Battery management method for kickboard, computer program for the same

Abstract

The present invention relates to a battery management method for a kickboard and a computer program for the same. According to an embodiment of the present invention, a battery management method for a kickboard and a computer program for the same are disclosed, configured to generate and provide to a user battery user data including the State of Charge (SOC) and driving range of the kickboard battery during driving by utilizing static element data including body information of the kickboard user, specifications of the kickboard battery, and weather information, and dynamic element data indicating the state of the kickboard battery during driving.

Description

Technology Field

[0001] The present invention relates to a battery management method for a kickboard and a computer program for the same, wherein the battery management method for a kickboard and the computer program for the same are configured to generate battery user data including the State of Charge (SOC) and driving range of the kickboard battery during driving by utilizing static element data including physical information of the kickboard user, specifications of the kickboard battery, and weather information, and dynamic element data indicating the state of the kickboard battery during driving, and provide such data to the user. Background Technology

[0003] Technologies for battery management of personal mobility devices such as electric scooters or electric bicycles have been proposed.

[0004] As an example of prior art, Korean published patent 10-2022-0097007 (July 7, 2022) relates to a system and method for providing a driving distance of an electric kickboard. It proposes a configuration that improves the accuracy of information regarding the driving distance by reducing errors caused by the slope of the driving path and errors caused by the user's weight, by comprising a driving information receiving step of receiving driving information and weight input information regarding a destination and a driving path before driving, and a providing step of providing an estimated current consumption and driving distance to the destination from the received information, and updating and providing the estimated current consumption and driving distance by checking the weight information of the user riding during driving.

[0005] However, the above-mentioned conventional technology has limitations in that it does not reflect the specifications of the kickboard battery and weather information, etc., because it obtains the estimated current consumption reflecting the slope of the driving path and the user's weight using a current consumption table.

[0006] As another example of prior art, Korean Published Patent 10-2022-0112434 (August 11, 2022) relates to a battery management and sharing system for personal mobility via a server, comprising: a battery station that stores and charges a battery having a battery identification code and generates battery status information according to charging and discharging, which is mounted on a personal mobility device; a user terminal having a user identification code that performs a rental procedure using a dedicated app to rent a battery charged at the battery station and return a used battery to the battery station; and a server that manages the battery status information and the rental procedure. The server proposes a configuration that manages the battery status information by receiving battery status information from the battery station or the user terminal based on the battery identification code and the user identification code.

[0007] However, the aforementioned conventional technology had a limitation in that it used a simple correction process that increased the driving distance and driving time as the user's weight decreased, the slope decreased, or the temperature approached the ambient temperature range.

[0008] As another example of prior art, Korean registered patent 10-2024765 (September 18, 2019) relates to a driving distance prediction device and proposes a configuration comprising: a measuring unit that measures the weight of a user mounted on the saddle of an electric bicycle; a calculating unit that calculates an estimated distance achievable by the motor using the remaining power of a battery that provides driving power to the motor moving the electric bicycle and the weight measured by the measuring unit; and a display unit installed on the handlebars of the electric bicycle and displaying the estimated distance.

[0009] However, the above-mentioned conventional technology has limitations in that it cannot reflect battery specifications and weather information, etc., because it is configured to calculate the estimated distance currently achievable using the remaining power of the battery based on a lookup table containing information on the distance achievable per unit of power according to body weight. Prior art literature

[0011] Republic of Korea Published Patent 10-2022-0097007 (July 7, 2022) Republic of Korea Published Patent 10-2022-0112434 (August 11, 2022) Republic of Korea Registered Patent 10-2024765 (September 18, 2019) The problem to be solved

[0012] The present invention has been devised in consideration of the above-mentioned problems, and aims to provide a battery management method for a kickboard and a computer program for the same, configured to generate and provide battery user data including the State of Charge (SOC) and driving range of the kickboard battery during driving by utilizing static element data including the kickboard user's body information, specifications of the kickboard battery, and weather information, and dynamic element data indicating the state of the kickboard battery during driving. means of solving the problem

