A power battery fault early warning method, a vehicle and a computer readable storage medium

Abstract

The application provides a power battery fault early warning method, a vehicle and a computer readable storage medium, and relates to the technical field of power. The method comprises the following steps: acquiring first information, wherein the first information comprises battery state data of a power battery, charging characteristic data of the power battery and user behavior data, the user behavior data comprises driving habit data and charging habit data of a user; then, inputting the first information into a battery fault early warning model to obtain an early warning level output by the battery fault early warning model and a solution strategy for the early warning level; and then, displaying early warning prompt information comprising the early warning level and the solution strategy. The technical scheme of the application can reduce the false positive rate of power battery fault early warning, improve the accuracy of power battery fault early warning, improve the timeliness of power battery fault early warning, and improve the user experience.

Description

Technical Field

[0001] This application relates to the field of power technology, specifically to a power battery fault early warning method, a vehicle, and a computer-readable storage medium. Background Technology

[0002] A power battery is a rechargeable battery that provides power to vehicles such as electric cars and electric trains. The absence of power battery failure is a fundamental condition for the normal operation of a vehicle. Therefore, predicting whether a power battery is about to fail and alerting users to address this impending failure in a timely manner has become a pressing technical problem.

[0003] In related technologies, the method for detecting whether a power battery is about to fail is to collect the temperature of the surface of the battery cell. If the temperature of the battery cell surface is greater than a fixed trigger threshold, such as 50 degrees Celsius, it indicates that the power battery is about to fail.

[0004] However, during fast charging, the instantaneous peak temperature of the battery cell surface can reach 51°C, but this does not actually reach a dangerous state. In this situation, the power battery is not actually in a dangerous state, but it will be falsely reported as an impending failure. The aforementioned method for detecting whether a power battery is about to fail has a high false alarm rate, for example, as high as 35%, resulting in a poor user experience. Summary of the Invention

[0005] This application provides a power battery fault early warning method, a vehicle, and a computer-readable storage medium, which can reduce the false alarm rate of power battery fault early warning, improve the accuracy of power battery fault early warning, improve the timeliness of power battery fault early warning, and enhance the user experience.

[0006] To achieve the above objectives, the embodiments of this application adopt the following technical solutions: Firstly, a method for early warning of power battery faults is provided. This method is applied to a vehicle, which contains a power battery. The method includes: acquiring first information; wherein the first information includes battery state data, charging characteristic data, and user behavior data, the user behavior data including driving habit data and charging habit data; inputting the first information into a battery fault early warning model to obtain a warning level and a resolution strategy for the warning level output by the battery fault early warning model; wherein the warning level characterizes the severity of a potential danger to the power battery, and the severity of the potential danger is positively correlated with the probability of a power battery fault; and displaying a warning message; wherein the warning message includes the warning level and the resolution strategy.

[0007] This application does not rely solely on the surface temperature of the battery cells to predict whether a power battery is about to malfunction. Instead, it comprehensively utilizes battery state data, charging characteristic data, and user behavior data to predict whether a power battery is about to malfunction. This reduces the false alarm rate of power battery malfunction warnings and improves their accuracy. Furthermore, when a power battery malfunction is predicted, it displays warning information including the warning level and resolution strategy, thereby improving the timeliness of power battery malfunction warnings and enhancing the user experience.

[0008] In conjunction with the first aspect, in one possible design, the power battery includes multiple cells, and the battery status data includes the voltage and temperature of each cell; the battery fault early warning model is used to determine the early warning level as the first early warning level when there is a cell whose temperature is greater than a temperature threshold and whose voltage change rate is greater than a voltage change rate threshold.

[0009] In this application, if the surface temperature of a battery cell exceeds a temperature threshold and the voltage change rate of that cell exceeds a voltage change rate threshold simultaneously, this is an extremely strong characteristic signal that thermal runaway of the power battery is about to occur. This is defined as the highest level, the first warning level, i.e., a red warning. This means that danger is very imminent, and continued operation may immediately lead to safety accidents such as fire, explosion, and electric shock. This reduces the false alarm rate of power battery fault warnings and improves their accuracy.

[0010] In conjunction with the first aspect, in one possible design approach, the battery fault early warning model is also used to determine the heat transfer relationship between cells in the power battery based on battery state data and thermal coupling coefficient when the early warning level is determined to be the first early warning level; predict the target set of cells whose surface temperature will exceed the temperature threshold based on the first cell and the heat transfer relationship; and determine the solution strategy for the first early warning level based on the target set of cells; wherein, the first cell is the cell whose surface temperature exceeds the temperature threshold.

[0011] In this application, the battery fault early warning model is also used to accurately locate the first cell where thermal runaway occurs and the cells that may be transferred heat based on the fault propagation probability matrix when the early warning level is determined to be the first early warning level, thereby improving the accuracy of the solution strategy.

[0012] In conjunction with the first aspect, in one possible design approach, the battery fault early warning model is also used to obtain a user behavior score when a cell's temperature is less than or equal to a temperature threshold, or its voltage change rate is less than or equal to a voltage change rate threshold. The user behavior score characterizes the impact of user behavior on the health of the power battery. The battery fault early warning model includes a user behavior scoring model, which is used to obtain a user behavior score based on the first information. If the user behavior score is greater than the user behavior score threshold, the early warning level is determined to be the third early warning level.

[0013] In this application, the battery fault early warning model is also used to determine the early warning level more accurately by using user behavior scoring when the early warning level is determined to be a non-first early warning level.

