Method and system for evaluating life of battery through ai-based electric vehicle operation data analysis

Abstract

According to an embodiment of the present invention, provided is a system for evaluating the life of a battery, the system including: a data collection unit for acquiring information about route information, a charging / discharging pattern, a driving pattern, and a cell state from an electric vehicle; and a life evaluation unit for predicting the life of the battery mounted on the electric vehicle by performing machine learning on the basis of the acquired information.

Description

Battery life evaluation method and system using AI-based electric vehicle operation data analysis

[0001] The present invention relates to a method and system for evaluating battery life through AI-based electric vehicle operation data analysis.

[0002] Vehicles can be equipped with various devices that operate by consuming electricity, and batteries are provided to store the power needed to supply these devices. Recently, electric vehicles that use batteries to generate power are on the rise.

[0003] The vehicle battery is the most critical component among the various components that make up an electric vehicle. Battery capacity and voltage determine driving range and maximum motor output. Because electric vehicle batteries are not removed from the vehicle once installed, unless they significantly impact driving, their deterioration cannot be assessed. Therefore, technology to diagnose the lifespan and condition of electric vehicle batteries, which directly impact the performance, reliability, and safety of electric vehicles, is crucial.

[0004] In the case of conventional methods for diagnosing the life and condition of batteries, most of them are based on lithium-ion battery deterioration models and are estimated by applying some data.

[0005] The matters described as background technology above are only intended to enhance understanding of the background of the present invention, and should not be taken as an admission that they correspond to prior art already known to those skilled in the art.

[0006] The purpose of the present invention is to solve the problems of the above-described prior art.

[0007] The purpose of the present invention is to enable accurate prediction of battery life by analyzing 17 factors affecting battery life and applying differential weights according to the degree of influence of each factor.

[0008] The purpose of the present invention is to provide information on a lifespan improvement plan to a user terminal of a user of an electric vehicle when the lifespan of the electric vehicle is below the average in relation to the total travel distance.

[0009] The purposes of the present invention are not limited to those mentioned above, and other purposes not mentioned will be clearly understood from the description below.

[0010] According to one embodiment of the present invention for achieving the above-described purpose, the present invention may include a data collection unit that obtains information on route information, charge / discharge patterns, driving patterns, and cell status from an electric vehicle; and a life evaluation unit that performs machine learning based on the obtained information to predict the life of a battery installed in the electric vehicle.

[0011] The above route information may be a total travel distance of the electric vehicle during a first period, the charge / discharge pattern may be an average DOD during a second period, an average discharge C-RATE during the second period, an accumulated discharge amount during the second period, an accumulated drivetrain power required during the second period, an average drivetrain power required during the second period, and an average SOC, the driving pattern may be an average speed during the second period, an average change in speed per second during the second period, an average accelerator position, a number of rapid accelerations, a number of rapid decelerations, a number of rapid stops, and a heater operation time during the second period, and the cell status may be an average cell voltage difference per second, a Cell Max average temperature, and a Cell Min average temperature during the second period.

[0012] In the above machine learning, weights of 028, 01, 022, 013, 011, 023, 007, 007, 008, 010, 012, 020, 100, 031, 045, 030, and 003 may be applied to the total moving distance of the electric vehicle during the first period, the average DOD during the second period, the average discharge C-RATE during the second period, the accumulated discharge amount during the second period, the accumulated drivetrain power required during the second period, the average drivetrain power required during the second period, the average SOC, the average speed during the second period, the average speed change per second during the second period, the average accelerator position during the second period, the number of rapid accelerations, the number of rapid decelerations, the number of rapid stops, the heater operation time, the average cell voltage difference per second during the second period, the Cell Max average temperature, and the Cell Min average temperature, respectively.

[0013] For the above machine learning, the random forest algorithm can be used.

[0014] According to an embodiment of the present invention, by analyzing 17 factors affecting battery life and applying differential weights according to the degree of influence of each factor, it is possible to accurately predict battery life.

[0015] According to an embodiment of the present invention, when the lifespan of an electric vehicle is below the average in relation to the total travel distance, information on a lifespan improvement plan can be provided to a user terminal of a user related to the electric vehicle.

