A method and device for predicting battery life

Abstract

The present application discloses a method and device for predicting the life of a battery pack. In this method, the health factor of each battery cell in a plurality of battery cells is obtained during one or more first charge and discharge time periods; based on the health factor of each battery cell in the one or more first charge and discharge time periods, the health factor of each corresponding battery cell in a second charge and discharge time period is predicted; based on the health factors of the plurality of battery cells in the second charge and discharge time period, the battery capacity of the battery pack in the second charge and discharge time period is determined; and based on the battery capacity of the battery pack in the second charge and discharge time period, the remaining service life of the battery pack is determined. Because the present application predicts the battery capacity of a battery pack based on the health factors of a plurality of battery cells, it avoids the consistency problem of the battery cells, improves the accuracy of the predicted battery capacity, and thus can accurately predict the remaining service life.

Description

Technical Field

[0001] The present application relates to the field of battery technology, and in particular to a method and device for predicting the life of a battery pack. Background Art

[0002] The electrification of transportation is a major strategic requirement for energy reform and the development of smart, environmentally friendly, and energy-efficient travel in countries around the world. Energy storage systems, as a key component, have long constrained the development of electric vehicles, electric ships, and electric aircraft. Lithium-ion batteries, with their high energy, high power, and long life, are considered the optimal choice for energy storage systems. Typically, battery cells are not used directly in electric vehicles. Instead, multiple cells are connected in series or parallel to form a battery pack to meet the high-capacity and high-power requirements of electric vehicles or energy storage power stations. To reduce battery operating costs and prevent energy storage system failure or even serious accidents, accurate prediction of battery pack life is necessary to develop targeted battery maintenance strategies and reduce operating costs.

[0003] Currently, due to differences in the manufacturing process between battery cells in a battery pack, as well as differences in operating temperature, loading amplitude, and aging status during use, the remaining service life of each battery cell is also inconsistent. Consequently, it is impossible to predict the remaining service life of the battery pack by predicting the remaining service life of each battery cell. Therefore, a method for predicting the life of a battery pack is urgently needed. Summary of the Invention

[0004] The present application provides a battery pack life prediction method and device for predicting the remaining life of a battery pack.

[0005] In a first aspect, a battery pack life prediction method is provided, which is applied to a battery pack, wherein the battery pack includes a plurality of battery cells. The method is described below using a battery management module that predicts the battery pack life as an execution subject. The method includes:

[0006] The battery management module obtains the health factor of each battery cell in one or more first charge and discharge time periods among multiple battery cells; predicts the health factor of each corresponding battery cell in a second charge and discharge time period based on the health factor of each battery cell in the one or more first charge and discharge time periods; then, determines the battery capacity of the battery pack in the second charge and discharge time period based on the health factors of the multiple battery cells in the second charge and discharge time period; and determines the remaining service life of the battery pack based on the battery capacity of the battery pack in the second charge and discharge time period.

[0007] In this method, the battery management module predicts the health factor of each battery cell in a second charge / discharge period based on the acquired health factor of each battery cell in one or more first charge / discharge period. The battery management module then determines the battery capacity of the battery pack in the second charge / discharge period based on the health factor of each battery cell in the second charge / discharge period. Finally, the remaining useful life of the battery pack is determined using the battery capacity of the battery pack in the second charge / discharge period. This method fully utilizes the health factors of multiple battery cells in the battery pack, avoids interference from cell consistency issues on the battery capacity prediction results, and improves the accuracy of determining the battery capacity of the battery pack in the second charge / discharge period, thereby improving the accuracy of determining the remaining useful life of the battery pack.

[0008] Optionally, the first charge and discharge time period is a charge and discharge time period that the battery pack has experienced.

[0009] Optionally, the second charge and discharge time period is a charge and discharge time period that the battery pack is about to experience.

[0010] Optionally, the one or more first charge and discharge time periods and the second charge and discharge time period are two adjacent time periods.