[0014] In conjunction with the first aspect, in one possible design approach, the battery fault early warning model is also used to obtain the charging health of the power battery from the battery status data when the user behavior score is less than or equal to the user behavior score threshold; wherein, the charging health is the ratio of the current chargeable capacity to the rated capacity of the power battery; and when the charging health of the power battery is greater than the charging health threshold, the early warning level is determined to be the second early warning level.

[0015] In this application, the battery fault early warning model is also used to determine the early warning level more accurately by utilizing the charging health of the power battery when the user behavior score is less than or equal to the user behavior score threshold.

[0016] In conjunction with the first aspect, in one possible design approach, the battery fault warning model is used to determine the solution strategy for the warning level as prompting the user to stop charging the vehicle's power battery when the warning level is the first warning level and the vehicle's power battery is being charged; or, the battery fault warning model is used to determine the solution strategy for the warning level as prompting the user to turn on the vehicle's ventilation system to cool the power battery and / or notifying the user to evacuate the vehicle when the warning level is the first warning level and the vehicle's power battery is being driven.

[0017] In this application, for different warning levels caused by different reasons, the application provides users with solutions for different reasons, thereby improving the effectiveness of the solutions.

[0018] In conjunction with the first aspect, in one possible design approach, the battery fault warning model is used to determine the solution strategy for the warning level as prompting the user to use driving operations that reduce damage to the power battery when the warning level is the second warning level; the battery fault warning model is used to determine the solution strategy for the warning level as prompting the user to use the target charging strategy when the warning level is the third warning level.

[0019] This application provides users with solutions for different scenarios, thereby improving the effectiveness of these solutions.

[0020] In conjunction with the first aspect, in one possible design approach, the battery fault early warning model includes an adaptation model, a first battery fault early warning sub-model, and a second battery fault early warning sub-model. The adaptation model is used to determine the data integrity of the first information. If the data integrity is greater than a first threshold, the early warning level and the solution strategy for the early warning level output by the first battery fault early warning sub-model are used as the early warning level and the solution strategy for the early warning level output by the battery fault early warning model. The first battery fault early warning sub-model is used to obtain the early warning level and the solution strategy for the early warning level based on the first information. If the data integrity is less than or equal to the first threshold, the early warning level and the solution strategy for the early warning level output by the second battery fault early warning sub-model are used as the early warning level and the solution strategy for the early warning level output by the battery fault early warning model. The second battery fault early warning sub-model is used to obtain the early warning level and the solution strategy for the early warning level based on the electrochemical parameters of the power battery.

[0021] In this application, the battery fault early warning model includes a first battery fault early warning sub-model and a second battery fault early warning sub-model. If the data integrity of the first information is high, the early warning level and corresponding solution strategy output by the first battery fault early warning sub-model are relied upon. This is because the second battery fault early warning sub-model can obtain a more accurate early warning level and corresponding solution strategy when the first information is relatively complete. If the data integrity of the first information is low, the early warning level and corresponding solution strategy output by the second battery fault early warning sub-model are relied upon. This is because the first battery fault early warning sub-model can obtain a relatively accurate early warning level and corresponding solution strategy based on the electrochemical parameters of the power battery when the first information is incomplete. In summary, regardless of the data integrity of the first information, a relatively accurate early warning level and corresponding solution strategy can be obtained by relying on models with different characteristics.

[0022] Secondly, embodiments of this application provide a vehicle, including: a processor; and a memory for storing processor-executable instructions, wherein the processor is used to execute the power battery fault warning method described in the first aspect.

[0023] Thirdly, embodiments of this application provide a computer-readable storage medium storing a computer program for executing the power battery fault early warning method described in the first aspect.

[0024] Fourthly, embodiments of this application provide a computer program product, which includes a computer program. When the computer program is executed by the processor of a computer device, it enables the computer device to execute the power battery fault warning method described in the first aspect.

[0025] The technical effects of any of the design methods in the second to fourth aspects can be found in the technical effects of different design methods in the first aspect, and will not be repeated here. Attached Figure Description

[0026] Figure 1 The diagram shown is an application scenario illustration of the power battery fault early warning method provided in an exemplary embodiment of this application.

[0027] Figure 2 The diagram shown is a flowchart illustrating a power battery fault early warning method provided in an exemplary embodiment of this application.

[0028] Figure 3 The diagram shown is a schematic diagram of obtaining the warning level using a battery fault warning model according to an exemplary embodiment of this application.

[0029] Figure 4 The diagram shown is a schematic representation of the process by which a battery fault warning model outputs a warning level, as provided in an exemplary embodiment of this application.

[0030] Figure 5 The diagram shown is a schematic representation of the training process of a battery fault early warning model provided in an exemplary embodiment of this application.

[0031] Figure 6 The diagram shown is a structural schematic of a power battery fault early warning device provided in an exemplary embodiment of this application. Detailed Implementation

[0032] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0033] Application Overview As described in the background section, the instantaneous peak temperature of the battery cell surface can reach 51°C during fast charging, but this does not actually reach a dangerous state. In this situation, the power battery is not actually in a dangerous state, but it will be falsely reported as an impending failure. The aforementioned method for detecting whether a power battery is about to fail has a high false alarm rate, for example, as high as 35%, resulting in a poor user experience.

[0034] To address the aforementioned technical problems, this application provides a power battery fault early warning method. This method is applied to a vehicle equipped with a power battery. The method includes: acquiring first information, which includes battery state data, charging characteristic data, and user behavior data, including user driving habit data and charging habit data; then, inputting the first information into a battery fault early warning model to obtain a warning level and a resolution strategy for the warning level output by the model; the warning level characterizes the severity of a potential danger to the power battery, with a higher severity indicating a greater probability of battery failure; and finally, displaying a warning message including the warning level and the resolution strategy.