[0016] Figure 1 is a schematic diagram of a battery life evaluation system according to one embodiment of the present invention.

[0017] This is a drawing showing the castle.

[0018] FIG. 2 shows the results of analyzing whether there were significant differences or changes between vehicles or within the operating period for each factor recorded and managed during the operation of an electric bus in order to identify factors affecting battery life according to one embodiment of the present invention.

[0019] Figure 3 is a diagram showing the learning results of each machine learning algorithm.

[0020] The following detailed description of the present invention refers to the accompanying drawings, which illustrate specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention, while different from each other, are not necessarily mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be implemented in other embodiments without departing from the spirit and scope of the present invention. Furthermore, it should be understood that the positions or arrangements of individual components within each disclosed embodiment may be modified without departing from the spirit and scope of the present invention. Accordingly, the following detailed description is not intended to be limiting, and the scope of the present invention is defined only by the appended claims, along with the full scope of equivalents to which such claims are entitled, if properly described. Like reference numerals in the drawings designate the same or similar functionality throughout the several aspects.

[0021] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings so that a person having ordinary skill in the art to which the present invention pertains can easily practice the present invention.

[0022] FIG. 1 is a diagram showing the configuration of a battery life evaluation system according to one embodiment of the present invention.

[0023] Referring to FIG. 1, a battery life evaluation system (100) according to one embodiment may be configured to include a data collection unit (110), an analysis unit (120), a machine learning unit (130), a life evaluation unit (140), and an improvement plan provision unit (150).

[0024] The life assessment system (100) according to one embodiment may be implemented as a known terminal or server. The terminal may be implemented in any form as long as it has computing capabilities and can communicate with the outside world.

[0025] The data collection unit (110), analysis unit (120), machine learning unit (130), life evaluation unit (140), and improvement provision unit (150) may be program modules or hardware capable of communicating with external devices. These program modules or hardware may be included in the life evaluation system (100) or other devices capable of communicating therewith in the form of an operating system, application program modules, and other program modules, and may be physically stored on various known memory devices. Meanwhile, these program modules or hardware include, but are not limited to, routines, subroutines, programs, objects, components, data structures, etc. that perform specific tasks or execute specific abstract data types, which will be described later according to the present invention.

[0026] The data collection unit (110) collects data shown in the following table from a vehicle equipped with a battery that is the subject of a life evaluation, for example, an electric bus.

[0027] According to one embodiment of the present invention, hundreds of factors collectable from batteries or battery-powered electric vehicles were analyzed for their correlation with battery life. The analysis was performed using algorithms such as machine learning, and the results identified 17 factors most strongly correlated with battery life.

[0028] Each item marked as “factor” in Table 1 below is a factor collected from each electric bus, and a total of 17 factors described above are collected.

[0029] Additionally, Table 1 below provides information on the categories (route information, charge / discharge pattern, driving pattern, cell status, target) for each factor, and how data for that factor is collected (sum or average) for each period (day or week).

[0030] Category Daily aggregate Weekly aggregate Item Factor Route information SUMSUMDISTANCE Total travel distance (weekly) Charge / discharge pattern Average Average DOD Daily average DOD (Max - Min) Average Average C_RATE Daily average discharge C-RATE SUMSUMDCH_DOD Daily cumulative discharge amount SUMSUMDRIVE_POWER_SUM Cumulative power required Average Average DRIVE_POWER_P Average power required (0 or more) Average Average SOC Average SOC Driving pattern Average Average SPEED Daily average speed (speed 0 or more) Average Average DIFF_SPEED Daily average speed change in seconds Average Average ACC_POS Daily average accelerator position (pos 0 or more) SUMSUMACC Number of sudden accelerations SUMSUMDEC Number of sudden decelerations SUMSUMDEC_STOP Number of sudden stops SUMSUMEVCU Heater on Count (seconds) Cell status Average Average CELL_DIFF Daily average cell voltage difference in seconds Average Average CELL_MXTMPCell Max Average temperature Average Average CELL_MNTMPCell Min Average temperature

[0031] In the table above, DOD (Depth of Discharge) is an indicator of the battery's discharge status, indicating the current percentage of discharge based on a fully charged battery. Additionally, C-RATE (Current Rate) is a value indicating the speed at which the battery is charged or discharged. The accelerator position is related to the driver's will to drive, with 0% indicating the ignition on and no accelerator pedal pressed, and 100% indicating full accelerator.

