A method and device for evaluating battery health

Abstract

The present application discloses a method and a device for evaluating battery health. The method includes: classifying a plurality of batteries into N battery categories according to the magnitude of voltage change per unit time, and then combining the index data of the second single battery in the first battery category as the training data for the first prediction model, the second prediction model, and the third prediction model. By using the trained first prediction model, second prediction model, and third prediction model, the first health value, the second health value, and the preset health value of the first single battery are obtained. Based on the foregoing health values, the target health value of the first single battery at the first moment in the first working step is further obtained. The battery category of the second single battery is the first battery category. The first working step includes: a discharging working step or a charging working step. Whether the first single battery is abnormal at the first moment in the first working step is evaluated through the above target health value, so as to achieve fault diagnosis, prediction, and early warning of the battery.

Description

Technical Field

[0001] This application relates to the technical field of batteries, and particularly to a method and device for evaluating battery health. Background Art

[0002] With the rapid development of battery technology in recent years, batteries are widely used in new energy vehicles and consumer electronic products. Currently, the batteries on the market have poor safety and are prone to catching fire or exploding. Therefore, it is of great significance to analyze and detect the operating data of batteries to predict the performance degradation and future life of batteries.

[0003] In the prior art, the analysis of the health status of batteries often involves collecting the operating parameters of batteries through battery monitoring devices and transmitting the operating parameters to a display device for corresponding display, so that the staff can perform predictive analysis on the above operating parameters. However, the analysis efficiency is low. Summary of the Invention

[0004] To solve the above technical problems, this application provides a method and device for evaluating battery health. The method includes: after classifying batteries according to the magnitude of voltage change per unit time, combining the index data of the second single battery in the first battery category as the training data of the first prediction model, the second prediction model, and the third prediction model, obtaining the first health value, the second health value, and the preset health value of the first single battery through the trained first prediction model, the second prediction model, and the third prediction model, and obtaining the target health value of the first single battery at the first moment in the first working step to evaluate whether the first single battery is abnormal at the first moment in the first working step, so as to achieve fault diagnosis, prediction, and early warning of the battery.

[0005] In a first aspect, this application provides a method for evaluating battery health, including:

[0006] Obtaining the index data of N single batteries corresponding to N channels at each moment in the first working step, where the index data of the single battery at each moment in the first working step includes any one or more of the following: the voltage, current, internal resistance, pressure, temperature, and capacity of the single battery at each moment in the first working step; where one channel corresponds to one single battery, and the first working step includes: a discharging working step or a charging working step;

[0007] According to the voltage of the single battery at each moment in the first working step obtained, obtaining the magnitude of voltage change per unit time of the single battery in the first working step, and based on the magnitude of voltage change per unit time in the first working step, dividing the N single batteries into M different battery categories, where each battery category corresponds to one or more single batteries, M is less than or equal to N, and both M and N are positive integers;

[0008] Output the predicted value of the internal resistance of the first single battery at the first moment in the first working step through the trained first prediction model, and output the predicted value of the pressure of the first single battery at the first moment in the first working step through the trained second prediction model, where the first single battery is any single battery in the first battery category; the first battery category is any one of the M different battery categories;

[0009] Perform a ratio operation on the internal resistance of the first single battery at the first moment in the first working step and the predicted value of the internal resistance at the first moment to obtain a first ratio, and use the first ratio as the first health value of the first single battery. Perform a ratio operation on the pressure of the first single battery at the first moment in the first working step and the predicted value of the pressure at the first moment to obtain a second ratio, and use the second ratio as the second health value of the first single battery;

[0010] Perform calculation processing on the first health value, the second health value, and the preset health value of the first single battery obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model to obtain the target health value of the first single battery at the first moment in the first working step, so as to evaluate whether the first single battery is abnormal at the first moment in the first working step, where the target health value of the first single battery is used to indicate whether the first single battery is abnormal.

[0011] In an optional implementation manner, performing calculation processing on the first health value, the second health value, and the preset health value of the first single battery obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model to obtain the target health value of the first single battery at the first moment in the first working step specifically includes:

[0012] Set the weight of the first health value of the first single battery to a first value, set the weight of the second health value of the first single battery to a second value, and set the weight of the preset health value of the first single battery obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model to a third value, where the sum of the first value, the second value, and the third value is 1;

[0013] Perform a multiplication operation on the first health value and the first value to obtain a fourth value, perform a multiplication operation on the second health value and the second value to obtain a fifth value, and perform a multiplication operation on the preset health value and the third value to obtain a sixth value;

[0014] Use the sum of the fourth value, the fifth value, and the sixth value as the target health value of the first single battery at the first moment in the first working step.