[0011] In one possible design, when the first charge and discharge time period is the charge and discharge time period that the battery pack has experienced, the battery management module obtains the working data of each battery cell in one or more first charge and discharge time periods; and, based on the working data of each battery cell in each first charge and discharge time period, determines the health factor of each corresponding battery cell in each first charge and discharge time period.

[0012] Through this design, the battery management module determines the health factor of the corresponding battery cell in each first charge and discharge time period by obtaining the working data of each battery cell in one or more first charge and discharge time periods, making full use of the working data of multiple battery cells and avoiding the interference of the consistency problem of battery cells on the subsequent prediction of battery capacity.

[0013] In one possible design, the battery management module obtains the working data of each battery cell in one or more first charge and discharge time periods, including: obtaining the discharge data of each battery cell collected at multiple sampling moments in each of the one or more first charge and discharge time periods; the health factor includes the capacity increment curve peak, the power variance, and the power difference variance; the battery management module determines the battery capacity increment of each battery cell corresponding to the multiple sampling moments in each first charge and discharge time period based on the discharge data of each battery cell collected at the multiple sampling moments in each first charge and discharge time period; and the battery management module uses the maximum value of the battery capacity increments of each battery cell corresponding to the multiple moments in each first charge and discharge time period as the capacity increment curve peak of the corresponding battery cell in the corresponding first charge and discharge time period; and the battery management module determines the power variance and power difference variance of each battery cell in each first charge and discharge time period based on the discharge data of each battery cell collected at the multiple sampling moments in each first charge and discharge time period.

[0014] Through this design, the battery management module can determine the power variance, power difference variance and capacity increment curve peak of each corresponding battery cell in each first charge and discharge time period based on the discharge data corresponding to multiple sampling moments of each battery cell in each first charge and discharge time period.

[0015] In a possible design, the discharge data includes a voltage value and a current value corresponding to the voltage value.

[0016] In one possible design, the battery management module divides the voltage values ​​of each battery cell collected at multiple sampling moments within each first charge and discharge time period to obtain multiple voltage value sets within each first charge and discharge time period. The difference between the maximum voltage value and the minimum voltage value in any voltage value set is equal to a set value. Based on the multiple voltage value sets within each first charge and discharge time period, the battery management module determines multiple current value sets corresponding to the multiple voltage value sets within each first charge and discharge time period. For each corresponding voltage value set and current value set, one or more current values ​​included in the current value set correspond to one or more voltage values ​​included in the corresponding voltage value set. Based on each voltage value set within each first charge and discharge time period and the current value set corresponding to each voltage value set, the battery management module determines the charge value corresponding to each voltage value set within each first charge and discharge time period. Based on the charge value corresponding to each voltage value set within each first charge and discharge time period, the battery management module determines the charge variance and charge difference variance of each battery cell within each first charge and discharge time period.

[0017] Through this design, the battery management module divides the voltage values ​​in the discharge data of each battery cell collected at multiple sampling moments in each first charge and discharge time period to obtain multiple voltage value sets, and determines the charge value corresponding to each voltage value set based on each voltage value set and the current value set corresponding to each voltage value set, and obtains the charge variance and charge difference variance of each battery cell in each first charge and discharge time period, thereby realizing the extraction of health factors.

[0018] In one possible design, the battery management module determines the battery capacity increment of each battery cell in each first charge and discharge time period corresponding to the multiple sampling moments based on the voltage value and corresponding current value of each battery cell collected at the multiple sampling moments in each first charge and discharge time period.

[0019] Through this design, the battery management module determines the battery capacity increment of each battery cell corresponding to multiple sampling moments in each first charge and discharge time period based on the discharge data of each battery cell collected at multiple sampling moments in each first charge and discharge time period, providing a data basis for subsequently determining the peak value of the capacity increment curve of each battery cell in each first charge and discharge time period.

[0020] In one possible design, the battery management module obtains a set of health factors for each battery cell; the health factor set includes the health factor of each battery cell during the charge and discharge time period that the battery pack has experienced, and the health factor of each battery cell during a predicted charge and discharge time period that the battery pack has not experienced; and the battery management module selects, from the health factor set, the health factors during one or more charge and discharge time periods that are adjacent to the second charge and discharge time period as the health factors during one or more first charge and discharge time periods.

