A battery pack discharge SOC calibration method

Abstract

The present invention belongs to the technical field of battery calibration. The present invention discloses a battery pack discharge SOC calibration method. The method comprises the following steps: extracting corresponding discharge characteristic information under different discharge state information based on a historical database; receiving a discharge SOC value and discharge characteristic information affecting a change in the discharge SOC value, establishing a machine learning model between the discharge SOC value and the discharge characteristic information, and extracting a calibration influence factor between the discharge SOC value and the discharge characteristic information based on the machine learning model; calculating a discharge stability coefficient based on the discharge characteristic information and the calibration influence factor, and matching a corresponding discharge calibration mode based on the discharge stability coefficient; reversely analyzing the calibration influence factor based on the discharge calibration mode and updating the calibration influence factor based on the discharge calibration mode; and calibrating the discharge SOC value based on the updated calibration influence factor.

Description

Technical Field

[0001] The present invention relates to the technical field of battery calibration, and more particularly to a battery pack discharge SOC calibration method. Background Art

[0002] Battery SOC refers to the current storage capacity of the battery and is one of the important parameters of battery performance. Since it cannot be directly and accurately measured, it can only be estimated through various estimation methods. In current practical applications, considering the impact of cost and computing power, the most commonly used method is to combine the ampere-hour integration method and the open circuit voltage method to estimate SOC. However, when the battery pack operates in the middle SOC range for a long time, since it cannot trigger specific conditions for calibration (such as full charge or empty), it is necessary to perform full charge and empty operations regularly to reset the SOC reference point, otherwise the error will gradually increase.

[0003] However, in actual applications, if the SOC cannot be automatically calibrated, most users will reset the SOC reference point to solve the battery pack failure problem only after the battery pack fails. The existing technology uses an incremental learning algorithm to perform online calibration of the SOC estimation. By continuously collecting battery operation data, the SOC estimation model is fine-tuned to reduce the error. It is also necessary to evaluate interference factors such as the type, purpose and working environment of the battery pack to reduce the SOC calibration deviation caused by external conditions.

[0004] In view of this, the present invention provides a battery pack discharge SOC calibration method. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a battery pack discharge SOC calibration method, which can comprehensively evaluate the impact of external conditions on SOC calibration and predict battery pack failures in advance, thereby timely maintaining and enhancing the service life of the battery pack.

[0006] To achieve the above object, the present invention provides the following technical solution: a battery pack discharge SOC calibration method, comprising the following steps:

[0007] Extracting corresponding discharge characteristic information under different discharge state information based on the historical database;

[0008] Receiving a discharge SOC value and discharge characteristic information that affects a change in the discharge SOC value, establishing a machine learning model between the discharge SOC value and the discharge characteristic information, and extracting a calibration influencing factor between the discharge SOC value and the discharge characteristic information based on the machine learning model;

[0009] A discharge stability coefficient is calculated based on the discharge characteristic information and the calibration influence factor, and a corresponding discharge calibration mode is matched based on the discharge stability coefficient; the calibration influence factor is reversely analyzed based on the discharge calibration mode and the calibration influence factor is updated based on the discharge calibration mode;

[0010] The discharge SOC value is calibrated based on the updated calibration influence factor.

[0011] In a preferred embodiment, the logic for obtaining the discharge characteristic information is:

[0012] The historical database includes several groups of historical discharge characteristic information and discharge state information collected in an experimental environment; the experimental environment is a data collection environment in which a tester controls the change of the discharge state information corresponding to the battery pack to collect various discharge characteristic information values;

[0013] The discharge characteristic information includes the discharge rate per unit time, the discharge depth, the battery pack temperature change value, the battery pack rated power and the discharge function requirements of the battery pack;

[0014] The discharge state information includes a rapid discharge state, a basic discharge state and an abnormal discharge state;

[0015] In the experimental environment, by controlling battery packs with different rated power and discharge function requirements, the discharge characteristic information of the battery packs under fast discharge state, basic discharge state and abnormal discharge state is collected.