[0014] According to one aspect of the present invention in light of the above objective, a battery management method for a kickboard executed on a battery management server linked via a network with a data input means, a data sensing means, and a data collection means comprises: 1) receiving static element data including body information of a kickboard user and an upper limit voltage, a lower limit voltage, and a capacity of a kickboard battery through the data input means; 2) receiving dynamic element data including current, voltage, and temperature of a kickboard battery and GPS coordinates of a kickboard according to a set cycle during driving of a kickboard through the data sensing means installed on the kickboard; and 3) receiving static element data including weather information based on the GPS coordinates of a kickboard through the data collection means. A battery management method for a kickboard is disclosed, comprising the step of: 4) when it is determined that the number of dynamic element data input according to a set cycle has accumulated to be N or more of a preset number, searching for battery information data satisfying a similarity condition in a battery information DB based on the static element data and the accumulated N dynamic element data, and generating battery user data including the State of Charge (SOC) and driving range of the kickboard battery based on the battery information data obtained as a search result, and providing it through a data output means.

[0015] According to another aspect of the present invention, a computer program stored in a medium is disclosed to be combined with hardware to execute a battery management method for the kickboard. Effects of the invention

[0017] The present invention has the advantage of being able to provide accurate battery user data by generating and providing battery user data during scooter riding using not only battery information such as the specifications of the scooter battery and the state of the scooter battery during scooter riding, but also the scooter user's physical information and weather information. Brief explanation of the drawing

[0019] FIG. 1 is an overall system configuration diagram in which a battery management method according to an embodiment of the present invention is executed. FIG. 2 is a configuration diagram of a battery management server according to an embodiment of the present invention, FIG. 3 is a schematic diagram from a hardware perspective of a battery management server according to an embodiment of the present invention, FIG. 4 is a configuration diagram of a kickboard and a kickboard terminal according to an embodiment of the present invention, FIGS. 5 to 8 are flowcharts of a battery management method according to an embodiment of the present invention, FIGS. 9 to 18 are schematic diagrams for explaining a battery management method according to an embodiment of the present invention. Specific details for implementing the invention

[0020] The present invention may be implemented in various other forms without departing from its technical concept or main features. Accordingly, the embodiments of the present invention are merely examples in all respects and should not be interpreted restrictively.

[0021] Terms such as "first," "second," etc., are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component.

[0022] When it is stated that a component is "connected" or "joined" to another component, it may be directly connected or joined to that other component, or there may be other components in between.

[0023] The singular expressions used in this application include the plural expressions unless the context clearly indicates otherwise. In this application, terms such as "comprising," "having," or "having" are intended to express the existence of the components or combinations thereof described in the specification, and do not preclude the possibility that other components or features may exist or be added.

[0025] Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the attached drawings.

[0026] FIG. 1 is an overall system configuration diagram in which a battery management method according to an embodiment of the present invention is executed, FIG. 2 is a configuration diagram of a battery management server according to an embodiment of the present invention, FIG. 3 is a schematic diagram from a hardware perspective of a battery management server according to an embodiment of the present invention, and FIG. 4 is a configuration diagram of a kickboard and a kickboard terminal according to an embodiment of the present invention.

[0027] A battery management method for a kickboard (10) according to one embodiment of the present invention is executed in a battery management server (100) that is linked through a network (1) with a data input means, a data sensing means, and a data collection means.

[0028] For example, the data input means may be a user terminal (200), the data input means may be an administrator terminal (300), the data sensing means may be a kickboard terminal (400), and the data collection means may be an API linkage module (104).

[0029] For example, the user terminal (200) and / or administrator terminal (300) may be a PC, smartphone, tablet, etc., equipped with a communication module and equipped with web browsing software or a battery management application that can connect to a battery management server (100). The user terminal (200) can be used by a user riding the kickboard (10) while driving.

[0030] For example, the data sensing means may be an embedded computer terminal installed in a kickboard terminal (400), equipped with a communication module, and capable of connecting to a battery management server (100) by running battery management software in the form of firmware.

[0031] The battery management server (100) uses static element data including the body information of the kickboard user, the specifications and weather information of the kickboard battery (20), and dynamic element data indicating the state of the kickboard battery (20) while the kickboard (10) is being driven, to generate battery user data including the State Of Charge (SOC) and driving distance of the kickboard battery (20) while the kickboard (10) is being driven, and provides this data to the user.

[0032] The battery management server (100) of the present embodiment includes, from a functional perspective, a user management module (101) that manages user registration and data transmission with a user terminal; a kickboard management module (102) that manages kickboard registration and data transmission with a kickboard terminal; a battery management module (103) that generates and provides battery user data during the operation of the kickboard (10) using static element data and dynamic element data; an API linkage module (104) that receives additional information by linking with an external server (e.g., a weather information provision server); and an operation module (105) that provides input / output processing functions for various information related to battery management and provides overall system management functions including an administrator mode. The battery management module (103) includes a data generator that searches for and returns data based on similarity condition judgment or generates and provides data using a trained deep learning model or machine learning model.