[0021] Through this design, the battery management module can select the health factors in one or more charging and discharging time periods that are adjacent to the second charging and discharging time period from the health factor set of each battery cell as the health factors in one or more first charging and discharging time periods, providing a data basis for subsequent prediction of the health factors of the battery cells in the second charging and discharging time period.

[0022] In one possible design, the battery management module inputs the health factor of each battery cell during one or more first charge and discharge time periods into a health factor attenuation model to obtain the corresponding health factor of each battery cell during the second charge and discharge time period. The one or more first charge and discharge time periods and the second charge and discharge time period are adjacent time periods.

[0023] Through this design, the battery management module can directly obtain the corresponding health factor of each battery cell in the second charge and discharge time period by inputting the health factor of each battery cell in one or more first charge and discharge time periods into the health factor attenuation model, thereby realizing rapid prediction of the health factor of the battery cell in the second charge and discharge time period.

[0024] In one possible design, the health factor attenuation model includes a double exponential sub-model and a neural network model; the battery management module inputs the number of charge and discharge times corresponding to the second charge and discharge time period into the double exponential sub-model to obtain the first candidate health factor of each corresponding battery cell in the second charge and discharge time period; wherein the number of charge and discharge times is the number of charge and discharge time periods experienced by the battery pack; the battery management module inputs the health factor of each battery cell in one or more of the first charge and discharge time periods into the neural network model to obtain the second candidate health factor of each battery cell in the second charge and discharge time period; based on the first candidate health factor and the second candidate health factor of each battery cell in the second charge and discharge time period, the health factor of each battery cell in the second charge and discharge time period is determined.

[0025] With this design, the battery management module obtains the first candidate health factor through the double exponential sub-model and the second candidate health factor through the neural network model, and then determines the health factor of the battery cell in the second charge and discharge time period based on the obtained first candidate health factor and the second candidate health factor, thereby improving the accuracy of the predicted health factor.

[0026] In one possible design, if the second candidate health factor is less than the product of the first candidate health factor and the set ratio, the battery management module uses the first candidate health factor as the health factor of each corresponding battery cell during the second charge and discharge time period; if the second candidate health factor is greater than or equal to the product of the first candidate health factor and the set ratio, the battery management module uses the second candidate health factor as the health factor of each corresponding battery cell during the second charge and discharge time period.

[0027] Through this design, the battery management module corrects the health factors predicted by the double exponential sub-model and the neural network model by comparing the second candidate health factor with the product of the first candidate health factor and the set ratio, thereby reducing the health factor prediction error and improving the health factor prediction accuracy.

[0028] In one possible design, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the third charge and discharge time period is the last charge and discharge time period experienced by the battery pack. The battery management module obtains the health factor of each battery cell during the third charge and discharge time period, uses the health factor of each battery cell during the third charge and discharge time period as a first sample health factor, and updates the corresponding health factor attenuation model based on the first sample health factor of each battery cell.

[0029] Through this design, the battery management module obtains the health factor of each battery cell in the third charge and discharge cycle, updates the corresponding health factor attenuation model, and can quickly complete the update of the health factor attenuation model, thereby improving the robustness of the health factor attenuation model, and further improving the accuracy of the health factor attenuation model in predicting the health factor of the battery cell in the charge and discharge time period.

[0030] In one possible design, the battery management module adjusts the double exponential sub-model based on the first sample health factor and the number of charge and discharge times corresponding to the third charge and discharge time period; inputs the first sample health factor into the neural network model, trains the fully connected layer of the neural network model, and updates the neural network model based on the training results.

[0031] With this design, the battery management module adjusts the dual exponential model based on the first sample health factor and the number of charges and discharges corresponding to the third charge and discharge time period. Furthermore, the battery management module trains the fully connected layer of the neural network model using the first sample health factor and updates the neural network model based on the training results, thereby rapidly updating the health factor decay model without the need to obtain the health factor of the second battery cell in real time to update the health factor decay model.