[0016] In a preferred embodiment, the training logic of the machine learning model is:

[0017] Each set of discharge characteristic information is used to generate a corresponding discharge SOC actual value according to the discharge state information when collected, and each set of discharge characteristic information is used as the input of the machine learning model. The machine learning model outputs the discharge SOC calibration value of each set of discharge characteristic information, takes the discharge SOC actual value as the prediction target, and takes minimizing the sum of the prediction accuracy of all discharge characteristic information as the training target; wherein the calculation formula for the prediction accuracy is: ,in, is the number of the discharge characteristic information, For prediction accuracy, For the The discharge SOC calibration value corresponding to the group discharge characteristic information, For the The actual discharge SOC value corresponding to the discharge characteristic information of the group is obtained; the machine learning model is trained until the sum of the prediction accuracies reaches convergence and the training is stopped.

[0018] In a preferred embodiment, the discharge state of the battery pack is recorded as , the operating unit time in the discharge state is marked as , then the discharge stability coefficient corresponding to the current battery pack is:

[0019] ;

[0020] For the In this state Discharge stability coefficient per unit time; For the In this state The discharge stability coefficient of the previous unit time of the unit time; ; Used to represent the current operating unit time of the battery pack; , and are all positive integers; For the In this state The discharge rate per unit time, For the In this state The change in discharge rate per unit time; For the In this state The depth of discharge per unit time, For the In this state The change in discharge depth per unit time; For the In this state The real-time temperature value of the battery pack per unit time, For the In this state The battery pack temperature change value per unit time; , and are the calibration influencing factors corresponding to the discharge characteristic information respectively.

[0021] In a preferred embodiment, the discharge stability coefficient determines the corresponding discharge calibration mode according to a plurality of preset value ranges, and the value range includes , and ,

[0022] when When the discharge calibration mode is marked as the first discharge mode, the discharge SOC change value is identified based on the first discharge mode. If the discharge SOC change value is within the preset first calibration interval, whether the battery SOC value at the current moment is within the preset second calibration interval is determined; if the battery SOC value is within the preset second calibration interval, the discharge SOC value is calibrated based on the calibration influence factor corresponding to the current discharge stability coefficient;

[0023] when When the discharge calibration mode is marked as the second discharge mode, the calibration influence factor is reconfigured based on the second discharge mode, and the calibration influence factor includes a fixed weight factor and an adjustment weight factor; the discharge SOC value is calibrated based on the adjustment weight factor;

[0024] when When the discharge calibration mode is marked as the third discharge mode, the third discharge mode is a safe discharge mode; the calibration influence factor is readjusted based on the third discharge mode, and the adjusted calibration influence factor is updated as the calibration influence factor under the discharge state at the next moment.

[0025] In a preferred embodiment, the analysis logic of the calibration influencing factor includes:

[0026] Integrate different discharge characteristic information into a discharge characteristic matrix through normalization processing; the discharge characteristic matrix uniformly encodes and synchronously transmits the discharge characteristic information in a DC code format;

[0027] Perform correlation analysis on calibration influence factors of each discharge characteristic information and the discharge stability coefficient, wherein the calibration influence factors include fixed weight factors and adjustment weight factors;

[0028] A bipartite graph is constructed for each discharge feature information and the discharge stability coefficient, and the maximum weight between the discharge feature information and the discharge stability coefficient is marked as a fixed weight factor;

[0029] The correlation coefficient between the determined fixed weight factor and the discharge stability coefficient is marked as an adjustment weight factor, and the linear correlation between the adjustment weight factor and the fixed weight factor has a value range between -1 and +1;

[0030] In a preferred embodiment, the updating logic of the calibration influence factor is:

[0031] The discharge calibration mode is maintained in the third discharge mode, and the current battery pack discharge characteristic information and the corresponding calibration influencing factor are extracted;

[0032] Based on the current calibration influencing factor, the battery pack is controlled to output a predetermined discharge voltage; and the measured values ​​of the discharge voltage and the discharge SOC value of the battery pack under the control of any two different predetermined discharge voltages are respectively obtained;

[0033] The voltage difference between the two measured discharge voltages and the predetermined discharge voltage is calculated; and the calibration influence factor is readjusted according to the ratio of the difference between the two voltage differences and the difference between two different output discharge SOC values.