[0033] In addition, the battery management server (100) of the present embodiment includes a user DB (111) that stores and manages user registration information, a kickboard DB (112) that stores and manages kickboard registration information, and a battery information DB (113) that stores and manages static element data and dynamic element data and stores and manages battery user data generated using these data.

[0034] Referring to FIG. 3, from a hardware perspective, the battery management server (100) of the present embodiment includes a memory (1002) for storing one or more commands and a processor (1004) for executing the one or more commands stored in the memory (1002), and is a computing device for executing a computer program stored on a medium to execute a battery management method for a kickboard (10). The battery management server (100) of the present embodiment may further include a data input / output interface (1006), a communication interface (1008), a data display means (1003), and a data storage means (1005).

[0036] FIGS. 5 to 8 are flowcharts of a battery management method according to an embodiment of the present invention, and FIGS. 9 to 18 are schematic diagrams for explaining a battery management method according to an embodiment of the present invention.

[0037] More specifically, FIGS. 5 and 6 are overall flowcharts of the battery management method of the present embodiment, FIG. 7 is a flowchart of the process of determining similarity conditions of the battery management method of the present embodiment, and FIG. 8 is a flowchart of the processing process according to the determination of the number of dynamic element data of the battery management method of the present embodiment.

[0038] 1) In step 1) the battery management server (100) receives static element data including the body information of the kickboard user and the upper voltage, lower voltage, and capacity of the kickboard battery (20) through the data input means (see FIG. 9 and FIG. 10). The static element data are data whose values ​​do not change over a short period of time.

[0039] Preferably, the body information of the kickboard user includes at least one of the kickboard user's height and weight.

[0040] For example, the upper voltage, lower voltage, and capacity of the kickboard battery (20) can be referenced from the battery specifications provided by the battery manufacturer.

[0041] When riding the kickboard (10), the load on the kickboard (10) increases according to the user's body weight, and the air resistance increases or decreases according to the user's height.

[0042] The maximum and minimum values ​​of the voltage output as the kickboard battery (20) discharges change according to the upper and lower voltage limits of the kickboard battery (20). These factors are determined during the manufacturing of the kickboard battery (20), and since the specifications of the battery change depending on the manufacturer, they must be accepted as static factors. Likewise, the capacity is also a factor determined during the manufacturing of the kickboard battery (20).

[0043] For example, the data input means for inputting the physical information of the kickboard user may be a user terminal (200) capable of inputting data of the kickboard user. Additionally, the data input means for inputting the upper voltage, lower voltage, and capacity of the kickboard battery (20) may be a manager terminal (300) capable of inputting data of the kickboard manager. User login or manager login may be performed for data input through the data input means.

[0044] For example, the administrator, as the entity managing the scooter rental service, inputs information about the scooter into the battery management server (100) in advance. Here, information refers to information that can identify the scooter, along with the upper and lower voltage limits and capacity of the battery. For example, the user refers to a person who uses the scooter rental service and rides the scooter when the request to rent the scooter is sent to the battery management server (100) and approved. The user may input personal information, such as height and weight, in advance through a process such as membership registration.

[0045] 2) In step 2) the battery management server (100) receives dynamic element data including the current, voltage, temperature of the kickboard battery (20) and the GPS coordinates of the kickboard (10) according to a set cycle during the driving of the kickboard (10) through the data sensing means installed on the kickboard (10) (see FIG. 9). The dynamic element data are data whose values ​​change over a short period of time.

[0046] The above data sensing means is a kickboard terminal (400) installed on the kickboard (10), and obtains the current, voltage, and temperature of the kickboard battery (20) through a current sensor (410), a voltage sensor (420), and a temperature sensor (430) installed on the kickboard battery (20), respectively, and obtains the GPS coordinates of the kickboard (10) through a GPS sensor (440) installed on the kickboard (10).