[0015] In an alternative embodiment, after calculating and processing the first health value, the second health value, and the preset health value of the first single cell obtained based on the predicted value of the full discharge capacity of the first single cell at the first moment in the first working step output by the trained third prediction model to obtain the target health value of the first single cell at the first moment in the first working step, the method further includes:

[0016] Determine whether the target health value of the first single cell at the first moment in the first working step is within a preset range. If the target health value is not within the preset range, it is determined that the first single cell is abnormal at the first moment in the first working step, and an alarm is given and the circuit connected to the first single cell is controlled to be cut off.

[0017] In an alternative embodiment, before the trained first prediction model outputs the predicted value of the internal resistance of the first single cell at the first moment in the first working step, the method further includes:

[0018] Obtain first training data for training the first prediction model, where

[0019] the first training data includes any one or more of the following: the voltage and current of multiple second single cells at each moment in the charging working step, the voltage and current of multiple second single cells at each moment in the discharging working step, the voltage change amount within a preset time in the charging working step, the current within a preset time in the charging working step, the voltage change amount within a preset time in the discharging working step, and the current within a preset time in the discharging working step; where the battery categories of the second single cells are all the first battery category;

[0020] Input the first training data into the first prediction model for training to obtain the trained first prediction model;

[0021] Input the first moment in the first working step into the trained first prediction model to obtain the predicted value of the internal resistance of the first single cell at the first moment in the first working step.

[0022] In an alternative embodiment, before the trained second prediction model outputs the predicted value of the pressure of the first single cell at the first moment in the first working step, the method further includes:

[0023] Obtain second training data for training the second prediction model, where

[0024] The second training data includes any one or more of the following: the pressure of each of the plurality of second single cells at each moment during the charging step, and the pressure of each of the plurality of second single cells at each moment during the discharging step; wherein, the battery categories of the plurality of second single cells are respectively the first battery category;

[0025] Input the second training data into the second prediction model for training to obtain a trained second prediction model;

[0026] Input the first moment in the first step into the trained second prediction model to obtain a predicted value of the pressure of the first single cell at the first moment in the first step.

[0027] In an alternative implementation, before calculating and processing the first health value, the second health value, and the preset health value of the first single cell obtained based on the predicted value of the full discharge capacity of the first single cell at the first moment in the first step output by the trained third prediction model to obtain the target health value of the first single cell at the first moment in the first step, it further includes:

[0028] Based on the predicted value of the full discharge capacity of the first single cell at the first moment in the first step output by the trained third prediction model, perform a ratio operation on the predicted value of the full discharge capacity and the preset value of the full discharge capacity of the first single cell at the first moment in the first step to obtain the preset health value of the first single cell, wherein the preset value of the full discharge capacity of the first single cell at the first moment in the first step is set by the manufacturer of the first single cell when the first single cell leaves the factory.

[0029] In an alternative implementation, before the predicted value of the full discharge capacity of the first single cell at the first moment in the first step output by the trained third prediction model, it further includes:

[0030] Obtain the third training data for training the third prediction model, wherein,

[0031] The third training data includes: the discharge capacity and the discharge temperature corresponding to the discharge cut-off voltage moment of each of the plurality of second single cells; wherein, the battery categories of the plurality of second single cells are respectively the first battery category;

[0032] Input the third training data into the third prediction model for training to obtain a trained third prediction model;

[0033] Input the first moment in the first step into the trained third prediction model to obtain a predicted value of the full discharge capacity of the first single cell at the first moment in the first step.

[0034] In an alternative embodiment, dividing the N single cells into M different cell categories based on the magnitude of the voltage change per unit time in the first process step includes:

[0035] Classifying the N single cells based on the magnitude of the voltage change per unit time in the discharging process step to divide the N single cells into M different cell categories; or,

[0036] Classifying the N single cells based on the magnitude of the voltage change per unit time in the charging process step to divide the N single cells into M different cell categories.