[0034] According to another aspect of the present invention, a battery pack discharge SOC calibration system is provided, which is based on the implementation of a battery pack discharge SOC calibration method according to one aspect of the present invention, and includes a discharge characteristic determination unit, a discharge SOC analysis unit, a discharge stability analysis unit and a calibration control unit;

[0035] A discharge characteristic determination unit, which extracts corresponding discharge characteristic information under different discharge state information based on a historical database;

[0036] A discharge SOC analysis unit receives a discharge SOC value and discharge characteristic information that affects a change in the discharge SOC value, establishes a machine learning model between the discharge SOC value and the discharge characteristic information, and extracts a calibration influencing factor between the discharge SOC value and the discharge characteristic information based on the machine learning model;

[0037] A discharge stability analysis unit calculates a discharge stability coefficient based on the discharge characteristic information and the calibration influence factor, matches a corresponding discharge calibration mode based on the discharge stability coefficient; reversely analyzes the calibration influence factor based on the discharge calibration mode and updates the calibration influence factor based on the discharge calibration mode;

[0038] The calibration control unit calibrates the discharge SOC value based on the updated calibration influence factor.

[0039] According to another aspect of the present invention, there is provided an electronic device, comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor;

[0040] The processor executes a battery pack discharge SOC calibration method according to one aspect of the present invention by calling a computer program stored in the memory.

[0041] According to another aspect of the present invention, there is provided a computer program product stored on a computer-readable medium, comprising a computer-readable program, which, when executed on an electronic device, provides a user input interface to implement a battery pack discharge SOC calibration method described in one aspect of the present invention.

[0042] Technical effects and advantages of a battery pack discharge SOC calibration method of the present invention:

[0043] The present invention extracts corresponding discharge characteristic information under different discharge state information based on a historical database, and establishes a machine learning model between the discharge SOC value and the discharge characteristic information, and then extracts the calibration influence factor, which can be applicable to various discharge scenarios and enhances the adaptability of the calibration method; and finally uses the updated calibration influence factor to calibrate the discharge SOC value, which helps to improve the accuracy of the discharge SOC value calibration, makes the calibration result closer to the actual discharge state, significantly improves the accuracy, adaptability and efficiency of the calibration, and helps to extend the battery life and improve the system reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A schematic diagram of a new energy battery pack fault detection system of the present invention;

[0045] Figure 2 This is a flow chart of the new energy battery pack fault detection method of the present invention;

[0046] Figure 3 It is a schematic diagram of the power monitoring circuit system of the present invention; DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] Example 1

[0049] See also Figure 1 As shown, a battery pack discharge SOC calibration system described in this embodiment includes a discharge characteristic determination unit 1, a discharge SOC analysis unit 2, a discharge stability analysis unit 3 and a calibration control unit 4, and the above-mentioned units are connected by wired and / or wireless connection to realize data transmission between the units;

[0050] A discharge characteristic determination unit 1 extracts corresponding discharge characteristic information under different discharge state information based on a historical database;

[0051] It should be noted that the logic for obtaining the discharge characteristic information is as follows:

[0052] The historical database includes several groups of historical discharge characteristic information and discharge state information collected in an experimental environment; the experimental environment is a data collection environment in which a tester controls the change of the discharge state information corresponding to the battery pack to collect various discharge characteristic information values;

[0053] The discharge characteristic information includes the discharge rate per unit time, the discharge depth, the battery pack temperature change value, the battery pack rated power and the discharge function requirements of the battery pack;

[0054] Specifically, the discharge rate, discharge depth and battery pack temperature change values ​​are tested through a series of testing equipment, such as a battery discharge tester used to measure the discharge performance of the battery, simulate different discharge conditions, and record the current, voltage, time and other parameters during the discharge process, thereby calculating the discharge rate and discharge depth. The temperature change of the battery pack during the discharge process is recorded by a temperature monitor.

[0055] The discharge rate indicates how fast the battery discharges. A high discharge rate may cause the battery to heat up and lose capacity faster, while a low discharge rate may extend the battery's service life.

[0056] The depth of discharge indicates the proportion of the amount of electricity that has been discharged from the battery during use to its rated capacity. Shallow discharge usually does not cause much damage to the battery, while deep discharge may accelerate battery aging and capacity decay.

[0057] The battery pack temperature change value standard refers to the temperature change of the battery during operation. Too high or too low temperature may cause battery capacity loss, increase in internal resistance or even battery damage.

[0058] The discharge state information includes a rapid discharge state, a basic discharge state and an abnormal discharge state;

[0059] In the experimental environment, by controlling battery packs with different rated power and discharge function requirements, the discharge characteristic information of the battery packs under fast discharge state, basic discharge state and abnormal discharge state is collected.