[0035] In this embodiment, instead of solely using the surface temperature of the battery cell to predict whether a power battery is about to malfunction, the system comprehensively utilizes battery state data, charging characteristic data, and user behavior data to predict this. This reduces the false alarm rate of power battery fault warnings and improves their accuracy. Furthermore, when a power battery fault is predicted, a warning message including the warning level and resolution strategy is displayed, thereby improving the timeliness of power battery fault warnings and enhancing the user experience.

[0036] Exemplary scenario The power battery fault early warning method provided in this application can be applied to various vehicles equipped with power batteries. Vehicles include various types, such as sedans, SUVs, buses, and trucks. This application does not impose any special limitations on the specific form of the vehicle.

[0037] The following is an illustrative diagram illustrating an application scenario of a power battery fault early warning method. For example, Figure 1 The diagram illustrates an application scenario of a power battery fault early warning method provided in an exemplary embodiment of this application. Figure 1 As shown, the scenario includes a vehicle 100, in which a power battery 101 and a display screen 102 are installed.

[0038] Vehicle 100 can acquire first information, which includes battery status data of power battery 101, charging characteristic data of power battery 101, and user behavior data, including user driving habit data and charging habit data; input the first information into the battery fault warning model to obtain the warning level and the solution strategy for the warning level output by the battery fault warning model. The warning level is used to characterize the severity of the danger of power battery 101. The higher the severity of the danger of power battery 101, the greater the probability (likelihood) of power battery 101 failing; then, display the warning prompt information including the warning level and the solution strategy on the display screen 102.

[0039] It should be understood that the above application scenario examples are only shown to facilitate understanding of the spirit and principles of this application, and the embodiments of this application are not limited thereto. Rather, the embodiments of this application can be applied to any applicable scenario.

[0040] Exemplary methods Figure 2 The diagram shown is a flowchart illustrating a power battery fault early warning method provided in an exemplary embodiment of this application. Figure 2 The method can be performed by a vehicle, such as Figure 1 Vehicle 100 is executed. For example... Figure 2 As shown, the power battery fault early warning method may include the following: 210: Obtain first information; wherein, the first information includes battery status data of the power battery, charging characteristic data of the power battery, and user behavior data, and the user behavior data includes user driving habit data and charging habit data.

[0041] In some embodiments, a preset control is displayed on the vehicle's screen, and the vehicle can perform step 210 in response to a user's tapping action on the control. It is understood that a tapping action is merely one example, and other gestures or voice actions are also possible, but not limited to these.

[0042] Battery status data may include any one or more of the following: voltage of each cell in the current power battery, surface temperature of each cell, internal resistance of each cell, and charging health of the power battery. The power battery may include one or more cells, which can be connected in series or in parallel, but are not limited to these methods. Charging health will be described in detail below.

[0043] Charging characteristic data can include charging curves and electrochemical impedance spectroscopy (EIS) data obtained from charging stations during the charging process over a historical period. The charging curve describes the change in terminal voltage of the power battery over time during charging. This curve reflects the battery's polarization characteristics, phase transition reactions, and kinetic performance through the voltage-time relationship. Electrochemical impedance spectroscopy (EIS), a commonly used technique for assessing battery health, captures key electrochemical characteristics such as internal resistance, charge transfer behavior, and capacitance dynamics by analyzing the battery's impedance response at different frequencies. EIS can be used to assess the state of lithium-ion batteries, such as the health status mentioned below.

[0044] User driving habit data can be time series data such as acceleration behavior data and braking behavior data.

[0045] Among these, acceleration behavior data can be the number of times of rapid acceleration (e.g., acceleration exceeding 3 meters per second squared) within a certain period. Similarly, braking behavior data can be the number of times of emergency braking (e.g., deceleration exceeding 4 meters per second squared) within a certain period. The sampling rate for user driving habit data within the aforementioned time period can be 1 Hz, but is not limited to this.

[0046] Charging habit data can include charging frequency data, charging time data, charging amount data, charging location data, charging method data, and remaining battery power at the start of charging.

[0047] Among these, charging frequency data refers to the number of times a user charges within a certain period, for example, an average of 3 times per week. Charging time data can refer to the duration of each charge, for example, an average of 4 hours per charge. Charging capacity data can refer to the average amount of electricity charged within a certain period. Charging location data can refer to the average number of times a user charges at home and the average number of times they charge at a public charging station within a certain period. Charging method data refers to the average number of fast charges and slow charges within a certain period.

[0048] 220: Input the first information into the battery fault early warning model to obtain the early warning level and the solution strategy for the early warning level output by the battery fault early warning model.

[0049] Figure 3 The diagram shown is a schematic representation of an exemplary embodiment of this application, illustrating the use of a battery fault early warning model to obtain early warning levels. Figure 3 As shown in Figure (a), the first information is input into the battery fault early warning model to obtain the early warning level and the solution strategy for the early warning level output by the battery fault early warning model.

[0050] The warning level is used to characterize the severity of the danger posed by the power battery. The higher the severity of the danger, the greater the probability of the power battery malfunctioning.

[0051] If the severity of the danger posed by the power battery is low, to the point of no danger, the warning level is low, and the warning level in this case can be called the fault-free level. If the severity of the danger posed by the power battery is high, it can be divided into the third warning level, the second warning level, and the first warning level according to the severity from low to high.

[0052] If colors are used to represent warning levels, the first warning level is red, the second warning level is orange, and the third warning level is blue. The first warning level (red) usually corresponds to abnormal voltage and temperature. These types of faults develop rapidly and have serious consequences (such as thermal runaway), and are considered high-risk warnings.