[0047] In the case of dynamic element data, it represents the change in the state of the kickboard battery (20) while the kickboard (10) is driving, and changes steadily over a short period of time. In the case of current, the current value increases rapidly in a short period when a lot of power is used, such as when the kickboard (10) goes uphill or when it is moving at a fast speed. In the case of voltage, it has the property of steadily decreasing while the kickboard (10) is driving. Temperature also has the property of steadily increasing while the kickboard (10) is driving, similar to voltage. In addition, the distance traveled by the kickboard (10) can be measured by checking the location of the kickboard (10) through the GPS sensor (440).

[0048] In step 3), the battery management server (100) receives static element data including weather information based on the GPS coordinates of the kickboard (10) through the data collection means (see FIG. 9 and FIG. 10).

[0049] Preferably, the weather information includes at least one of temperature and humidity corresponding to the GPS coordinates of the kickboard (10).

[0050] When it is the first data of a ride, the location of the scooter is checked via GPS, and the temperature and humidity of the ride are received using an external service such as a weather API (e.g., Korea Meteorological Administration API).

[0051] The above data collection means may be an API integration module (104) that receives weather information from a weather information providing server (500). The API integration module (104) receives weather information by linking with the weather API of the weather information providing server (500).

[0052] Although the temperature and humidity of the driving environment change over time, they are factors that do not change significantly during the short period of driving the scooter (10), so the temperature and humidity can be said not to change during a single ride and are classified as static element data.

[0053] In step 4), the battery management server (100) determines whether the number of dynamic element data input according to the setting cycle has accumulated to a preset number of N or more (see FIG. 5 and FIG. 8).

[0054] When it is determined that the number of dynamic element data input according to the set cycle has accumulated to be more than N, a battery information DB (112) satisfying similarity conditions is searched based on the static element data and the accumulated N dynamic element data, and battery user data including the State of Charge (SOC) and driving range of the kickboard battery (20) is generated based on the battery information data obtained as a search result and provided through a user-side data output means. For example, the user-side data output means may be a user terminal (200).

[0055] FIG. 13 is an example of dynamic element data input according to a set period. The battery management server (100) obtains and uses N data from the driving records of the kickboard (10) collected according to a stored set period (e.g., 10 seconds). The dynamic element data is composed of the current, voltage, and temperature of the kickboard battery (20) and is used together with static element data received in advance from the user.

[0056] For example, if the set cycle is 10 seconds and N is set to 10, the N data points used to search for battery information data satisfying similarity conditions could be 10 data points corresponding to 0, 10, 20, ..., 90 seconds (when the current time point is 90 seconds), 10 data points corresponding to 10, 20, ..., 100 seconds (when the current time point is 100 seconds), or 10 data points corresponding to 20, ..., 110 (when the current time point is 110 seconds). In other words, N data points including the current time point are extracted and used from dynamic element data input during driving.

[0057] In order to generate battery user data including the State of Charge (SOC) and driving range of the kickboard battery (20), the battery information data obtained from the search result may include the hourly SOC, driving range per hour, discharge capacity, and the current time point in the corresponding discharge cycle.

[0058] For example, accumulated N dynamic element data can be understood as a group of 1st to Nth dynamic element data, with the most recently input dynamic element data being the Nth data. The 1st dynamic element data can be viewed as data measured at the first set cycle after driving.

[0059] Preferably, in step 4), the battery information data obtained from the search result is obtained based on time point data on the discharge cycle according to the setting period of the dynamic element data.

[0060] Figure 14 is an example of battery information data obtained from a search result.

[0061] Based on similarity from the input data, a discharge cycle is found from the battery information data of the battery information DB (112) and a discharge cycle is generated. When generating the discharge cycle, information regarding the current time is returned together so that the input data can be identified at what point in the discharge cycle based on the battery information data obtained as a search result.

[0062] In the case of FIG. 14, the time taken for one discharge cycle is 0 to 1800 seconds until the SOC of the kickboard battery (20) goes from 100% to 0%, and information regarding the current time point is returned regarding which point in the discharge cycle with a set period of 10 seconds corresponds to the input N data during the time from 0 to 1800 seconds. FIG. 14 exemplifies a case where the input time point of the Nth data corresponds to the current time point, and the current time point is 240 seconds during the discharge cycle with a set period of 10 seconds during the time from 0 to 1800 seconds.

[0063] Therefore, in the case of FIG. 14, battery user data including the State of Charge (SOC) and driving range of the scooter battery of the scooter (10) currently being driven is generated based on the data at the time point (current time point) when the discharge cycle data (current time point) in the battery information data of the battery information DB (112) is 240 seconds, and is provided through a user-side data output means.