[0060] A discharge SOC analysis unit 2 receives a discharge SOC value and discharge characteristic information that affects a change in the discharge SOC value, establishes a machine learning model between the discharge SOC value and the discharge characteristic information, and extracts a calibration influencing factor between the discharge SOC value and the discharge characteristic information based on the machine learning model;

[0061] It should be noted that: the calibration influence factor and discharge characteristic information are substituted into the machine learning model to calculate the discharge SOC calibration value; the original SOC value is corrected or adjusted by calibrating the calibration influence factor to make it closer to the actual discharge SOC value. The discharge SOC calibration value is continuously verified by the actual discharge SOC value to ensure its accuracy and reliability;

[0062] The training logic of the machine learning model is:

[0063] Each set of discharge characteristic information is used to generate a corresponding discharge SOC actual value according to the discharge state information when collected, and each set of discharge characteristic information is used as the input of the machine learning model. The machine learning model outputs the discharge SOC calibration value of each set of discharge characteristic information, takes the discharge SOC actual value as the prediction target, and takes minimizing the sum of the prediction accuracy of all discharge characteristic information as the training target; wherein the calculation formula for the prediction accuracy is: ,in, is the number of the discharge characteristic information, For prediction accuracy, For the The discharge SOC calibration value corresponding to the group discharge characteristic information, For the The actual discharge SOC value corresponding to the discharge characteristic information of the group is obtained; the machine learning model is trained until the sum of the prediction accuracies reaches convergence and the training is stopped.

[0064] Wherein, the machine learning model is any one of a deep neural network model or a deep belief network model.

[0065] The discharge stability analysis unit 3 calculates a discharge stability coefficient based on the discharge characteristic information and the calibration influence factor, matches a corresponding discharge calibration mode based on the discharge stability coefficient; reversely analyzes the calibration influence factor based on the discharge calibration mode and updates the calibration influence factor based on the discharge calibration mode;

[0066] The discharge state of the battery pack is recorded as , the operating unit time in the discharge state is marked as , then the discharge stability coefficient corresponding to the current battery pack is:

[0067] ;

[0068] For the In this state Discharge stability coefficient per unit time; For the In this state The discharge stability coefficient of the previous unit time of the unit time; ; Used to represent the current operating unit time of the battery pack; , and are all positive integers; For the In this state The discharge rate per unit time, For the In this state The change in discharge rate per unit time; For the In this state The depth of discharge per unit time, For the In this state The change in discharge depth per unit time; For the In this state The real-time temperature value of the battery pack per unit time, For the In this state The battery pack temperature change value per unit time; , and are respectively calibration influencing factors corresponding to the discharge characteristic information; the discharge stability coefficient determines the corresponding discharge calibration mode according to a plurality of pre-set value ranges.

[0069] It should be noted that: the higher the discharge stability coefficient corresponding to the current battery pack, the more stable the discharge of the current battery pack, the higher the possibility that the detected data is within the normal range, and the lower the possibility of abnormality. Similarly, the lower the discharge stability coefficient corresponding to the current battery pack, the more unstable the discharge of the current battery pack, the lower the possibility that the detected data is within the normal range, and the higher the possibility of abnormality.

[0070] The calibration control unit calibrates the discharge SOC value based on the updated calibration influence factor.

[0071] Specifically, the value range corresponding to the discharge stability coefficient includes , and ,

[0072] when When the discharge calibration mode is marked as the first discharge mode, the discharge SOC change value is identified based on the first discharge mode. If the discharge SOC change value is within the preset first calibration interval, whether the battery SOC value at the current moment is within the preset second calibration interval is determined; if the battery SOC value is within the preset second calibration interval, the discharge SOC value is calibrated based on the calibration influence factor corresponding to the current discharge stability coefficient;

[0073] It should be noted that: in the first discharge mode, the discharge characteristic information of the battery pack itself is not stable, and the influence of the calibration influence factor on the discharge characteristic information is not obvious. The calibration influence factor can only ensure that when the battery pack is working normally, the discharge SOC value is within the predicted range. Therefore, in the first discharge mode, the discharge SOC value is calibrated by the calibration influence factor;