[0053] The second warning level (orange warning) usually corresponds to the battery's health status, which is of medium risk and reflects the aging trend of the power battery, thus having mid-term warning value. The third warning level (blue warning) usually corresponds to abnormal mechanical expansion force, which is an indicator of early chemical failure, earlier than temperature rise and voltage changes. It belongs to the low to medium risk warning category, but still has early warning value.

[0054] A green alert generally indicates that the system is normal or in a safe state.

[0055] The following examples illustrate the corresponding solutions for different warning levels.

[0056] For example, if the warning level is the highest level and the vehicle is currently charging its battery, the solution is to prompt the user to stop charging the battery.

[0057] For example, if the warning level is the highest level and the vehicle is currently in motion, the solution would be to prompt the user to turn on the vehicle's ventilation system to cool the power battery and / or notify the user to evacuate the vehicle.

[0058] For example, if the warning level is the second warning level, the solution is to prompt the user to use driving operations that reduce damage to the power battery.

[0059] For example, if the warning level is the third warning level, the solution is to prompt the user to use the target charging strategy.

[0060] In some other embodiments, the vehicle can determine a resolution strategy for the warning level via a cloud server. Specifically, the vehicle can send the warning level, a resolution strategy 1 for the warning level, and first information to the cloud server. The cloud server has a resolution strategy generation model set up. The cloud server can input the warning level, the resolution strategy 1 for the warning level, and the first information into the resolution strategy generation model, obtain the resolution strategy 2 output by the resolution strategy generation model, and feed the resolution strategy 2 back to the vehicle. The content of resolution strategy 2 is more extensive than that of resolution strategy 1.

[0061] 230: Displays early warning information; the early warning information includes the early warning level and the solution strategy.

[0062] For example, if the warning level is the first warning level and the vehicle is currently charging its power battery, the output warning message will include the first warning level and a prompt to the user to stop charging the vehicle's power battery.

[0063] For example, if the warning level is the first warning level and the vehicle is currently in motion, the output warning information will include the first warning level, a prompt to the user to turn on the vehicle's ventilation system to cool the power battery, and / or a notification to the user to evacuate the vehicle.

[0064] For example, if the warning level is the second warning level, the output warning information will include prompts to the user to use driving operations that reduce damage to the power battery.

[0065] For example, if the warning level is the third warning level, the output warning information includes the third warning level and a prompt for the user to use the target charging strategy. In this embodiment, instead of solely using the temperature of the battery cell surface to predict whether the power battery will malfunction, it comprehensively utilizes the battery state data, charging characteristic data, and user behavior data to predict whether the power battery will malfunction. This reduces the false alarm rate of power battery malfunction warnings and improves their accuracy. Furthermore, when a malfunction is predicted, a warning information including the warning level and resolution strategy is displayed, thus improving the timeliness of power battery malfunction warnings and enhancing the user experience.

[0066] The following is an example of an implementation scheme for the battery fault early warning model to output the early warning level in step 220 above. For example, Figure 4 The diagram shown is a schematic representation of the process by which a battery fault early warning model outputs an early warning level, as provided in an exemplary embodiment of this application. Figure 4 As shown, the method includes the following steps 410 and 470: 410: Input the first information into the battery fault early warning model.

[0067] 420: Determine if there exists a cell whose temperature is greater than a temperature threshold and whose voltage change rate is greater than a voltage change rate threshold.

[0068] A power battery can have one or more cells. If the surface temperature of one cell i is greater than the temperature threshold, it means that a violent exothermic side reaction has begun to occur inside the power battery. The exothermic side reaction can refer to the decomposition of the solid electrolyte interface (SEI). The SEI film is a passivation layer formed by the reduction reaction of the electrolyte on the surface of the negative electrode (such as graphite, silicon-based materials, etc.) during the first charge and discharge process of a lithium-ion battery.

[0069] If the voltage change rate of cell i is greater than the voltage change rate threshold, this is an extremely large change rate, indicating a sudden voltage drop, which usually means that a short circuit is occurring inside the power battery.

[0070] If the surface temperature of a cell i is greater than the temperature threshold, and the voltage change rate of cell i is greater than the voltage change rate threshold, these two situations occur simultaneously. This is a very strong characteristic signal that thermal runaway of the power battery is about to occur. Thermal runaway refers to the temperature of the cell surface being greater than the temperature threshold within a preset time period. In other words, thermal runaway of a cell refers to the phenomenon that the temperature rises sharply and triggers a chain reaction when the heat generated by the internal chemical reaction of the battery cell cannot be dissipated in time during charging or use.

[0071] If the surface temperature of a cell i is greater than the temperature threshold and the voltage change rate of cell i is greater than the voltage change rate threshold, the output warning level is the first warning level, i.e., red warning. This means that danger is very imminent and continued operation may immediately cause safety accidents such as fire, explosion, and electric shock.

[0072] The temperature threshold can be 65 degrees (°C), and the voltage change rate threshold can be 5 millivolts per second (5 mV / s).

[0073] Specifically, if the temperature of a battery cell exceeds a temperature threshold and its voltage change rate exceeds a voltage change rate threshold, the warning level is set to the first warning level, i.e., step 330 is executed. Otherwise, user behavior scores are further obtained to determine the warning level, i.e., step 340 is executed. 430: The warning level is set as the first warning level.

[0074] 440: Obtain user behavior scores; whereby user behavior scores are used to characterize the impact of user behavior on the health of the power battery.

[0075] In one example, the battery fault warning model includes a user behavior scoring model, which is used to obtain a user behavior score based on the first piece of information. The higher the user behavior score, the less damage the user's behavior causes to the health of the power battery, and the lower the severity of the danger to the power battery.