[0032] In one possible design, the battery management module inputs the health factors of the multiple battery cells during the second charge and discharge time period into a capacity estimation model to obtain a candidate battery capacity value for the battery pack during the second charge and discharge time period. If the candidate battery capacity value is less than or equal to a capacity threshold, the battery management module uses the candidate battery capacity value as the battery capacity of the battery pack during the second charge and discharge time period. If the candidate battery capacity value is greater than the capacity threshold, the battery management module continues to input the health factors of the multiple battery cells during the second charge and discharge time period into the capacity estimation model until the obtained candidate battery capacity value is less than or equal to the capacity threshold.

[0033] With this design, when the battery management module inputs the health factors of multiple battery cells in the second charge and discharge time period into the capacity estimation model, it obtains a candidate battery capacity value of the battery pack in the second charge and discharge time period; and the battery management module outputs the candidate battery capacity value as the battery capacity only when the candidate battery capacity value is less than or equal to the capacity threshold, thereby avoiding the need to predict the remaining service life of the battery pack when the battery capacity in the second charge and discharge time period is sufficient, thereby reducing the workload.

[0034] In one possible design, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the fourth charge and discharge time period is the last charge and discharge time period experienced by the battery pack. The battery management module determines the battery capacity of the battery pack during the fourth charge and discharge time period and the health factor of each battery cell during the fourth charge and discharge time period based on the operating data of each battery cell in the plurality of battery cells during the fourth charge and discharge time period. The battery management module uses the health factor of each battery cell during the fourth charge and discharge time period as a second sample health factor; and uses the battery capacity of the battery pack during the fourth charge and discharge time period as a sample battery capacity; and constructs the capacity estimation model based on the second sample health factor and the sample battery capacity.

[0035] Through this design, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the battery management module uses the health factor and battery capacity of the battery pack in the fourth charge and discharge cycle to construct a capacity estimation model. This achieves the rapid and accurate construction of a capacity estimation model suitable for the battery pack without the need to obtain the second sample health factor and sample battery capacity in real time, thereby improving the robustness of the capacity estimation model and further improving the accuracy of the capacity estimation model in predicting the battery capacity of the battery pack in the charge and discharge time period.

[0036] In one possible design, the battery management module obtains a target number of charge and discharge times; the target number of charge and discharge times is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge; determines the difference between the calibrated battery capacity of the battery pack and the battery capacity of the battery pack in the second charge and discharge time period; determines the ratio of the target number of charge and discharge times to the difference; and multiplies the ratio by the battery capacity of the battery pack in the second charge and discharge time period as the remaining service life of the battery pack.

[0037] With this design, the battery management module can accurately determine the remaining service life of the battery pack according to the target number of charge and discharge times and the predicted battery capacity of the battery pack in the second charge and discharge time period.

[0038] In one possible design, the battery management module obtains a target number of charge and discharge cycles; the target number of charge and discharge cycles is used to characterize the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge; determines the number of charge and discharge cycles corresponding to the second charge and discharge time period when the battery capacity of the battery pack in the second charge and discharge time period is equal to a minimum capacity threshold; and uses the charge and discharge cycles as the actual service life of the battery pack; and uses the difference between the actual service life and the target number of charge and discharge cycles as the remaining service life of the battery pack.

[0039] With this design, the battery management module can accurately determine the remaining service life of the battery pack by using the target charge and discharge times and the charge and discharge times corresponding to the second charge and discharge time period when the predicted battery capacity is equal to the minimum capacity threshold.

[0040] In a second aspect, a battery pack life prediction device is provided, which is applied to a battery pack, wherein the battery pack includes a plurality of battery cells; wherein the device includes:

[0041] an acquiring unit, configured to acquire a health factor of each battery cell of the plurality of battery cells during one or more first charge and discharge time periods; the health factor being used to characterize a capacity decay characteristic of the battery cell;

[0042] a first prediction unit, configured to predict a health factor of each corresponding battery cell in a second charge and discharge time period based on the health factor of each battery cell in one or more of the first charge and discharge time periods;

[0043] a second prediction unit, configured to determine a battery capacity of the battery pack in the second charge and discharge time period according to health factors of the plurality of battery cells in the second charge and discharge time period;

[0044] A determining unit is configured to determine the remaining service life of the battery pack according to the battery capacity of the battery pack in the second charge and discharge time period.