[0032] Sudden acceleration refers to driving at a rate of 5 to 8 km / h (5 km / h for trucks, 6 km / h for buses, 8 km / h for taxis) from a speed of 6.0 km / h or more. Sudden deceleration refers to driving at a rate of 8 to 14 km / h (8 km / h for trucks, 9 km / h for buses, 14 km / h for taxis) or more while the speed is 6.0 km / h or more. In addition, sudden stop refers to driving at a rate of 8 to 14 km / h (8 km / h for trucks, 9 km / h for buses, 14 km / h for taxis) or more while the speed is 5.0 km / h or less.

[0033] Heater on Count is defined as the amount of time the vehicle's heater is running. In one embodiment, Heater on Count is defined as the accumulated heater running time over a day or week, and the unit may be "seconds".

[0034] The Cell Max average temperature is the average of the maximum temperature of each of the multiple cells present in the battery, and the Cell Min average temperature is the average of the minimum temperature of each of the multiple cells present in the battery.

[0035] The analysis unit (120) performs a function of analyzing whether each of the above 17 factors has some degree of correlation with the battery life, i.e., SOH.

[0036] For example, FIG. 2 shows the results of analyzing whether there were significant differences or changes between vehicles or within the operating period for each factor recorded and managed during the operation of an electric bus to identify factors affecting battery life according to one embodiment of the present invention.

[0037] Referring to (a) of Fig. 2, it can be confirmed that the cell voltage difference is stabilized after the battery module is replaced, and referring to (b) of Fig. 2, it can be seen that the brake discharge current and the charging current intersect abruptly during driving, and that when the brake is lightly pressed, more charging current is generated.

[0038] Additionally, referring to (c) of FIG. 2, it can be seen that braking during driving causes rapid alternation of charging and discharging, thereby increasing the cell voltage difference. In other words, it can be seen that the cell voltage difference is an influencing factor on the battery life.

[0039] Referring to (d) of Fig. 2, it can be seen that approximately 11 kW of power is consumed when the heater is in operation, and that whether the heater is in operation is an influencing factor on the battery life.

[0040] The analysis unit (120) analyzes the degree of correlation between the extracted factors and battery life through the above analysis.

[0041] According to one embodiment, the machine learning unit (130) performs learning on the relationship between 17 factors and battery life based on the results analyzed by the analysis unit (120), and thereby analyzes the degree to which each of the 17 factors affects the battery life, i.e., the degree of influence.

[0042] The applicant performed the above learning based on the Random Forest, XGB, and Light GBM algorithms among machine learning algorithms, and the results are as shown in Fig. 3.

[0043] Figures 3 (a) to (c) show the results of learning based on the Random Forest, XGB, and Light GBM algorithms, respectively. In each graph, the horizontal axis represents the actual value, and the vertical axis represents the predicted value.

[0044] As can be seen from Figure 3, the random forest algorithm produces predicted values ​​with the smallest deviation from the actual values.

[0045] Therefore, in one embodiment, the influence of each factor was analyzed by learning the relationship between 17 factors and battery life based on the random forest algorithm, and the results are as follows.