[0037] In an alternative embodiment, the training data of the first prediction model includes: the index data of multiple second single cells under the first cell category; the training data of the second prediction model includes: the index data of multiple second single cells under the first cell category; or,

[0038] The training data of the third prediction model includes: the index data of multiple second single cells under the first cell category.

[0039] In a second aspect, the present application provides an evaluation device, including: a memory and a processor connected to the memory, the memory is used to store application program code, and the processor is configured to call the program code to execute the battery health evaluation method described in the first aspect.

[0040] The present application discloses a battery health evaluation method and an evaluation device, including: after classifying multiple batteries into N battery categories by the magnitude of the voltage change per unit time, combining the index data of the second single cells in the first battery category as the training data of the first prediction model, the second prediction model, and the third prediction model, and obtaining the first health value, the second health value, and the preset health value of the first single cell through the trained first prediction model, the second prediction model, and the third prediction model. Based on the foregoing health values, the target health value of the first single cell at the first moment in the first process step is further obtained, where the battery category of the second single cell is the first battery category; the first process step includes: a discharging process step or a charging process step. Evaluating whether the first single cell is abnormal at the first moment in the first process step through the above target health value can achieve fault diagnosis, prediction, and early warning of the battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] To more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0042] Figure 1 is a schematic flowchart of a method for evaluating battery health provided by the present application;

[0043] Figure 2 is a schematic structural diagram of an evaluation device provided by the present application. Detailed implementation manners

[0044] The following will clearly and completely describe the technical solutions in the present application with reference to the accompanying drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0045] It should be noted that the first, second, and third in the present application are only used to distinguish different prediction models, single cells, health values, training data, etc., and have no other special meanings, and should not limit the scope of protection of the present application.

[0046] Figure 1 is a schematic flowchart of a method for evaluating battery health provided by the present application. As Figure 1 shown, the method for evaluating battery health may include but is not limited to the following steps:

[0047] S101. Obtain the index data of N single cells corresponding to N channels at each moment in the first working step.

[0048] In the present application, the index data of the aforementioned single cells at each moment in the first working step may include but is not limited to any one or more of the following: the voltage of the single cell at each moment in the first working step, the current at each moment in the first working step, the internal resistance of the single cell at each moment in the first working step, the pressure of the single cell at each moment in the first working step, the temperature of the single cell at each moment in the first working step, the capacity of the single cell at each moment in the first working step;

[0049] Among them, one channel corresponds to one single cell, that is, one channel is used to charge or discharge one single cell; the first working step may include but is not limited to: a discharging working step, or a charging working step.

[0050] Among them, each moment in the first working step may be each moment during the discharging process, or each moment during the charging process;

[0051] Preferably, the internal resistance of the above single cell at each moment in the first step may include, but is not limited to, the DC internal resistance of the single cell at each moment in the first step.

[0052] It should be noted that the above channel is the channel for charging or discharging the single cell.

[0053] It should be noted that the index data of the N single cells at each moment in the first step is the index data of each single cell in the N single cells at each moment in the first step respectively.

[0054] Specifically, the device for obtaining the index data of the N single cells corresponding to the N channels at each moment in the first step may be a battery detection device.

[0055] It should be noted that the single cell may include, but is not limited to, a single cell applied or integrated in an unmanned aerial vehicle, a single cell applied or integrated in an electric vehicle, a single cell applied or integrated in a two-wheeled vehicle, or a single cell applied or integrated in consumer electronic products (such as notebooks, tablets, etc.).

[0056] S102. Obtain the magnitude of the voltage change per unit time of the single cell in the first step according to the voltage of the single cell at each moment in the first step, and divide the N single cells into M different battery categories based on the magnitude of the voltage change per unit time in the first step.

[0057] In this application, each battery category may respectively include one or more single cells, M is less than or equal to N, where M and N are both positive integers.

[0058] For example, the M different battery categories may include, but are not limited to: the first battery category, the second battery category,..., the (M - 1)th battery category, the Mth battery category.

[0059] For example, the first battery category may be the battery category to which the single cell belongs when the magnitude of the voltage change per unit time during the discharge process of the single cell is 500 mv - 600 mv; the second battery category may be the battery category to which the single cell belongs when the magnitude of the voltage change per unit time during the discharge process of the single cell is 600 mv - 700 mv.