[0045] In one possible design, the first charge and discharge time period is a charge and discharge time period that the battery pack has experienced;

[0046] The acquisition unit is specifically configured to:

[0047] Acquire operating data of each battery cell during one or more of the first charge and discharge time periods;

[0048] According to the working data of each battery cell in each first charge and discharge time period, a health factor of each corresponding battery cell in each first charge and discharge time period is determined.

[0049] In one possible design, the health factor includes a capacity increment curve peak, a power variance, and a power difference variance; and the acquisition unit is specifically configured to:

[0050] Obtaining discharge data of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods within one or more of the first charge and discharge time periods;

[0051] determining, based on the discharge data of each battery cell collected at a plurality of sampling moments in each of the first charge and discharge time periods, a battery capacity increment of each battery cell corresponding to the plurality of sampling moments in each of the first charge and discharge time periods;

[0052] taking the maximum value of the battery capacity increments of each battery cell corresponding to the multiple moments in each of the first charge and discharge time periods as the peak value of the capacity increment curve of the corresponding battery cell in the corresponding first charge and discharge time period; and

[0053] According to the discharge data of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods, the power variance and power difference variance of each battery cell in each of the first charge and discharge time periods are determined.

[0054] In a possible design, the discharge data includes a voltage value and a current value corresponding to the voltage value.

[0055] In one possible design, the acquiring unit is specifically configured to:

[0056] The voltage values ​​of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods are divided to obtain multiple voltage value sets in each of the first charge and discharge time periods; the difference between the maximum voltage value and the minimum voltage value in any voltage value set is equal to the set value;

[0057] Determining, based on the multiple voltage value sets within each of the first charging and discharging time periods, multiple current value sets corresponding one-to-one to the multiple voltage value sets within each of the first charging and discharging time periods, wherein for a corresponding voltage value set and a corresponding current value set, one or more current values ​​included in the current value set are in one-to-one correspondence with one or more voltage values ​​included in the corresponding voltage value set;

[0058] Determining, according to each voltage value set and the current value set corresponding to each voltage value set in each first charging and discharging time period, a power value corresponding to each voltage value set in each first charging and discharging time period;

[0059] According to the power values ​​corresponding to each voltage value set in each first charge and discharge time period, the power variance and power difference variance of each battery cell in each first charge and discharge time period are determined.

[0060] In one possible design, the acquiring unit is specifically configured to:

[0061] According to the voltage value and corresponding current value of each battery cell collected at multiple sampling moments in each first charge and discharge time period, the battery capacity increment of each battery cell corresponding to the multiple sampling moments in each first charge and discharge time period is determined.

[0062] In one possible design, the acquiring unit is specifically configured to:

[0063] Obtaining a health factor set for each battery cell; the health factor set includes a health factor of each battery cell during a charge and discharge time period that the battery pack has experienced, and a health factor of each battery cell during a predicted charge and discharge time period that the battery pack has not experienced;

[0064] Health factors in one or more charging and discharging time periods that are adjacent to the second charging and discharging time period are selected from the health factor set as health factors in one or more first charging and discharging time periods.

[0065] In one possible design, the first prediction unit is specifically configured to:

[0066] The health factor of each battery cell in one or more of the first charge and discharge time periods is input into a health factor attenuation model to obtain the corresponding health factor of each battery cell in the second charge and discharge time period; the one or more first charge and discharge time periods and the second charge and discharge time period are adjacent time periods.