[0074] when When the discharge calibration mode is marked as the second discharge mode, the calibration influence factor is reconfigured based on the second discharge mode, and the calibration influence factor includes a fixed weight factor and an adjustment weight factor; the discharge SOC value is calibrated based on the adjustment weight factor;

[0075] It should be noted that: in the second discharge mode, the current discharge characteristic information of the battery pack itself is relatively stable, ensuring that when the battery pack is working normally, the discharge SOC value is within the predicted range. Therefore, in the first discharge mode, the discharge SOC value is calibrated by the calibration influence factor. The calibration influence factor can provide real-time feedback on the current operating mode of the battery pack, so as to more accurately obtain the discharge SOC value, and the calibration of the discharge SOC value is affected by repeatedly adjusting the calibration influence factor.

[0076] It should also be noted that the main factors affecting the change of the battery pack discharge SOC value are the discharge rate, discharge depth, and battery pack temperature change value; however, in actual applications, there is other discharge characteristic information. With the development of science and technology and the enhancement of data analysis capabilities, other discharge characteristic information that affects the change of the battery pack discharge SOC value can be included in the statistical scope in the later stage, and the influencing factors can be adapted and calibrated according to the corresponding discharge characteristic information.

[0077] The analysis logic of the calibration influencing factor includes:

[0078] Integrate different discharge characteristic information into a discharge characteristic matrix through normalization processing; the discharge characteristic matrix uniformly encodes and synchronously transmits the discharge characteristic information in a DC code format;

[0079] Perform correlation analysis on calibration influence factors of each discharge characteristic information and the discharge stability coefficient, wherein the calibration influence factors include fixed weight factors and adjustment weight factors;

[0080] A bipartite graph is constructed for each discharge feature information and the discharge stability coefficient, and the maximum weight between the discharge feature information and the discharge stability coefficient is marked as a fixed weight factor;

[0081] The correlation coefficient between the determined fixed weight factor and the discharge stability coefficient is marked as an adjustment weight factor, and the linear correlation between the adjustment weight factor and the fixed weight factor has a value range between -1 and +1.

[0082] It should be noted that as long as the historical database contains enough samples, the accuracy of the fixed weight factor can be more ensured, and the discharge characteristic information of the battery pack discharge performance under different discharge states can be predicted; in other words, if the accuracy of the fixed weight factor reaches a certain level, the adjusted weight factor can approach 0, and the dependence on the adjusted weight factor can be gradually reduced.

[0083] when When the discharge calibration mode is marked as the third discharge mode, the third discharge mode is a safe discharge mode; the calibration influence factor is readjusted based on the third discharge mode, and the adjusted calibration influence factor is updated as the calibration influence factor under the discharge state at the next moment.

[0084] The updating logic of the calibration influence factor is:

[0085] The discharge calibration mode is maintained in the third discharge mode, and the current battery pack discharge characteristic information and the corresponding calibration influencing factor are extracted;

[0086] Based on the current calibration influencing factor, the battery pack is controlled to output a predetermined discharge voltage; and the measured values ​​of the discharge voltage and the discharge SOC value of the battery pack under the control of any two different predetermined discharge voltages are respectively obtained;

[0087] The voltage difference between the two measured discharge voltages and the predetermined discharge voltage is calculated; and the calibration influence factor is readjusted according to the ratio of the difference between the two voltage differences and the difference between two different output discharge SOC values.

[0088] It should be noted that: when the battery pack is in the third discharge mode, the ratio of the difference between the two voltage differences and the difference between the two different output discharge SOC values ​​reflects the change in voltage error under different discharge SOCs, and can therefore be used to adjust the calibration influence factor to more accurately reflect the actual discharge characteristics of the battery pack, depending on the current calibration influence factor and the required adjustment direction (increase or decrease).

[0089] Example 2

[0090] See also Figure 2 As shown, the part not described in detail in this embodiment is described in the first embodiment, and a battery pack discharge SOC calibration method is provided, comprising the following steps:

[0091] Extracting corresponding discharge characteristic information under different discharge state information based on the historical database;

[0092] Receiving a discharge SOC value and discharge characteristic information that affects a change in the discharge SOC value, establishing a machine learning model between the discharge SOC value and the discharge characteristic information, and extracting a calibration influencing factor between the discharge SOC value and the discharge characteristic information based on the machine learning model;

[0093] A discharge stability coefficient is calculated based on the discharge characteristic information and the calibration influence factor, and a corresponding discharge calibration mode is matched based on the discharge stability coefficient; the calibration influence factor is reversely analyzed based on the discharge calibration mode and the calibration influence factor is updated based on the discharge calibration mode;

[0094] The discharge SOC value is calibrated based on the updated calibration influence factor.