[0064] If the kickboard (10) continues to drive and the time point data (current time point) on the discharge cycle reaches 250 seconds, battery user data including the State of Charge (SOC) and driving range of the kickboard battery of the kickboard (10) currently being driven is generated based on the time point when the time point data (current time point) on the discharge cycle (current time point) is 250 seconds in the battery information data of the battery information DB (112) and the data at that time point and subsequent times, and is updated and provided in real time through the user-side data output means.

[0065] Meanwhile, the battery information DB (112) is configured to include two or more battery information data obtained according to different driving conditions, and can be configured in advance before the execution of the battery management method of the present embodiment, and can be configured to further accumulate and increase the battery information data during the execution of the battery management method of the present embodiment.

[0066] For example, each battery information data includes static element data and dynamic element data received through steps 1) to 3) while driving the scooter (10) until the SOC of the scooter battery (20) becomes 0% from 100%, and is configured to include the SOC of the scooter battery (20) and the distance traveled calculated according to a set cycle during driving of the scooter (10) based on the received static element data and dynamic element data. Preferably, for one scooter battery (20), driving to obtain battery information data can be performed until the scooter battery (20) reaches EOL (End of Life), and charging and discharging of the scooter battery (20) is performed, and dynamic element data is collected according to a set cycle in each discharge cycle. In addition, such driving can be performed for one static element data and then for another static element data, thereby obtaining multiple dynamic element data for each of various static element data conditions.

[0067] FIG. 11 is an example of static element data and dynamic element data for configuring battery information data of the battery information DB (112).

[0068] In the case of Fig. 11, data is collected during a single scooter ride until the fully charged battery is completely discharged and the State of Charge (SOC) changes from 100% to 0%. At this time, static element data such as the scooter user's height and weight, the upper and lower voltage limits and capacity of the scooter battery, and the temperature and humidity are collected and recorded. Additionally, dynamic element data such as the scooter battery's voltage, current, temperature, and the scooter's GPS coordinates are collected by measuring them using a sensor at a set interval (e.g., 10 seconds).

[0069] After data collection is complete, the discharge cycle number is recorded for data processing, and the SOC for each time period is calculated using the measured current value.

[0070]

[0071] SOC is calculated using the formula above. Here, t represents the current time, and SOC0 represents the initial SOC. Also, I represents the discharge current, and T represents the time interval (set period) used for measurement.

[0072] Similar to SOC, the travel distance is also calculated by using GPS coordinate values ​​for each time period.

[0073] In addition, the State of Health (SOH) is calculated by determining the capacity of the corresponding discharge cycle through the total amount of current used. That is, the capacity of the corresponding discharge cycle is obtained by integrating the current amount with respect to time. And the SOH is It is calculated as follows. Measurement is terminated when the SOH reaches 70%, which is set as the EOL (End of Life). For example, if the initial battery capacity is 2 Ah and the capacity of the t-th discharge cycle is 1.4 Ah, then in that cycle, the SOH reaches 70% and the EOL is reached, so a replacement operation is performed on the battery and a record is made on the server that the battery has been replaced.

[0074] Finally, the number of discharge cycles, the capacity of the discharge cycles, the total driving distance, the SOC by time period, and the remaining driving distance are added to the collected data to create the data as shown in FIG. 12, which is used as battery information data for the battery information DB (112), and is also used as input data for training a deep learning model or a machine learning model. Here, the discharge cycle is counted as one cycle for the number of times the SOC is used from 100% to 0%.

[0075] When collecting data, ensure that other factors, excluding the above elements, can be controlled in an environment where data quality is maintained.

[0076] FIG. 11 illustrates a case in which, in static element data, the user's physical information (height, weight) is 160 cm and 55 kg, the upper limit voltage, lower limit voltage, and capacity of the kickboard battery (20) are 29.4 V to 25.9 V and 8 Ah, and the weather information (temperature, humidity) is -1.3℃ and 76.6%.

[0077] Additionally, FIG. 11 illustrates a case where time of 0 to 1800 seconds is taken for the SOC of the kickboard battery (20) to go from 100% to 0%, and in dynamic element data, during driving of the kickboard (10), the current of the kickboard battery (20) changes in the range of 2.05 to 1.95 A according to the set cycle (10 seconds), the voltage changes from 28.10 to 26.45 V, the temperature changes from 5.1 to 25.1 ℃, and the GPS coordinates (latitude, longitude) of the kickboard (10) change from 37.124, 128.21 to 37.126, 128.15.