[0076] 450: Determine whether the user behavior score is greater than the user behavior score threshold.

[0077] The user behavior scoring threshold is a critical threshold used to determine whether the warning level of the power battery is the third warning level. For example, the user behavior scoring threshold could be 0.7.

[0078] Specifically, if the user behavior score is greater than the user behavior score threshold, it indicates that the user's behavior is less damaging to the health of the power battery, and the severity of the danger to the power battery is lower. The warning level is then set to the third warning level, i.e., step 460 is executed. If the user behavior score is less than or equal to the user behavior score threshold, it indicates that the user's behavior is more damaging to the health of the power battery, and the severity of the danger to the power battery is higher. The charging health of the power battery is then continuously assessed to further determine the degree of danger, i.e., step 470 is executed.

[0079] 460: The warning level has been set as Level 3.

[0080] 470: Obtain the charging health of the power battery from the battery status data.

[0081] State of Health (SOH), also known as charging health status, measures the degree to which a battery retains its performance compared to its original, brand-new state. One calculation method is to use the ratio of the current chargeable capacity to the battery's rated capacity. For example, a brand-new battery with a rated capacity of 60 kWh, after several years of use, may now only be able to charge a maximum of 51 kWh. Its charging health status would then be: 51 kWh / 60 kWh = 85%, meaning its capacity has decreased by 15%, and it currently only retains 85% of its original healthy capacity.

[0082] It is understandable that, since charging health generally only decreases after several years and remains constant for a long period of time, while the impact of user behavior on the power battery varies greatly, the warning level can be determined by using user behavior scores first, and then by using charging health to determine the warning level.

[0083] 480: Determine whether the charging health of the power battery is greater than the charging health threshold.

[0084] The charging health threshold is a critical threshold used to determine whether the warning level of a power battery is at the second warning level. For example, the charging health threshold could be 0.75.

[0085] Specifically, if the charging health status of the power battery is greater than the charging health threshold, it indicates that the remaining healthy capacity of the power battery is relatively large, and it is determined that the power battery is not in danger, i.e., proceed to step 490. If the charging health status of the power battery is less than or equal to the charging health threshold, it indicates that the remaining healthy capacity of the power battery is relatively small, and there is a certain degree of danger. The warning level is determined to be the second warning level, i.e., proceed to step 4100.

[0086] 490: Confirm that the power battery is not dangerous.

[0087] 4100: The warning level is set as the second warning level.

[0088] In related technologies, only temperature sensors can determine if the surface temperature of a certain battery cell exceeds a temperature threshold. This means that only the heat source is known, but it's unknown which other battery cells the heat from the cell exceeding the threshold will transfer to, i.e., the heat propagation path is unknown. In the embodiments of this application, by determining the heat transfer relationship between battery cells, the heat source can be located, and other battery cells about to experience thermal runaway can also be identified, thus accurately locating the battery cell about to experience thermal runaway.

[0089] Specifically, the battery fault early warning model is also used to determine the heat transfer relationship between cells in the power battery based on battery state data and thermal coupling coefficient when the early warning level is determined to be the first early warning level; to predict the target set of cells whose surface temperature will exceed the temperature threshold based on the first cell whose surface temperature exceeds the temperature threshold and the heat transfer relationship; and to determine the solution strategy for the first early warning level based on the target set of cells.

[0090] The heat transfer relationship can be represented by a fault propagation probability matrix; where the size of the fault propagation probability matrix is ​​M*N, N is the number of cells, and each element P-ij in the fault propagation probability matrix represents the probability that if the temperature on the surface of cell i is greater than the temperature threshold, the temperature on the surface of the adjacent cell j will also be greater than the temperature threshold.

[0091] If the value of element P-ij is greater than or equal to the probability threshold, then if the surface temperature of cell i is greater than the temperature threshold, cell j is identified as the target cell. For example, the probability threshold could be 0.9. The following is an example of a 3*3 fault propagation probability matrix, as detailed in Table 1 below.

[0092] Table 1 As shown in Table 1, P_12 equals 0.9, which means that if the surface temperature of cell 1 is greater than the temperature threshold, there is a 90% probability that the surface temperature of cell 2 will also exceed the temperature threshold. If the surface temperature of cell 1 is greater than the temperature threshold, cell 2 is identified as the target cell.

[0093] P_23=0.95, which means that if the surface temperature of cell 2 is greater than the temperature threshold, there is a 95% probability that the surface temperature of cell 3 will also exceed the temperature threshold. If the surface temperature of cell 2 is greater than the temperature threshold, cell 2 is identified as the target cell.

[0094] Battery status data can include the voltage of each cell in the current power battery, the surface temperature of each cell, and the internal resistance of each cell.

[0095] The thermal coupling coefficient, also known as the edge weight, is calculated based on the thermal conductivity and current distribution. The thermal coupling coefficient between battery cells typically refers to the thermal conductivity (such as the thermal conductivity of a thermally conductive silicone pad). Its value can be determined in the following ways: Method 1 (Experimental measurement, pre-stored in the vehicle): The thermal resistance (Rth) between the cells of the power battery is directly measured using a thermal resistance tester or a thermal conductivity meter, and the thermal coupling coefficient is determined by the following formula (1): (1) Where A is the heat dissipation area.

[0096] Method 2 (Theoretical Estimation): If the cell spacing and heat dissipation material parameters (such as thermal conductivity λ, thickness, etc.) of the power battery are known, the thermal coupling coefficient can be determined by the following formula (2): (2) Where L is the length of the heat conduction path and ΔT is the temperature difference.