[0095] The logic for obtaining the discharge characteristic information:

[0096] The historical database includes several groups of historical discharge characteristic information and discharge state information collected in an experimental environment; the experimental environment is a data collection environment in which a tester controls the change of the discharge state information corresponding to the battery pack to collect various discharge characteristic information values;

[0097] The discharge characteristic information includes the discharge rate per unit time, the discharge depth, the battery pack temperature change value, the battery pack rated power and the discharge function requirements of the battery pack;

[0098] The discharge state information includes a rapid discharge state, a basic discharge state and an abnormal discharge state;

[0099] In the experimental environment, by controlling battery packs with different rated power and discharge function requirements, the discharge characteristic information of the battery packs under fast discharge state, basic discharge state and abnormal discharge state is collected.

[0100] The training logic of the machine learning model is:

[0101] Each set of discharge characteristic information is used to generate a corresponding discharge SOC actual value according to the discharge state information when collected, and each set of discharge characteristic information is used as the input of the machine learning model. The machine learning model outputs the discharge SOC calibration value of each set of discharge characteristic information, takes the discharge SOC actual value as the prediction target, and takes minimizing the sum of the prediction accuracy of all discharge characteristic information as the training target; wherein the calculation formula for the prediction accuracy is: ,in, is the number of the discharge characteristic information, For prediction accuracy, For the The discharge SOC calibration value corresponding to the group discharge characteristic information, For the The actual discharge SOC value corresponding to the discharge characteristic information of the group is obtained; the machine learning model is trained until the sum of the prediction accuracies reaches convergence and the training is stopped.

[0102] The discharge state of the battery pack is recorded as , the operating unit time in the discharge state is marked as , then the discharge stability coefficient corresponding to the current battery pack is:

[0103] ;

[0104] For the In this state Discharge stability coefficient per unit time; For the In this state The discharge stability coefficient of the previous unit time of the unit time; ; Used to represent the current operating unit time of the battery pack; , and are all positive integers; For the In this state The discharge rate per unit time, For the In this state The change in discharge rate per unit time; For the In this state The depth of discharge per unit time, For the In this state The change in discharge depth per unit time; For the In this state The real-time temperature value of the battery pack per unit time, For the In this state The battery pack temperature change value per unit time; , and are the calibration influencing factors corresponding to the discharge characteristic information respectively.

[0105] The discharge stability coefficient determines the corresponding discharge calibration mode according to a plurality of preset value ranges, the value range includes , and ,

[0106] when When the discharge calibration mode is marked as the first discharge mode, the discharge SOC change value is identified based on the first discharge mode. If the discharge SOC change value is within the preset first calibration interval, whether the battery SOC value at the current moment is within the preset second calibration interval is determined; if the battery SOC value is within the preset second calibration interval, the discharge SOC value is calibrated based on the calibration influence factor corresponding to the current discharge stability coefficient;

[0107] when When the discharge calibration mode is marked as the second discharge mode, the calibration influence factor is reconfigured based on the second discharge mode, and the calibration influence factor includes a fixed weight factor and an adjustment weight factor; the discharge SOC value is calibrated based on the adjustment weight factor;

[0108] when When the discharge calibration mode is marked as the third discharge mode, the third discharge mode is a safe discharge mode; the calibration influence factor is readjusted based on the third discharge mode, and the adjusted calibration influence factor is updated as the calibration influence factor under the discharge state at the next moment.

[0109] The analysis logic of the calibration influencing factor includes:

[0110] Integrate different discharge characteristic information into a discharge characteristic matrix through normalization processing; the discharge characteristic matrix uniformly encodes and synchronously transmits the discharge characteristic information in a DC code format;

[0111] Perform correlation analysis on calibration influence factors of each discharge characteristic information and the discharge stability coefficient, wherein the calibration influence factors include fixed weight factors and adjustment weight factors;

[0112] A bipartite graph is constructed for each discharge feature information and the discharge stability coefficient, and the maximum weight between the discharge feature information and the discharge stability coefficient is marked as a fixed weight factor;

[0113] The correlation coefficient between the determined fixed weight factor and the discharge stability coefficient is marked as an adjustment weight factor, and the linear correlation between the adjustment weight factor and the fixed weight factor has a value range between -1 and +1.