[0078] Figure 12, generated based on Figure 11, illustrates a case where the distance traveled (km) is driven from 0 to 14.80 until the SOC of the kickboard battery (20) becomes 0% from 100%. Figure 12 illustrates a case where the discharge capacity is 7.8 Ah as battery information data in the first discharge cycle for the kickboard (10). For example, if static element data and dynamic element data are obtained in the second discharge cycle for the kickboard (10), the discharge cycle becomes 2.

[0079] The above battery information DB (112) is configured to include two or more battery information data obtained according to different driving conditions (static element data, dynamic element data).

[0080] As a variation, when a user rides the scooter (10) until the SOC of the scooter battery (20) goes from 100% to 0%, the data obtained from the ride can be stored in the battery information DB (112) as battery information data and used. Through this, the battery information DB (112) can further accumulate battery information data along with the users' use of the scooter.

[0081] For example, in step 4) above, the search for battery information data satisfying the similarity condition can be performed as follows.

[0082] 41) In step, the battery management server (100) searches the battery information DB (112) to determine whether there is battery information data that satisfies the similarity condition with the static element data.

[0083] In step 42), the battery management server (100) searches for battery information data that satisfies the similarity condition of step 41) and determines whether there is battery information data that satisfies the similarity condition with the accumulated N dynamic element data.

[0084] In step 43), the battery management server (100) obtains a search result in which battery information data satisfying the similarity condition of step 42) is battery information data satisfying the similarity condition of step 4).

[0085] Preferably, the determination of whether the similarity condition is satisfied in step 41) above can be made by vectorizing the static element data and battery information data by min-max normalization, respectively, and determining whether the cosine similarity between each vectorized vector exceeds a preset threshold.

[0086] In addition, the determination of whether the similarity condition is satisfied in step 42) above can be made by vectorizing the accumulated N dynamic element data and battery information data by min-max normalization, respectively, and determining whether the cosine similarity between each vectorized vector exceeds a preset threshold.

[0087] Cosine similarity is calculated by substituting the difference in angle between two vectors into the cosine, using the following formula.

[0088]

[0089] Therefore, to calculate cosine similarity, vectorization must be performed on each data point. First, the input static component data is vectorized to examine its similarity. To perform this vectorization, minimum and maximum values ​​are determined for each component, and Min-Max Normalization is conducted. Here, the minimum and maximum values ​​for each component are arbitrarily determined based on empirical results. Subsequently, the normalized static component data is combined into a single vector.

[0090] Referring to Fig. 15, Min-Max Normalization can be configured as follows.

[0091] For example, in the case of South Korea, since the lowest temperature never drops below -40 ℃, -40 ℃ is designated as the minimum value, and since the highest temperature has never exceeded 50 ℃, 50 ℃ is designated as the maximum value. And regarding the current temperature It is calculated using the formula. When the temperature is 24 ℃ as in the example in Fig. 15 Since this is the case, normalization is performed to 0.71. Similarly, min-max normalization is performed by pre-defining the maximum and minimum values ​​for each data point.

[0092] Similarity checking for dynamic element data is performed in the same manner as similarity checking for static element data. Finally, if the similarity check results show that both static and dynamic element data exceed a threshold, data is returned from the search results.

[0093] Meanwhile, in step 4) above, if it is determined that the number of dynamic element data input according to the set cycle has not accumulated to more than N, the SOC of the kickboard battery (20) estimated using OCV (Open Circuit Voltage) is provided as battery user data.

[0094] If the amount of data loaded during driving is less than N, battery information data satisfying the similarity condition is searched from the battery information DB to provide battery user data; since there is insufficient data, it is assumed to be a previous driving cycle, and the current SOC is estimated using the OCV method. If the amount of data loaded during driving is N or more, it is used as input to determine the similarity condition to obtain the discharge cycle and the current time point.

[0095] There are many known technologies for estimating and calculating the SOC of a battery using OCV (Open Circuit Voltage), and as such, they can be understood through, for example, Korean registered patent 10-1696313 (January 09, 2017), a detailed explanation is omitted.