[0060] It should be noted that when dividing the N single cells into M different battery categories, that is, each of the M battery categories may include one or more single cells, and the M battery categories in total include N single cells.

[0061] It should be noted that based on the voltages of the single cells at various moments in the first process step, the magnitude of the voltage change per unit time of the single cells in the first process step is obtained. Based on the magnitude of the voltage change per unit time in the first process step, the device that divides the N single cells into M different cell categories can be a battery detection device.

[0062] S103. Output the predicted value of the internal resistance of the first single cell at the first moment in the first process step through the trained first prediction model, and output the predicted value of the pressure of the first single cell at the first moment in the first process step through the trained second prediction model.

[0063] In this application, the first single cell is any single cell in the first cell category; the first cell category is any one of the M different cell categories;

[0064] The first moment in the first process step may include, but is not limited to: the first moment in the discharging process step, or the first moment in the charging process step;

[0065] The training data of the first prediction model includes: the index data of multiple second single cells in the first cell category; the training data of the first prediction model is also the first training data;

[0066] The training data of the second prediction model includes: the index data of multiple second single cells in the first cell category; the training data of the second prediction model is also the second training data.

[0067] The first moment in the first process step can be the first moment during the discharging process, and the first moment in the first process step can also be the first moment during the charging process;

[0068] It should also be noted that the first moment in the first process step may include, but is not limited to: the first moment in the discharging process step, or the first moment in the charging process step; that is to say, the first moment is any moment during the discharging process, or the first moment is any moment during the charging process;

[0069] It should be noted that before outputting the predicted value of the internal resistance of the first single cell at the first moment in the first process step through the trained first prediction model, the following steps may also be included, but are not limited to:

[0070] Obtain the first training data for training the first prediction model, where,

[0071] The first training data may include, but is not limited to, any one or more of the following: the voltage and current of each of the multiple second single cells at each moment during the charging step, the voltage and current of each of the multiple second single cells at each moment during the discharging step, the voltage change amount of each of the multiple second single cells within a preset time during the charging step, the current of each of the multiple second single cells within a preset time during the charging step, the voltage change amount of each of the multiple second single cells within a preset time during the discharging step, and the current of each of the multiple second single cells within a preset time during the discharging step; wherein, the battery categories of the second single cells are all the first battery category;

[0072] Input the first training data into the first prediction model for training to obtain a trained first prediction model;

[0073] Input the first moment in the first step into the trained first prediction model to obtain a predicted value of the internal resistance of the first single cell at the first moment in the first step.

[0074] It should be noted that the battery detection device can output a predicted value of the internal resistance of the first single cell at the first moment in the first step through the trained first prediction model, and output a predicted value of the pressure of the first single cell at the first moment in the first step through the trained second prediction model.

[0075] S104: Perform a ratio operation on the internal resistance of the first single cell at the first moment in the first step and the predicted value of the internal resistance at the first moment to obtain a first ratio, and use the first ratio as the first health value of the first single cell. Perform a ratio operation on the pressure of the first single cell at the first moment in the first step and the predicted value of the pressure at the first moment to obtain a second ratio, and use the second ratio as the second health value of the first single cell.

[0076] In this application, performing a ratio operation on the internal resistance of the first single cell at the first moment in the first step and the predicted value of the internal resistance at the first moment to obtain a first ratio may specifically include, but is not limited to:

[0077] Perform a ratio operation on the internal resistance of the first single cell at the first moment in the first step and the predicted value of the internal resistance of the first single cell at the first moment in the first step to obtain a first ratio.

[0078] Perform a ratio operation on the pressure of the first single cell at the first moment in the first step and the predicted value of the pressure at the first moment to obtain a second ratio, which may specifically include, but is not limited to:

[0079] Perform a ratio operation on the pressure of the first single cell at the first moment in the first process step and the predicted value of the pressure of the first single cell at the first moment in the first process step to obtain a second ratio.