[0114] The updating logic of the calibration influence factor is:

[0115] The discharge calibration mode is maintained in the third discharge mode, and the current battery pack discharge characteristic information and the corresponding calibration influencing factor are extracted;

[0116] Based on the current calibration influencing factor, the battery pack is controlled to output a predetermined discharge voltage; and the measured values ​​of the discharge voltage and the discharge SOC value of the battery pack under the control of any two different predetermined discharge voltages are respectively obtained;

[0117] The voltage difference between the two measured discharge voltages and the predetermined discharge voltage is calculated; and the calibration influence factor is readjusted according to the ratio of the difference between the two voltage differences and the difference between two different output discharge SOC values.

[0118] Example 3

[0119] like Figure 3As shown, for the part not described in detail in this embodiment, please refer to the description of Embodiment 1. This embodiment provides an electronic device, including: a processor and a memory, wherein the memory stores a computer program that can be called by the processor;

[0120] The processor executes a battery pack discharge SOC calibration method according to one aspect of the present invention by calling a computer program stored in the memory.

[0121] This embodiment also provides a computer program product stored on a computer-readable medium, including a computer-readable program, which, when executed on an electronic device, provides a user input interface to implement the battery pack discharge SOC calibration method.

[0122] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0123] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A battery pack discharge SOC calibration method, characterized in that: The following steps are involved: Extracting corresponding discharge characteristic information under different discharge state information based on the historical database; Receiving a discharge SOC value and discharge characteristic information that affects a change in the discharge SOC value, establishing a machine learning model between the discharge SOC value and the discharge characteristic information, and extracting a calibration influencing factor between the discharge SOC value and the discharge characteristic information based on the machine learning model; A discharge stability coefficient is calculated based on the discharge characteristic information and the calibration influencing factor, and a corresponding discharge calibration mode is matched based on the discharge stability coefficient; Reversely analyzing the calibration influence factor based on the discharge calibration mode and updating the calibration influence factor based on the discharge calibration mode; The discharge stability coefficient determines the corresponding discharge calibration mode according to a plurality of preset value ranges, the value range includes , and , when When the discharge calibration mode is marked as the first discharge mode, the discharge SOC change value is identified based on the first discharge mode. If the discharge SOC change value is within the preset first calibration interval, whether the battery SOC value at the current moment is within the preset second calibration interval is determined; if the battery SOC value is within the preset second calibration interval, the discharge SOC value is calibrated based on the calibration influence factor corresponding to the current discharge stability coefficient; when When the discharge calibration mode is marked as a second discharge mode, the calibration influence factor is reconfigured based on the second discharge mode, and the calibration influence factor includes a fixed weight factor and an adjustment weight factor; Calibrate the discharge SOC value based on the adjustment weight factor; when When , the discharge calibration mode is marked as a third discharge mode, and the third discharge mode is a safe discharge mode; Re-adjusting the calibration influence factor based on the third discharge mode, and updating the adjusted calibration influence factor as the calibration influence factor in the discharge state at the next moment; The analysis logic of the calibration influencing factor includes: Integrate different discharge feature information into a discharge feature matrix through normalization; The discharge characteristic matrix uniformly encodes and synchronously transmits the discharge characteristic information in a DC code format; Perform correlation analysis on calibration influence factors of each discharge characteristic information and the discharge stability coefficient, wherein the calibration influence factors include fixed weight factors and adjustment weight factors; A bipartite graph is constructed for each discharge feature information and the discharge stability coefficient, and the maximum weight between the discharge feature information and the discharge stability coefficient is marked as a fixed weight factor; The correlation coefficient between the determined fixed weight factor and the discharge stability coefficient is marked as an adjustment weight factor, and the linear correlation between the adjustment weight factor and the fixed weight factor has a value range between -1 and +1; The updating logic of the calibration influence factor is: The discharge calibration mode is maintained in the third discharge mode, and the current battery pack discharge characteristic information and the corresponding calibration influencing factor are extracted; Based on the current calibration influence factor, controlling the battery pack to output a predetermined discharge voltage; Respectively obtaining the measured values ​​of the discharge voltage and the discharge SOC value of the battery pack under the control of any two different predetermined discharge voltages; Calculating a voltage difference between the two measured discharge voltages and the predetermined discharge voltage; Re-adjusting the calibration influence factor according to the ratio of the difference between the two voltage differences and the difference between the two different output discharge SOC values; The discharge SOC value is calibrated based on the updated calibration influence factor.