[0096] Meanwhile, in step 4) above, if it is determined that the number of dynamic element data input according to the set cycle has accumulated to be more than N, and if battery information data satisfying the similarity condition is not found in the battery information DB (112) based on the static element data and the accumulated N dynamic element data, the process is performed as follows.

[0097] The above static element data and the accumulated N dynamic element data are input as input data to a trained deep learning model or machine learning model, and the SOC and driving range of the kickboard battery (20) are obtained as output data, and the output data is provided as battery user data.

[0098] Preferably, the trained deep learning model or machine learning model may be a model based on a Generative Adversarial Network (GAN) structure.

[0099] FIG. 16 shows the structure of a generative model based on a GAN structure. The deep learning model or machine learning model of the present embodiment can be configured as follows.

[0100] Generative Adversarial Networks (GANs) are artificial intelligence algorithms used in unsupervised learning. They are deep learning models composed of a generator that creates virtual data samples and a discriminator that determines whether an input data sample is real data, and are constructed through adversarial training between the generator and the discriminator. The generator aims to generate new data by extracting it from a GAN-based deep learning model after learning the probability distribution of real data.

[0101] In this embodiment, a discriminator determines whether the input data is actual data by using one cycle of driving data as input. Subsequently, using pre-collected driving data, only a portion of the data within that cycle is combined with static data. The combined data is used as input to a generator to create an entire cycle. The generated entire cycle is used as input to a discriminator to determine whether it is generated data or actual data. The generator and discriminator are trained by repeating this process of discrimination using actual data as input and discrimination using data generated by the generator as input. Once sufficient training is complete, data is generated using only the generator. By performing data generation when similarity is low—that is, when there is a large discrepancy with previously collected data—data can be generated and used, thereby returning data that is more suitable for the given input situation.

[0102] Meanwhile, in step 4) above, if it is determined that the capacity of the kickboard battery (20) has reached EOL (End of Life) based on the battery information data obtained from the search result, EOL arrival data can be generated and provided through the administrator's data output means.

[0103] In this case, the administrator-side data output means that generates and provides the EOL arrival data may be an administrator terminal (300).

[0104] As described above, the State of Health (SOH) can be calculated by determining the capacity of the corresponding discharge cycle through the total amount of current used. For example, if the End of Life (EOL) is set when the SOH reaches 70%, and the initial battery capacity is 2 Ah and the capacity of the t-th discharge cycle is 1.4 Ah, then it can be determined that the SOH reached 70% in that cycle and the EOL was reached.

[0106] FIG. 17 is an example of a web service that uses generated data. The web service provides data to the user based on data generated by the battery management server (100). The web screen displays the estimated battery level by time as a graph, shows the estimated driving distance along with the current battery level, and shows the accumulated driving distance to date.

[0107] FIG. 18 is an example of combining a map service with generated data. A service such as a map API is used to select a starting point and an ending point, and then the driving distance is received along with the driving route. After that, the battery management server (100) uses the generated data to calculate the estimated remaining battery level upon arrival when driving along the route and displays it to the user.

[0108] The SOC and driving range of the kickboard battery (20) can be provided by converting them into the remaining battery capacity, accumulated driving range, etc. through simple calculation.

[0110] Embodiments of the present invention include a program for performing operations implemented by various computers and a computer-readable recording medium recording the same. The computer-readable recording medium may include program instructions, data files, data structures, etc., either alone or in combination. The medium may be one specifically designed and configured for the present invention or one known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs, DVDs, and USB drives; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. Explanation of the symbols

[0112] 10: Kickboard 20: Scooter Battery 100: Battery Management Server 104: API Integration Module 200: User terminal 300: Administrator Terminal 400: Kickboard terminal 410: Current sensor 420: Voltage sensor 430: Temperature sensor 440: GPS sensor 500: Weather Server