[0080] In this application, after calculating and processing the first health value, the second health value, and the preset health value of the first single cell obtained according to the predicted value of the full discharge capacity of the first single cell at the first moment in the first process step output by the trained third prediction model to obtain the target health value of the first single cell at the first moment in the first process step, the following steps may be included but are not limited to:

[0081] Judge whether the target health value of the first single cell at the first moment in the first process step is within the preset range. If the target health value is not within the preset range, it is determined that the first single cell is abnormal at the first moment in the first process step, and an alarm is given and the circuit connected to the first single cell is controlled to be cut off. Or, if the target health value is not within the preset range, it is determined that the first single cell is abnormal, and an alarm is given and the first single cell is bypassed.

[0082] Preferably, the above preset range can be set by the user. For example, the preset range can be set to 0.7 - 1;

[0083] It should be noted that before the predicted value of the pressure of the first single cell at the first moment in the first process step is output by the trained second prediction model, the following steps may be included but are not limited to:

[0084] Obtain the second training data for training the second prediction model, where

[0085] The second training data may include but are not limited to any one or more of the following: the pressure of each second single cell at each moment in the charging process step among multiple second single cells, the pressure of each second single cell at each moment in the discharging process step among multiple second single cells; among them, the battery categories of the multiple second single cells are all the first battery category;

[0086] Input the second training data into the second prediction model for training to obtain the trained second prediction model;

[0087] Input the first moment in the first process step into the aforementioned trained second prediction model to obtain the predicted value of the pressure of the first single cell at the first moment in the first process step.

[0088] It should be noted that the battery detection device performs a ratio operation on the internal resistance of the first single battery at the first moment in the first process step and the predicted value of the internal resistance at the first moment, obtains a first ratio, and uses the first ratio as the first health value of the first single battery. The battery detection device also performs a ratio operation on the pressure of the first single battery at the first moment in the first process step and the predicted value of the pressure at the first moment, obtains a second ratio, and uses the second ratio as the second health value of the first single battery.

[0089] S105: Calculate and process the first health value, the second health value, and the preset health value of the first single battery obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first process step output by the trained third prediction model, to obtain the target health value of the first single battery at the first moment in the first process step, so as to evaluate whether the first single battery is abnormal at the first moment in the first process step.

[0090] In this application, the target health value of the first single battery is used to indicate whether the first single battery is abnormal; the training data of the third prediction model includes: index data of multiple second single batteries under the first battery category.

[0091] In this application, calculating and processing the first health value, the second health value, and the preset health value of the first single battery obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first process step output by the trained third prediction model to obtain the target health value of the first single battery at the first moment in the first process step may specifically include but is not limited to the following steps:

[0092] Step 1: Set the weight of the first health value of the first single battery to a first value, set the weight of the second health value of the first single battery to a second value, and set the weight of the preset health value of the first single battery obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first process step output by the trained third prediction model to a third value, where the sum of the first value, the second value, and the third value is 1;

[0093] Among them, the aforementioned preset health value is obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first process step output by the trained third prediction model. For the specific obtaining process, see the following text;

[0094] Step 2: Perform a multiplication operation on the first health value and the first value to obtain a fourth value, perform a multiplication operation on the second health value and the second value to obtain a fifth value, and perform a multiplication operation on the preset health value and the third value to obtain a sixth value;

[0095] Step 3: Use the sum of the fourth value, the fifth value, and the sixth value as the target health value of the first single battery at the first moment in the first process step.

[0096] Regarding the above steps 1-3, the first value is represented as x1, the second value is represented as x2, and the third value is represented as x3. That is, x1 + x2 + x3 = 1. The first health value is represented as s1, the second health value is represented as s2, and the preset health value is represented as s3. Then the fourth value can be represented as s1 * x1, the fifth value of the health value can be represented as s2 * x2, and the sixth value of the health value can be represented as s3 * x3. Taking the target health value represented as SOH as an example, the calculation process of SOH is as follows:

[0097] SOH = s1 * x1 + s2 * x2 + s3 * x3, where x1 + x2 + x3 = 1.

[0098] It should be noted that before calculating and processing the first health value, the second health value, and the preset health value of the first single battery obtained based on the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model to obtain the target health value of the first single battery at the first moment in the first working step, the following steps may also be included but are not limited to:

[0099] Based on the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model, perform a ratio operation on the obtained predicted value of the full discharge capacity and the preset value of the full discharge capacity of the first single battery at the first moment in the first working step to obtain the preset health value of the first single battery. Among them, the preset value of the full discharge capacity of the first single battery at the first moment in the first working step can be set by the manufacturer of the first single battery when the first single battery leaves the factory.