2. A battery pack discharge SOC calibration method according to claim 1, characterized in that: The logic for obtaining the discharge characteristic information: The historical database includes several groups of historical discharge characteristic information and discharge state information collected in an experimental environment; the experimental environment is a data collection environment in which a tester controls the change of the discharge state information corresponding to the battery pack to collect various discharge characteristic information values; The discharge characteristic information includes the discharge rate per unit time, the discharge depth, the battery pack temperature change value, the battery pack rated power and the discharge function requirements of the battery pack; The discharge state information includes a rapid discharge state, a basic discharge state and an abnormal discharge state; In the experimental environment, by controlling battery packs with different rated power and discharge function requirements, the discharge characteristic information of the battery packs under fast discharge state, basic discharge state and abnormal discharge state is collected.

3. A battery pack discharge SOC calibration method according to claim 1, characterized in that: The training logic of the machine learning model is: Each set of discharge characteristic information is used to generate a corresponding discharge SOC actual value according to the discharge state information when collected, and each set of discharge characteristic information is used as the input of the machine learning model. The machine learning model outputs the discharge SOC calibration value of each set of discharge characteristic information, takes the discharge SOC actual value as the prediction target, and takes minimizing the sum of the prediction accuracy of all discharge characteristic information as the training target; wherein the calculation formula for the prediction accuracy is: ,in, is the number of the discharge characteristic information, For prediction accuracy, For the The discharge SOC calibration value corresponding to the group discharge characteristic information, For the The actual discharge SOC value corresponding to the discharge characteristic information of the group is obtained; the machine learning model is trained until the sum of the prediction accuracies reaches convergence and the training is stopped.

4. A battery pack discharge SOC calibration method according to claim 3, characterized in that: The discharge state of the battery pack is recorded as , the operating unit time in the discharge state is marked as , then the discharge stability coefficient corresponding to the current battery pack is: ； For the In this state Discharge stability coefficient per unit time; For the In this state The discharge stability coefficient of the previous unit time of the unit time; ; Used to represent the current operating unit time of the battery pack; , and All are positive integers; For the In this state The discharge rate per unit time, For the In this state The change in discharge rate per unit time; For the In this state The depth of discharge per unit time, For the In this state The change in discharge depth per unit time; For the In this state The real-time temperature value of the battery pack per unit time, For the In this state The battery pack temperature change value per unit time; , and are the calibration influencing factors corresponding to the discharge characteristic information respectively.

5. A battery pack discharge SOC calibration system, which is based on the implementation of a battery pack discharge SOC calibration method according to any one of claims 1 to 4, characterized in that: It includes a discharge characteristic determination unit, a discharge SOC analysis unit, a discharge stability analysis unit and a calibration control unit; A discharge characteristic determination unit, which extracts corresponding discharge characteristic information under different discharge state information based on a historical database; A discharge SOC analysis unit receives a discharge SOC value and discharge characteristic information that affects a change in the discharge SOC value, establishes a machine learning model between the discharge SOC value and the discharge characteristic information, and extracts a calibration influencing factor between the discharge SOC value and the discharge characteristic information based on the machine learning model; A discharge stability analysis unit, which calculates a discharge stability coefficient based on discharge characteristic information and a calibration influencing factor, and matches a corresponding discharge calibration mode based on the discharge stability coefficient; Reversely analyzing the calibration influence factor based on the discharge calibration mode and updating the calibration influence factor based on the discharge calibration mode; The calibration control unit calibrates the discharge SOC value based on the updated calibration influence factor.

6. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; The processor executes the battery pack discharge SOC calibration method according to any one of claims 1 to 4 by calling the computer program stored in the memory.

7. A computer program product stored on a computer readable medium, characterized in that It comprises a computer readable program, which, when executed on an electronic device, provides a user input interface to implement a battery pack discharge SOC calibration method as described in any one of claims 1 to 4.

Images