Claims

Claim 1 A battery management method for a kickboard executed on a battery management server linked via a network with a data input means, a data sensing means, and a data collection means, comprising: 1) receiving static element data including body information of the kickboard user and an upper limit voltage, a lower limit voltage, and a capacity of the kickboard battery through the data input means; 2) receiving dynamic element data including the current, voltage, and temperature of the kickboard battery and the GPS coordinates of the kickboard according to a set cycle during riding of the kickboard through the data sensing means installed on the kickboard; and 3) receiving static element data including weather information based on the GPS coordinates of the kickboard through the data collection means. and 4) when it is determined that the number of dynamic element data input according to a set cycle has accumulated to be N or more as pre-set, search for battery information data satisfying a similarity condition in a battery information DB based on the static element data and the accumulated N dynamic element data, and generate battery user data including the State of Charge (SOC) and driving range of the kickboard battery based on the battery information data obtained as a search result, and provide it through a user-side data output means; wherein the battery information DB is pre-configured to include two or more battery information data obtained according to different driving conditions, and each battery information data includes static element data and dynamic element data received through steps 1) to 3) while driving the kickboard in one discharge cycle of the kickboard battery, and is configured to include the SOC of the kickboard battery and driving distance calculated according to a set cycle during driving of the kickboard based on the received static element data and dynamic element data. Claim 2 A battery management method for a kickboard according to claim 1, characterized in that, in step 4), the battery information data obtained from the search result is obtained based on time point data on the discharge cycle according to the setting period of the dynamic element data. Claim 3 A battery management method for a kickboard according to claim 1, characterized in that, in step 4), if it is determined that the number of dynamic element data input according to the set cycle has not accumulated to be more than a preset N, the SOC of the kickboard battery estimated using OCV (Open Circuit Voltage) is provided as battery user data. Claim 4 A battery management method for a kickboard according to claim 1, wherein in step 4), when it is determined that the number of dynamic element data input according to the set cycle has accumulated to be N or more as pre-set, and when battery information data satisfying a similarity condition is not searched in the battery information DB based on the static element data and the accumulated N dynamic element data, the static element data and the accumulated N dynamic element data are input as input data to a trained deep learning model or machine learning model, the SOC and driving range of the kickboard battery are obtained as output data, and the output data is provided as battery user data. Claim 5 A battery management method for a kickboard according to claim 4, characterized in that the trained deep learning model or machine learning model is a model based on a Generative Adversarial Network (GAN) structure. Claim 6 A battery management method for a kickboard according to claim 1, wherein in step 4), the search for battery information data satisfying the similarity condition comprises: 41) a step of searching a battery information DB to determine whether there is battery information data satisfying the similarity condition with the static element data; 42) a step of searching for battery information data satisfying the similarity condition of step 41) to determine whether there is battery information data satisfying the similarity condition with the accumulated N dynamic element data; and 43) a step of obtaining a search result in which the battery information data satisfying the similarity condition of step 42) is the battery information data satisfying the similarity condition of step 4). Claim 7 A battery management method for a kickboard according to claim 6, wherein the determination of whether the similarity condition is satisfied in step 41) is determined by vectorizing the static element data and battery information data by min-max normalization, respectively, and determining whether the cosine similarity between each vectorized vector exceeds a preset threshold, and the determination of whether the similarity condition is satisfied in step 42) is determined by vectorizing the accumulated N dynamic element data and battery information data by min-max normalization, respectively, and determining whether the cosine similarity between each vectorized vector exceeds a preset threshold. Claim 8 A battery management method for a kickboard according to claim 1, characterized in that, in step 4), if it is determined that the capacity of the kickboard battery has reached EOL (End of Life) based on battery information data obtained from the search result, EOL arrival data is generated and provided through a data output means on the administrator side. Claim 9 A battery management method for a kickboard according to claim 1, characterized in that the body information of the kickboard user includes at least one of the height and weight of the kickboard user, and the weather information includes at least one of the temperature and humidity corresponding to the GPS coordinates of the kickboard. Claim 10 A battery management method for a kickboard according to claim 1, wherein the data input means for inputting the body information of the kickboard user is a user terminal capable of inputting data of the kickboard user, the data input means for inputting the upper limit voltage, lower limit voltage, and capacity of the kickboard battery is a manager terminal capable of inputting data of the kickboard manager, the data sensing means is a kickboard terminal installed on the kickboard, wherein the current, voltage, and temperature of the kickboard battery are obtained through a current sensor, a voltage sensor, and a temperature sensor installed on the kickboard battery, respectively, and the GPS coordinates of the kickboard are obtained through a GPS sensor installed on the kickboard, and the data collection means is an API linkage module that receives weather information from a weather information providing server. Claim 11 A battery management method for a kickboard according to claim 8, characterized in that the administrator-side data output means for generating and providing the EOL arrival data is an administrator terminal. Claim 12 delete Claim 13 delete Claim 14 A computer program stored on a medium to execute a battery management method for a kickboard according to any one of claims 1 to 11 in combination with hardware.

Images