[0067] In one possible design, the health factor attenuation model includes a double exponential sub-model and a neural network model; and the first prediction unit is specifically configured to:

[0068] Inputting the number of charge and discharge times corresponding to the second charge and discharge time period into the double exponential sub-model to obtain a first candidate health factor for each battery cell in the second charge and discharge time period; the number of charge and discharge times is the number of charge and discharge time periods experienced by the battery pack;

[0069] Inputting the health factor of each battery cell in one or more of the first charge and discharge time periods into the neural network model to obtain a second candidate health factor of each corresponding battery cell in the second charge and discharge time period;

[0070] The health factor of each battery cell in the second charge and discharge time period is determined according to the first candidate health factor and the second candidate health factor of each battery cell in the second charge and discharge time period.

[0071] In one possible design, the first prediction unit is specifically configured to:

[0072] If the second candidate health factor is less than the product of the first candidate health factor and the set ratio, the first candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period;

[0073] If the second candidate health factor is greater than or equal to the product of the first candidate health factor and the set ratio, the second candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period.

[0074] In one possible design, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the third charge and discharge time period is the last charge and discharge time period experienced by the battery pack; and the acquisition unit is further configured to:

[0075] Obtaining a health factor of each battery cell during the third charge and discharge time period;

[0076] The first prediction unit is further configured to:

[0077] The health factor of each battery cell in the third charge and discharge time period is used as a first sample health factor, and the corresponding health factor attenuation model is updated according to the first sample health factor of each battery cell.

[0078] In one possible design, the first prediction unit is specifically configured to:

[0079] adjusting the double exponential sub-model according to the first sample health factor and the number of charge and discharge times corresponding to the third charge and discharge time period;

[0080] The first sample health factor is input into the neural network model, the fully connected layer of the neural network model is trained, and the neural network model is updated according to the training results.

[0081] In one possible design, the second prediction unit is specifically configured to:

[0082] inputting the health factors of the plurality of battery cells in the second charge and discharge time period into a capacity estimation model to obtain a candidate value of the battery capacity of the battery pack in the second charge and discharge time period;

[0083] If the candidate battery capacity value is less than or equal to the capacity threshold, using the candidate battery capacity value as the battery capacity of the battery pack in the second charge and discharge time period;

[0084] If the battery capacity candidate value is greater than the capacity threshold, the health factors of the multiple battery cells in the second charge and discharge time period are continuously input into the capacity estimation model until the obtained battery capacity candidate value is less than or equal to the capacity threshold.

[0085] In one possible design, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the fourth charge and discharge time period is the last charge and discharge time period experienced by the battery pack; and the acquiring unit is further configured to:

[0086] determining, based on operating data of each battery cell in the fourth charge and discharge time period, a battery capacity of the battery pack in the fourth charge and discharge time period and a health factor of each battery cell in the fourth charge and discharge time period;

[0087] The second prediction unit is further configured to:

[0088] using the health factor of each battery cell in the fourth charge and discharge time period as a second sample health factor; and using the battery capacity of the battery pack in the fourth charge and discharge time period as a sample battery capacity;

[0089] The capacity estimation model is constructed according to the second sample health factor and the sample battery capacity.

[0090] In one possible design, the determining unit is specifically configured to:

[0091] Obtaining a target charge and discharge number; the target charge and discharge number is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge;

[0092] determining a difference between a calibrated battery capacity of the battery pack and a battery capacity of the battery pack during the second charge and discharge time period;

[0093] Determining a ratio of the target charge and discharge times to the difference;

[0094] The product of the ratio and the battery capacity of the battery pack in the second charge and discharge time period is used as the remaining service life of the battery pack.

[0095] In one possible design, the determining unit is specifically configured to:

[0096] Obtaining a target charge and discharge number; the target charge and discharge number is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge;

[0097] determining the number of charge and discharge times corresponding to the second charge and discharge time period when the battery capacity of the battery pack in the second charge and discharge time period is equal to a minimum capacity threshold; and using the number of charge and discharge times as the actual service life of the battery pack;

[0098] The difference between the actual service life and the target charge and discharge times is used as the remaining service life of the battery pack.

[0099] In a third aspect, a battery module is provided, comprising a battery pack and the battery pack life prediction device described in the second aspect. When the battery pack is in operation, the battery pack life prediction device executes the method in the first aspect or any possible implementation of the first aspect.