[0100] It should be noted that before the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model, the following steps may also be included but are not limited to:

[0101] Step 1: Obtain the third training data for training the third prediction model, where

[0102] the third training data includes: the discharge capacity and discharge temperature corresponding to each second single battery at the discharge cut-off voltage moment among multiple second single batteries; among them, the battery categories of the multiple second single batteries are all the first battery category;

[0103] Step 2: Input the third training data into the third prediction model for training to obtain the trained third prediction model;

[0104] Step 3: Input the first moment in the first working step into the trained third prediction model to obtain the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step.

[0105] For example, input the discharge cut-off voltage U1 into the third prediction model to obtain the discharge capacity C1 and the discharge temperature T1 corresponding to the discharge cut-off voltage U1; according to the temperature compensation formula, C11 = C1 - (T1 - C) * D, to obtain the temperature-compensated discharge capacity C11, and then according to the formula C22 = A * C11 + B, to obtain the predicted value of the full discharge capacity of the first single battery. Wherein, C is the average discharge temperature of the above-mentioned multiple second single batteries during the discharge process, D is the capacity difference of the above-mentioned multiple second single batteries compared with the average discharge temperature C; wherein, * represents the multiplication sign, C11 is the discharge capacity after temperature compensation, C22 is the predicted value of the predicted full discharge capacity, and A and B are constants, A represents the capacity prediction coefficient, and B is the capacity difference.

[0106] In the present application, based on the magnitude of the voltage change per unit time in the first working step, N single batteries are divided into M different battery categories, which may include but are not limited to the following process:

[0107] Based on the magnitude of the voltage change per unit time in the discharge working step, the N single batteries are classified based on the magnitude of the voltage change per unit time in the discharge working step, so as to divide the N single batteries into M different battery categories; or,

[0108] Based on the magnitude of the voltage change per unit time in the charging working step, the N single batteries are classified based on the magnitude of the voltage change per unit time in the charging working step, so as to divide the N single batteries into M different battery categories.

[0109] It should be noted that the battery detection device calculates and processes the first health value, the second health value, and the preset health value of the first single battery obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model, to obtain the target health value of the first single battery at the first moment in the first working step, so as to evaluate whether the first single battery is abnormal at the first moment in the first working step.

[0110] It should be noted that the devices for implementing the battery health evaluation method in the present application may include but are not limited to: battery detection devices, battery management devices, or other battery health monitoring devices.

[0111] Figure 1 It is only used to illustrate the embodiments of the present application and should not limit the protection scope of the present application.

[0112] The present application provides an evaluation device, Figure 2 which is a schematic structural diagram of the evaluation device provided by the present application. In the embodiments of the present application, Figure 2 the evaluation device therein may be an execution Figure 1A battery detection device, a battery management device, or other battery health monitoring devices for the evaluation method of battery health. Specifically,

[0113] such as Figure 2 shown, the evaluation device 20 may include but is not limited to: a processor 21 and a memory 22;

[0114] It should be noted that the memory 22 is coupled to the processor 21, and the memory 22 can be used to store Figure 1 the application code of the evaluation method of battery health in the described embodiment. The processor 21 is configured to call Figure 1 the application code of the evaluation method of battery health in the described embodiment to execute Figure 1 the evaluation method of battery health in the embodiment. Specifically,

[0115] The processor 21 can be used for:

[0116] Obtain the index data of N single cells corresponding to N channels at each moment in the first working step;

[0117] According to the voltages of the single cells at each moment in the first working step obtained, obtain the magnitude of the voltage change per unit time of the single cells in the first working step. Based on the magnitude of the voltage change per unit time in the first working step, divide the N single cells into M different battery categories;

[0118] Output the predicted value of the internal resistance of the first single cell at the first moment in the first working step through the trained first prediction model, and output the predicted value of the pressure of the first single cell at the first moment in the first working step through the trained second prediction model;

[0119] Perform a ratio operation on the internal resistance of the first single cell at the first moment in the first working step and the predicted value of the internal resistance at the first moment to obtain a first ratio, and use the first ratio as the first health value of the first single cell. Perform a ratio operation on the pressure of the first single cell at the first moment in the first working step and the predicted value of the pressure at the first moment to obtain a second ratio, and use the second ratio as the second health value of the first single cell;

[0120] Perform calculation processing on the first health value, the second health value, and the preset health value of the first single cell obtained according to the predicted value of the full discharge capacity of the first single cell at the first moment in the first working step output by the trained third prediction model to obtain the target health value of the first single cell at the first moment in the first working step, so as to evaluate whether the first single cell is abnormal at the first moment in the first working step.