[0097] Based on the target set of battery cells, the solution strategy for the first warning level could be to alert the user that battery cell A has overheated. According to thermal field simulation, the heat will be conducted to the core battery cell B after 60 seconds, and it is recommended to immediately carry out targeted cooling of the areas where battery cells A and B are located.

[0098] In some other embodiments, the battery fault early warning model may include a first battery fault early warning sub-model and a second battery fault early warning sub-model. The first battery fault early warning sub-model is used to obtain the early warning level and the solution strategy for the early warning level based on first information. The scheme for obtaining the early warning level and the solution strategy based on the first information has been described above and will not be repeated here. The second battery fault early warning sub-model is used to obtain the early warning level and the solution strategy based on the electrochemical parameters of the power battery. The electrochemical parameters of the power battery may be lithium-ion concentration distribution, solid-liquid phase potential, electrolyte diffusion, etc.

[0099] Based on the battery fault early warning model, which includes a first battery fault early warning sub-model and a second battery fault early warning sub-model, the battery fault early warning model is used to determine the data integrity of the first information. If the data integrity is greater than a first threshold, the early warning level and the corresponding solution strategy output by the first battery fault early warning sub-model are used as the early warning level and corresponding solution strategy output by the battery fault early warning model. If the data integrity is less than or equal to the first threshold, the early warning level and the corresponding solution strategy output by the second battery fault early warning sub-model are used as the early warning level and corresponding solution strategy output by the battery fault early warning model. Data integrity refers to the ratio between the quantity of first information and the target total quantity of first information. The target total quantity of first information can be set in advance. For example, if the target total quantity of first information is 100 points to be received per second, but 20 points are lost, and the quantity of first information is 80, then the data integrity is 0.8.

[0100] For example, such as Figure 3 As shown in Figure (b), the battery fault early warning model includes a first battery fault early warning sub-model, a second battery fault early warning sub-model, and an adaptation model. First information is input into the first battery fault early warning sub-model, which is used to output early warning level 1 and a solution strategy 1 for the early warning level based on the first information. The second battery fault early warning sub-model is used to output early warning level 2 and a solution strategy 2 for the early warning level based on the electrochemical parameters of the power battery.

[0101] An adaptation model is used to determine the data integrity of the first information. If the data integrity is greater than a first threshold, the warning level 1 and the solution strategy 1 for the warning level output by the first battery fault warning sub-model are used as the warning level and solution strategy for the warning level output by the battery fault warning model. If the data integrity is less than or equal to the first threshold, the warning level 2 and the solution strategy 2 for the warning level output by the second battery fault warning sub-model are used as the warning level and solution strategy for the warning level output by the first battery fault warning sub-model. In this embodiment, the battery fault warning model includes a first battery fault warning sub-model and a second battery fault warning sub-model. If the data integrity of the first information is high, the warning level and solution strategy for the warning level output by the first battery fault warning sub-model are relied upon, because the second battery fault warning sub-model can obtain a more accurate warning level and solution strategy for the warning level when the first information is relatively complete. If the completeness of the first information is low, the warning level and corresponding solution strategy output by the second battery fault warning sub-model will be relied upon. This is because the first battery fault warning sub-model can obtain a relatively accurate warning level and solution strategy based on the electrochemical parameters of the power battery when the first information is incomplete. In summary, regardless of the completeness of the first information, relatively accurate warning levels and solution strategies can be obtained by relying on models with different characteristics.

[0102] In some embodiments, the first information is processed by any one or more of time synchronization, spatial mapping, feature normalization and vibration noise removal to obtain processed first information, and the processed first information is input into the battery fault early warning model.

[0103] Time synchronization refers to the use of a Precision Time Protocol (PTP) to collect information from multiple devices under the same clock count. Spatial mapping refers to the use of visual or laser sensors on a charging station to detect the coordinates of a hot spot or bulge point on the battery pack. These coordinates need to be converted into coordinates within the battery pack coordinate system recognized by the vehicle's battery management system (BMS). This allows for accurate matching to a specific battery cell.

[0104] Feature normalization refers to the fact that voltage (unit: V, numerical range: approximately 2.5-4.2), temperature (unit: °C, numerical range: approximately -20 to 60), and internal resistance (unit: mΩ, numerical range: approximately 0.1-1) have vastly different dimensions and numerical ranges. When these features are directly input into a model (such as a GNN), the model will determine their importance based on the magnitude of the input feature's value. For example, the model might consider features with larger values ​​(such as temperature) to be more important, while features with smaller values ​​(such as internal resistance) to be less important.

[0105] If one feature has a large numerical range (e.g., temperature -20 to 60°C), while another feature has a small numerical range (e.g., internal resistance 0.1 to 1), the model might mistakenly consider the feature with the larger numerical value (temperature) more important because its changes appear to have a greater impact on the model output. Conversely, the feature with the smaller numerical value (internal resistance) might be ignored by the model because its changes appear to have a small impact on the model output.

[0106] Feature normalization (such as min-max normalization or standardization) eliminates the influence of dimensions and numerical ranges by mapping the numerical ranges of different features to the same scale (e.g., 0~1 or a standard normal distribution). Vehicle vibration affects sensor readings. Every object has its natural frequency. If the frequency of vehicle vibration matches the natural frequency of the sensor's internal structure, resonance occurs. This causes the internal components of the sensor to undergo deformation or movement far greater than normal, resulting in severely distorted readings or even mechanical damage to the sensor. In this embodiment, vibration noise removal technology can be used to remove vibration noise from the first information. Specifically, in one example, vibration noise removal refers to using the vehicle's existing acceleration sensor (typically used in airbags, ESP systems, etc.) or additional vibration sensors to collect vibration data (including acceleration ranges of 0.5-2G) in real time under driving, bump-crossing, and braking conditions. Simultaneously, signal processing techniques such as high-pass filtering and wavelet transform are used to separate the "vibration noise" component from the raw voltage / temperature signal collected by the BMS. In this way, a pure battery signal and a vibration signal representing environmental interference are obtained.