[0100] In a fourth aspect, an energy storage system is provided, comprising a controller and the battery module described in the third aspect above, wherein when the controller controls the operation of the battery module, the battery module executes the method in the first aspect above or any possible implementation of the first aspect.

[0101] In a fifth aspect, the present application provides a computer-readable storage medium having a computer program or instruction stored thereon. When the computer program or instruction is executed, the computer executes the method in the above-mentioned first aspect or any possible implementation of the first aspect.

[0102] In a sixth aspect, the present application provides a computer program product. When a computer executes the computer program product, the computer executes the method in the above-mentioned first aspect or any possible implementation of the first aspect.

[0103] For the beneficial effects of the second to fourth aspects mentioned above, please refer to the description of the beneficial effects of the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0104] Figure 1 A schematic diagram of the architecture of a possible battery pack for the solution provided in an embodiment of the present application;

[0105] Figure 2 A schematic diagram of the distribution of battery cells in a possible battery pack of the solution provided in an embodiment of the present application;

[0106] Figure 3 A flow chart of a possible battery life prediction method for the solution provided in an embodiment of the present application;

[0107] Figure 4A flow chart of a possible power variance and a method for determining power difference variance of the solution provided in an embodiment of the present application;

[0108] Figure 5 A schematic diagram of the structure of a possible health factor attenuation model for the solution provided in an embodiment of the present application;

[0109] Figure 6 A schematic diagram of a possible update process of a health factor attenuation model for the solution provided in an embodiment of the present application;

[0110] Figure 7 A flowchart illustrating a possible training process of a health factor attenuation model for the solution provided in an embodiment of the present application;

[0111] Figure 8 A schematic diagram of a possible battery charging and discharging strategy for the solution provided in an embodiment of the present application;

[0112] Figure 9 A schematic diagram of a possible health factor fusion scheme provided in an embodiment of the present application;

[0113] Figure 10 A flowchart of a possible battery capacity prediction method for the solution provided in an embodiment of the present application;

[0114] Figure 11 A complete flowchart of a possible battery pack life prediction method for the solution provided in an embodiment of the present application;

[0115] Figure 12 A complete flowchart of another possible battery pack life prediction method according to the solution provided in the embodiment of the present application;

[0116] Figure 13 A schematic diagram of the structure of a possible battery pack life prediction system for the solution provided in an embodiment of the present application;

[0117] Figure 14 A schematic diagram of the structure of a battery pack life prediction device according to an embodiment of the present application;

[0118] Figure 15 A schematic structural diagram of a battery module according to an embodiment of the present application;

[0119] Figure 16 A schematic diagram of the architecture of a possible energy storage system for the solution provided in an embodiment of the present application. DETAILED DESCRIPTION

[0120] The present invention provides a method and device for predicting battery life. The method and device are based on the same concept. Since the method and device solve similar problems, the implementation of the device and method can refer to each other, and the repeated parts will not be repeated.

[0121] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the embodiments of the present application will be further described in detail below with reference to the accompanying drawings. In the description of the embodiments of the present application, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features.

[0122] To facilitate understanding, exemplary descriptions of concepts related to this application are provided for reference.

[0123] 1) Remaining useful life (RUL) primarily refers to the remaining useful life of a system after a period of operation. Accurately predicting a system's RUL can significantly reduce losses caused by system downtime and improve system reliability. Furthermore, when the RUL reaches a lower limit, the user can be alerted so that they can take timely measures to ensure safety.

[0124] 2) The ampere-hour integration method does not consider the internal mechanism of the battery. It estimates the battery's state of charge by integrating time and current based on certain external characteristics of the system, such as current, time, and temperature compensation, and sometimes adding certain compensation coefficients.

[0125] 3) The least squares method is a mathematical optimization technique. It finds the best function matching the data by minimizing the sum of squared errors. Least squares can be used to easily obtain unknown data and minimize the sum of squared errors between the obtained data and the actual data. Least squares can also be used for curve fitting. Other optimization problems can also be formulated using least squares by minimizing energy or maximizing entropy.