[0121] It can be understood that for Figure 2 the specific implementation manner of the evaluation device 20, reference can be made toFigure 1 The method embodiments are not described in detail herein.

[0122] It should be understood that the evaluation device 20 is only an example provided in the embodiments of the present application, and the evaluation device 20 may have more or fewer components than those shown, two or more components may be combined, or different configurations of components may be implemented.

[0123] Those of ordinary skill in the art can realize that the devices and algorithm steps of each example described in combination with the embodiments disclosed in the present application can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0124] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the devices described above can refer to the corresponding processes in the foregoing method embodiments and will not be described in detail herein.

[0125] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the composition and steps of each example have been described. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0126] The embodiments of the method or device described above are merely illustrative. In addition, the displayed or discussed couplings or direct couplings or communication connections to each other may be indirect couplings or communication connections through some interfaces, devices, or units, and may also be electrical, mechanical, or other forms of connection.

[0127] Based on such understanding, the technical solution of the present application, in essence or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs that can store program codes.

[0128] As described above, the above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for evaluating battery health, characterized in that, Including: Obtaining index data of N single cells corresponding to N channels at each moment in a first working step, where the index data of the single cells at each moment in the first working step includes any one or more of the following: voltage, current, internal resistance, pressure, temperature, and capacity of the single cells at each moment in the first working step; wherein, one channel corresponds to one single cell, and the first working step includes: a discharging working step or a charging working step; Based on the voltage of the single cells at each moment in the first working step obtained, obtaining the magnitude of the voltage change per unit time of the single cells in the first working step, and based on the magnitude of the voltage change per unit time in the first working step, dividing the N single cells into M different cell categories, where each cell category corresponds to one or more single cells, M is less than or equal to N, and both M and N are positive integers; Outputting a predicted value of the internal resistance of a first single cell at a first moment in the first working step through a trained first prediction model, and outputting a predicted value of the pressure of the first single cell at the first moment in the first working step through a trained second prediction model, where the first single cell is any single cell in a first cell category; the first cell category is any one of the M different cell categories; Performing a ratio operation on the internal resistance of the first single cell at the first moment in the first working step and the predicted value of the internal resistance at the first moment to obtain a first ratio, and using the first ratio as the first health value of the first single cell, performing a ratio operation on the pressure of the first single cell at the first moment in the first working step and the predicted value of the pressure at the first moment to obtain a second ratio, and using the second ratio as the second health value of the first single cell; Performing calculation processing on the first health value, the second health value, and a preset health value of the first single cell obtained based on the predicted value of the full discharge capacity of the first single cell at the first moment in the first working step output by a trained third prediction model to obtain a target health value of the first single cell at the first moment in the first working step, so as to evaluate whether the first single cell is abnormal at the first moment in the first working step, where the target health value of the first single cell is used to indicate whether the first single cell is abnormal.

2. The battery health assessment method according to claim 1, characterized in that, Performing calculation processing on the first health value, the second health value, and a preset health value of the first single cell obtained based on the predicted value of the full discharge capacity of the first single cell at the first moment in the first working step output by a trained third prediction model to obtain a target health value of the first single cell at the first moment in the first working step, specifically including: Setting the weight of the first health value of the first single cell to a first value, setting the weight of the second health value of the first single cell to a second value, and setting the weight of the preset health value of the first single cell obtained based on the predicted value of the full discharge capacity of the first single cell at the first moment in the first working step output by a trained third prediction model to a third value, where the sum of the first value, the second value, and the third value is 1; Perform a multiplication operation on the first health value and the first value to obtain a fourth value, perform a multiplication operation on the second health value and the second value to obtain a fifth value, and perform a multiplication operation on the preset health value and the third value to obtain a sixth value; Use the sum of the fourth value, the fifth value, and the sixth value as the target health value of the first single battery at the first moment in the first working step.