[0107] The training process of the battery fault early warning model will be described below.

[0108] Figure 5 The diagram shown is a schematic representation of the training process of a battery fault early warning model provided in an exemplary embodiment of this application. Figure 5 The method can be executed by training equipment, which can be a computer, server, etc. For example... Figure 5 As shown, the training process includes the following: 510: Obtain the first training sample, wherein the first training sample includes the first information.

[0109] 520: Input the first training sample into the battery fault early warning model to obtain the early warning level and solution strategy.

[0110] The battery fault early warning model is used to perform time-domain feature processing and frequency-domain feature processing on the first training sample to obtain time-domain features and frequency-domain features. Then, the time-domain features and frequency-domain features are subjected to feature cross processing to obtain cross features. Based on the cross features, the early warning level and solution strategy are obtained.

[0111] Time-domain features refer to statistical characteristics directly calculated over time. Examples include the mean (reflecting average level) and variance (reflecting fluctuation). For voltage, time-domain features can also include the maximum, minimum, and slope within a sliding window.

[0112] Frequency domain features are obtained by converting a signal from the time domain to the frequency domain using Fourier transform (FFT) or wavelet transform.

[0113] Feature crossover refers to the mathematical combination of different features to generate new features that better reveal the problem. For example, ΔV×ΔT (the product of the rate of change of voltage and the rate of change of temperature).

[0114] 530: Determine the first loss value based on the early warning level, the resolution strategy, and the target.

[0115] Specifically, the target warning level and target solution strategy are the true labels for the warning level and solution strategy output by the battery fault warning model.

[0116] 540: The battery fault warning model is trained based on the first loss value to obtain the trained battery fault warning model.

[0117] Exemplary device Figure 6 The diagram shown is a structural schematic of a power battery fault early warning device provided in an exemplary embodiment of this application. Figure 6 As shown, the power battery fault early warning device 600 includes: an acquisition module 610, a determination module 620, and a display module 630.

[0118] The acquisition module 610 acquires first information, including battery status data, charging characteristic data, and user behavior data, which includes user driving habits and charging habits. The determination module 620 inputs the first information into a battery fault warning model to obtain the warning level and the solution strategy for the warning level. The warning level characterizes the severity of the danger posed by the power battery, and the severity of the danger is positively correlated with the probability of a power battery failure. The display module 630 displays warning information, including the warning level and the solution strategy.

[0119] This application provides a power battery fault early warning device. Instead of relying solely on the temperature of the battery cell surface to predict whether a fault is imminent, it comprehensively utilizes battery state data, charging characteristic data, and user behavior data to predict the potential fault. This reduces the false alarm rate and improves the accuracy of power battery fault early warnings. Furthermore, when a fault is predicted, it displays a warning message including the warning level and resolution strategy, thereby improving the timeliness of power battery fault early warnings and enhancing the user experience.

[0120] According to one embodiment of this application, the power battery includes multiple cells, and the battery status data includes the voltage and temperature of each cell; the battery fault early warning model is used to determine the early warning level as the first early warning level when there is a cell whose temperature is greater than a temperature threshold and whose voltage change rate is greater than a voltage change rate threshold.

[0121] According to one embodiment of this application, the battery fault early warning model is further used to, when the early warning level is determined to be a first early warning level, determine the heat transfer relationship between cells in the power battery based on battery state data and thermal coupling coefficient; predict a set of target cells whose surface temperature will exceed a temperature threshold based on the first cell and the heat transfer relationship; and determine a solution strategy for the first early warning level based on the set of target cells; wherein, the first cell is the cell whose surface temperature exceeds the temperature threshold.

[0122] According to one embodiment of this application, the battery fault early warning model is further used to obtain a user behavior score when the temperature of a battery cell is less than or equal to a temperature threshold, or the voltage change rate is less than or equal to a voltage change rate threshold; wherein, the user behavior score is used to characterize the impact of user behavior on the health of the power battery, the battery fault early warning model includes a user behavior score model, the user behavior score model is used to obtain a user behavior score based on first information; when the user behavior score is greater than the user behavior score threshold, the early warning level is determined to be the third early warning level.

[0123] According to one embodiment of this application, the battery fault early warning model is further used to obtain the charging health of the power battery from the battery status data when the user behavior score is less than or equal to the user behavior score threshold; wherein, the charging health is the ratio of the current chargeable capacity to the rated capacity of the power battery, etc.; when the charging health of the power battery is greater than the charging health threshold, the early warning level is determined to be the second early warning level.

[0124] According to one embodiment of this application, the battery fault warning model is used to determine the solution strategy for the warning level as prompting the user to stop charging the vehicle's power battery when the warning level is the first warning level and the vehicle's power battery is being charged; or, the battery fault warning model is used to determine the solution strategy for the warning level as prompting the user to turn on the vehicle's ventilation system to cool the power battery and / or notifying the user to evacuate the vehicle when the warning level is the first warning level and the vehicle's power battery is being driven.

[0125] According to one embodiment of this application, the battery fault early warning model is used to determine the solution strategy for the warning level as prompting the user to use driving operations that reduce damage to the power battery when the warning level is the second warning level; the battery fault early warning model is used to determine the solution strategy for the warning level as prompting the user to use the target charging strategy when the warning level is the third warning level.