[0046] Classification Factor Influence Direction Influence Detailed Influence Route Information Total travel distance (weekly) - Acceleration of SOH degradation when travel distance increases after 1400km per week 0.28 Cell status Daily average cell voltage difference in seconds - SOH degradation accelerates as cell voltage difference increases 0.45 Cell Max average temperature section SOH degradation is less in the section where Cell Max temperature is 25 to 30 degrees 0.30 SOH degradation is less in the section where Cell Min average temperature + Cell Min temperature is 15 to 20 degrees 0.03 Driving pattern Number of sudden stops - SOH degradation accelerates when the number of sudden stops increases 1.00 Heater on Count (in seconds) - SOH degradation accelerates when heater usage time increases 0.31 Number of sudden decelerations - SOH degradation accelerates when the number of sudden decelerations increases 0.20 Number of sudden accelerations - SOH degradation accelerates in the section from 0 to 500 times 0.12 Daily average accelerator position - 15 to 20, 35 to 40 SOH degradation acceleration in the section 0.10 Daily average speed change per second - SOH degradation acceleration when the number of rapid accelerations increases 0.08 Daily average speed (speed 0 or more) + 0.07 Average power consumption of the drivetrain before charging and discharging pattern (0 or more) Section 30~40KWh SOH degradation is small in the section 0.23 Daily average discharge C-RATE - SOH degradation acceleration when the C-rate increases 0.22 Cumulative discharge amount - SOH degradation acceleration when the cumulative discharge amount increases 0.13 Cumulative power consumption - SOH degradation acceleration when the cumulative power consumption increases 0.11 Daily average DOD (Max - Min) - SOH degradation acceleration when the daily average DOD is deeply used 0.10 Average SOC + 0.07

[0047] The life evaluation unit (140) applies the influence of each of the 17 factors obtained above as a weight. That is, after obtaining the 17 factors from a specific electric vehicle, for example, an electric bus, and applying the weights to each factor to perform machine learning, the life of the battery installed in the electric vehicle, i.e., the SOH, can be predicted. The improvement plan provision unit (150) performs the function of providing an improvement plan when the evaluated life compared to the total travel distance is below the average. Among the 17 factors described above, the charge / discharge pattern and the driving pattern are factors that can have a positive effect on the battery life by improving the pattern.

[0048] Accordingly, information on the need for improvement can be provided regarding factors causing SOH degradation acceleration in Table 2 above. The information can be provided to a user terminal (not shown) capable of communicating with the system (100) according to one embodiment.

[0049] According to one embodiment, by analyzing 17 factors affecting battery life and applying differential weights according to the degree of influence of each, a prediction of battery life can be made accurately.

[0050] According to one embodiment, when the lifespan of an electric vehicle is below average compared to the total distance traveled, information on a lifespan improvement plan may be provided to a user terminal of a user related to the electric vehicle.

[0051] The foregoing description of the present invention is for illustrative purposes only, and those skilled in the art will readily appreciate that the present invention can be readily modified into other specific forms without altering the technical spirit or essential characteristics of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single entity may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined manner.

[0052] The scope of the present invention is indicated by the claims described below, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention.

Claims

1. A data collection unit that obtains information on route information, charging / discharging patterns, driving patterns, and cell status from electric vehicles; A battery life evaluation system, comprising a life evaluation unit that performs machine learning based on acquired information to predict the life of a battery installed in the electric vehicle.

2. In paragraph 1, The above route information is the total travel distance of the electric vehicle during the first period, The above charge / discharge pattern is the average DOD during the second period, the average discharge C-RATE during the second period, the accumulated discharge amount during the second period, the accumulated drivetrain power required during the second period, the average drivetrain power required during the second period, and the average SOC. The above driving pattern is the average speed during the second period, the average change in speed per second during the second period, the average accelerator position during the second period, the number of sudden accelerations, the number of sudden decelerations, the number of sudden stops, and the heater operation time. The above cell status is a battery life evaluation system, which is an average cell voltage difference per second during a second period, an average Cell Max temperature, and an average Cell Min temperature.

3. In paragraph 2, In the above machine learning, A battery life evaluation system, wherein the Cell Min average temperature is applied with weights of 028, 01, 022, 013, 011, 023, 007, 007, 008, 010, 012, 020, 100, 031, 045, 030, and 003 respectively to the total moving distance of the electric vehicle during the first period, the average DOD during the second period, the average discharge C-RATE during the second period, the accumulated discharge amount during the second period, the accumulated drivetrain power required during the second period, the average drivetrain power required during the second period, the average SOC, the average speed during the second period, the average second-by-second speed change during the second period, the average accelerator position during the second period, the number of rapid accelerations, the number of rapid decelerations, the number of sudden stops, the heater operation time, the average second-by-second cell voltage difference during the second period, the Cell Max average temperature, and the Cell Min average temperature.

4. In paragraph 1, In the above machine learning, battery life evaluation using the random forest algorithm System.

Images