3. The battery health evaluation method according to claim 2, wherein, After calculating and processing the first health value, the second health value, and the preset health value of the first single battery obtained according to the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model to obtain the target health value of the first single battery at the first moment in the first working step, it further includes: Judge whether the target health value of the first single battery at the first moment in the first working step is within the preset range. If the target health value is not within the preset range, determine that the first single battery is abnormal at the first moment in the first working step, and give an alarm and control to cut off the circuit connected to the first single battery.

4. The method for evaluating battery health according to claim 1, wherein, Before the trained first prediction model outputs the predicted value of the internal resistance of the first single battery at the first moment in the first working step, it further includes: Obtain the first training data for training the first prediction model, where the first training data includes any one or more of the following: the voltage and current of each of the multiple second single batteries at each moment in the charging working step, the voltage and current of each of the multiple second single batteries at each moment in the discharging working step, the voltage change amount of each of the multiple second single batteries within a preset time in the charging working step, the current of each of the multiple second single batteries within a preset time in the charging working step, the voltage change amount of each of the multiple second single batteries within a preset time in the discharging working step, and the current of each of the multiple second single batteries within a preset time in the discharging working step; where the battery categories of the second single batteries are all the first battery category; Input the first training data into the first prediction model for training to obtain the trained first prediction model; Input the first moment in the first working step into the trained first prediction model to obtain the predicted value of the internal resistance of the first single battery at the first moment in the first working step.

5. The battery health assessment method according to claim 1, wherein Before the trained second prediction model outputs the predicted value of the pressure of the first single battery at the first moment in the first working step, it further includes: Obtain the second training data for training the second prediction model, where the second training data includes any one or more of the following: the pressure of each of the multiple second single batteries at each moment in the charging working step, the pressure of each of the multiple second single batteries at each moment in the discharging working step; where the battery categories of the multiple second single batteries are all the first battery category; Input the second training data into the second prediction model for training to obtain the trained second prediction model; Input the first moment in the first working step into the trained second prediction model to obtain the predicted value of the pressure of the first single battery at the first moment in the first working step.

6. The method for evaluating battery health according to claim 1, wherein, Before calculating and processing the first health value, the second health value, and the preset health value of the first single battery obtained based on the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step output by the trained third prediction model to obtain the target health value of the first single battery at the first moment in the first working step, it further includes: Output the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step by the trained third prediction model, and perform a ratio operation on the predicted value of the full discharge capacity and the preset value of the full discharge capacity of the first single battery at the first moment in the first working step to obtain the preset health value of the first single battery, where the preset value of the full discharge capacity of the first single battery at the first moment in the first working step is set by the manufacturer of the first single battery when it leaves the factory.

7. The battery health assessment method according to claim 6, wherein Before outputting the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step by the trained third prediction model, it further includes: Obtain the third training data for training the third prediction model, where The third training data includes: the discharge capacity and discharge temperature corresponding to each of the multiple second single batteries at the discharge cut-off voltage moment; where the battery categories of the multiple second single batteries are all the first battery category; Input the third training data into the third prediction model for training to obtain the trained third prediction model; Input the first moment in the first working step into the trained third prediction model to obtain the predicted value of the full discharge capacity of the first single battery at the first moment in the first working step.

8. The method for evaluating battery health according to claim 1, wherein, The dividing the N single batteries into M different battery categories based on the magnitude of the voltage change per unit time in the first working step includes: Based on the magnitude of the voltage change per unit time in the discharge working step, classify the N single batteries based on the magnitude of the voltage change per unit time in the discharge working step to divide the N single batteries into M different battery categories; or, Based on the magnitude of the voltage change per unit time in the charging working step, classify the N single batteries based on the magnitude of the voltage change per unit time in the charging working step to divide the N single batteries into M different battery categories.

9. The method for evaluating battery health according to claim 1, wherein, The training data of the first prediction model includes: the index data of multiple second single batteries under the first battery category; the training data of the second prediction model includes: the index data of multiple second single batteries under the first battery category; or, The training data of the third prediction model includes: the index data of multiple second single batteries under the first battery category.

10. An evaluation device for battery health, characterized in that, Includes: A memory and a processor connected to the memory, the memory is used to store application program code, and the processor is configured to call the program code to execute the battery health assessment method according to any one of claims 1-9.

Images