[0126] According to one embodiment of this application, a battery fault early warning model includes an adaptation model, a first battery fault early warning sub-model, and a second battery fault early warning sub-model. The adaptation model is used to determine the data integrity of first information. If the data integrity is greater than a first threshold, the early warning level and the solution strategy for the early warning level output by the first battery fault early warning sub-model are used as the early warning level and the solution strategy for the early warning level output by the battery fault early warning model. The first battery fault early warning sub-model is used to obtain the early warning level and the solution strategy for the early warning level based on the first information. If the data integrity is less than or equal to the first threshold, the early warning level and the solution strategy for the early warning level output by the second battery fault early warning sub-model are used as the early warning level and the solution strategy for the early warning level output by the battery fault early warning model. The second battery fault early warning sub-model is used to obtain the early warning level and the solution strategy for the early warning level based on the electrochemical parameters of the power battery.

[0127] This application also provides a computer-readable storage medium storing a computer program for executing the trailer hook control method provided in any of the above embodiments.

[0128] This application also provides a computer program product, which includes a computer program. When the computer program is executed by the processor of a computer device, the computer device is able to execute the power battery fault early warning method or model training method provided in any of the above embodiments.

[0129] All of the above-mentioned optional technical solutions can be combined in any way to form optional embodiments of this application, and will not be described in detail here.

[0130] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0131] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0132] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0133] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0134] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0135] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program verification codes, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0136] It should be noted that in the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0137] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0138] The above are merely preferred embodiments of this application and are not intended to limit this application. Any modifications or equivalent substitutions made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for early warning of power battery faults, characterized in that, The method is applied to a vehicle in which a power battery is installed, and the method includes: Obtain first information; wherein, the first information includes battery status data of the power battery, charging characteristic data of the power battery, and user behavior data, wherein the user behavior data includes user driving habit data and charging habit data; The first information is input into the battery fault early warning model to obtain the early warning level output by the battery fault early warning model and the solution strategy for the early warning level; wherein, the early warning level is used to characterize the severity of the danger of the power battery, the severity of the danger of the power battery is positively correlated with the probability of the power battery malfunctioning, the power battery includes multiple cells, and the battery status data includes the voltage and temperature of each cell; The battery fault early warning model is used to determine the early warning level as the first early warning level when a cell's temperature is greater than a temperature threshold and its voltage change rate is greater than a voltage change rate threshold. The battery fault early warning model is also used to obtain a user behavior score when the temperature of a battery cell is less than or equal to a temperature threshold, or the voltage change rate is less than or equal to a voltage change rate threshold; wherein, the user behavior score is used to characterize the impact of user behavior on the health of the power battery, the battery fault early warning model includes a user behavior scoring model, the user behavior scoring model is used to obtain a user behavior score based on first information; when the user behavior score is greater than the user behavior score threshold, the early warning level is determined to be the third early warning level; The battery fault early warning model is also used to obtain the charging health of the power battery from the battery status data when the user behavior score is less than or equal to the user behavior score threshold; wherein, the charging health is the ratio of the current chargeable capacity to the rated capacity of the power battery; and when the charging health of the power battery is greater than the charging health threshold, the early warning level is determined as the second early warning level. Display warning information; wherein, the warning information includes the warning level and the solution strategy.

2. The power battery fault early warning method according to claim 1, characterized in that, The battery fault early warning model is also used to determine the heat transfer relationship between cells in the power battery based on battery status data and thermal coupling coefficient when the early warning level is determined to be the first early warning level. Based on the first cell and the heat transfer relationship, a target cell set whose surface temperature will exceed the temperature threshold is predicted. The solution strategy for the first warning level is determined based on the target set of battery cells; wherein the first battery cell is a battery cell whose surface temperature exceeds a temperature threshold.

3. The power battery fault early warning method according to claim 1, characterized in that, The battery fault early warning model is used to determine the solution strategy for the warning level as prompting the user to stop charging the vehicle's power battery when the warning level is the first warning level and the vehicle's power battery is being charged. or; The battery fault early warning model is used to determine the solution strategy for the warning level when the warning level is the first warning level and the vehicle's power battery is in motion. The solution is to prompt the user to turn on the vehicle's ventilation system to cool the power battery and / or notify the user to evacuate the vehicle.

4. The power battery fault early warning method according to claim 1, characterized in that, The battery fault early warning model is used to determine the solution strategy for the warning level as prompting the user to use driving operations that reduce damage to the power battery when the warning level is the second warning level. The battery fault early warning model is used to determine the solution strategy for the warning level as prompting the user to use the target charging strategy when the warning level is the third warning level.

5. The power battery fault early warning method according to claim 1, characterized in that, The battery fault early warning model includes an adaptation model, a first battery fault early warning sub-model, and a second battery fault early warning sub-model. The adaptation model is used to determine the data integrity of the first information; wherein, the data integrity refers to the ratio between the quantity of the first information and the target total quantity of the first information; If the data integrity is greater than the first threshold, then the warning level output by the first battery fault warning sub-model and the solution strategy for the warning level are used as the warning level output by the battery fault warning model and the solution strategy for the warning level; wherein, the first battery fault warning sub-model is used to obtain the warning level and the solution strategy for the warning level based on the first information. If the data integrity is less than or equal to the first threshold, then the warning level output by the second battery fault warning sub-model and the solution strategy for the warning level are used as the warning level output by the battery fault warning model and the solution strategy for the warning level; wherein, the second battery fault warning sub-model is used to obtain the warning level and the solution strategy for the warning level based on the electrochemical parameters of the power battery.

6. A vehicle, characterized in that, include: processor; Memory used to store the processor's executable instructions. The processor is used to execute the power battery fault early warning method according to any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which is used to execute the power battery fault early warning method according to any one of claims 1 to 5.

Images