Claims (4)

An IoT terminal that is connected to an electric vehicle's OBD-II terminal and CAN module and provides vehicle data including time information, data for calculating the cumulative number of charges at the time of data transmission, and battery remaining health value (SOH) data to a platform via the Internet, and a platform that is connected to the Internet and stores data received from the IoT,
In the above platform, a step for receiving and storing vehicle data including data for calculating the number of accumulated charges at the time of data transmission and data for battery remaining value (SOH) from various types of electric vehicles equipped with various types of batteries; a step for dividing the received vehicle data by vehicle type and battery type and configuring a dataset D/B corresponding to each; a step for calculating the number of accumulated charges from the data for calculating the number of accumulated charges of the dataset D/B configured by vehicle type and battery type; a step for mapping the number of accumulated charges calculated by vehicle type and battery type and the SOH of the battery classified by vehicle type and battery type; a step for generating a baseline prediction model by vehicle type and battery type by a regression model using the number of accumulated charges at the time of data transmission as an independent variable and the SOH of the battery at the time of data transmission as a dependent variable from the mapped number of accumulated charges and SOH of the vehicle type and battery type; A method for predicting the future residual value of an electric vehicle battery through analysis of connected car vehicle data, characterized by performing a step of predicting the future residual value (SOH in the future) of the battery according to the cumulative number of charging cycles using a baseline prediction model for each vehicle type and battery type. An IoT terminal that is connected to an electric vehicle's OBD-II terminal and CAN module and provides vehicle data including time information, data for calculating the cumulative number of charges at the time of data transmission, and battery remaining health value (SOH) data to a platform via the Internet, and a platform that is connected to the Internet and stores data received from the IoT,
In the above platform, a step for receiving and storing vehicle data including data for calculating the number of cumulative charges at the time of data transmission from an individual electric vehicle and data for battery remaining value (SOH); a step for configuring a dataset D/B for vehicle data of an individual electric vehicle; a step for calculating the number of cumulative charges of an individual electric vehicle from the data for calculating the number of cumulative charges of the dataset D/B of the individual electric vehicle; a step for mapping the calculated number of cumulative charges of an individual electric vehicle and the battery remaining value (SOH) of an individual electric vehicle stored in the dataset; a step for generating a baseline prediction model of an individual electric vehicle by a regression model in which the cumulative number of charges at the time of data transmission is an independent variable and the battery remaining value (SOH) at the time of data transmission is a dependent variable from the mapped number of cumulative charges and SOH of an individual electric vehicle; A method for predicting the future state of health (SOH) of an electric vehicle battery through analysis of connected car vehicle data, characterized in that it comprises: a step of predicting the future state of health (SOH) of a battery according to the cumulative number of charging cycles using a baseline prediction model of the individual electric vehicle;
In paragraph 1 or 2,
A method for predicting the future residual value of an electric vehicle battery, characterized in that the baseline prediction model for each vehicle type and battery type or the baseline prediction model for an individual electric vehicle is periodically updated using vehicle data collected in real time.
In any one of the clauses 1 to 3,
The above platform applies the cumulative number of charging of an individual electric vehicle to a baseline prediction model by vehicle type and battery type and obtains a function of the cumulative number of charging of an individual electric vehicle and time to predict a future point-in-time battery residual value (SOH in the future) of an individual electric vehicle over time, or applies the cumulative number of charging of an individual electric vehicle to a baseline prediction model for the individual electric vehicle and obtains a function of the cumulative number of charging of an individual electric vehicle and time to predict a future point-in-time battery residual value (SOH in the future) of an individual electric vehicle over time, and provides a result of predicting a future point-in-time battery residual value (SOH in the future) over time to a smartphone of an individual electric vehicle user. A method for predicting a future point-in-time battery residual value.

Images