[0126] 4) Long short term memory (LSTM) is a special type of recurrent neural network (RNN) that can learn long-term dependency information.

[0127] 5) Gaussian process regression (GPR) is a non-parametric model that uses Gaussian process (GP) priors to perform regression analysis on data.

[0128] 6) The charge and discharge time period refers to the time period consisting of the fully charged charging time and the fully discharged discharge time of the battery pack.

[0129] In the description of the embodiments of the present application, "and / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent three situations: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. At least one referred to in this application refers to one or more; multiple refers to two or more. In addition, it should be understood that in the description of this application, words such as "first" and "second" are only used for the purpose of distinguishing the description, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying order. The embodiments of the present application will be described in detail below in conjunction with the accompanying drawings.

[0130] Currently, due to differences in the manufacturing process between battery cells in a battery pack, as well as differences in operating temperature, loading amplitude, and aging status during use, the remaining service life of each battery cell is also inconsistent. Consequently, it is impossible to predict the remaining service life of the battery pack by predicting the remaining service life of each battery cell. Therefore, a method for predicting the life of a battery pack is urgently needed.

[0131] In order to predict the remaining service life of a battery pack, an embodiment of the present application provides a battery pack life prediction method, which is applied to a battery pack, wherein the battery pack includes a plurality of battery cells. The following description is made using a battery management module that predicts the life of a battery pack as the execution body: In this method, the battery management module obtains the health factor of each battery cell in one or more first charge and discharge time periods among a plurality of battery cells; wherein the health factor is used to characterize the capacity attenuation characteristics of the battery cell. The battery management module predicts the health factor of each corresponding battery cell in a second charge and discharge time period based on the health factor of each battery cell in one or more first charge and discharge time periods. After determining the battery capacity of the battery pack in the second charge and discharge time period based on the health factors of the plurality of battery cells in the second charge and discharge time period, the battery management module determines the remaining service life of the battery pack based on the battery capacity of the battery pack in the second charge and discharge time period.

[0132] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0133] Figure 1 A schematic diagram of a possible battery pack architecture applicable to the solution provided in the embodiment of the present application. Figure 1 As shown, the battery pack includes a battery management module and a plurality of battery cells. The battery cells can be connected in parallel or in series, and the battery cells are composed of at least one battery cell connected in series. For example, Figure 2 As shown, a battery cell is composed of 3 battery cells connected in series, and a battery pack is composed of 4 battery cells connected in parallel.

[0134] The battery management module may include a processor and a memory, wherein the memory is used to store the health factor of each battery cell in the battery pack during a previously experienced charge and discharge period or a predicted health factor during a previously experienced charge and discharge period.

[0135] The processor obtains the health factor of each battery cell in one or more charge and discharge time periods from the memory.

[0136] In some possible implementations, the processor determines the health factor of each battery cell in the next charge / discharge period following the health factor of each battery cell in one or more charge / discharge period. The processor then determines the battery capacity of the battery pack in the next charge / discharge period based on the health factors of the multiple battery cells in the next charge / discharge period, and further determines the remaining service life of the battery pack based on the obtained battery capacity.

[0137] In other embodiments, the battery management module may also be located outside the battery pack to manage the battery pack according to a configured battery management strategy.

[0138] The solution provided in this application is introduced below with reference to specific embodiments.

[0139] Figure 3 A schematic diagram of a battery pack life prediction method provided in an embodiment of the present application is applied to a battery pack including a plurality of battery cells. The following description is based on a battery management module that predicts the battery pack life as the execution body; wherein the battery management module can be located inside the battery module, such as Figure 1 The battery management module in the battery pack shown in FIG. 1 may also be located outside the battery pack. Figure 3 As shown, the method includes:

[0140] S301: A battery management module obtains a health factor of each battery cell among a plurality of battery cells in one or more first charge and discharge time periods.

[0141] It should be noted that the health factor is used to characterize the capacity decay characteristics of battery cells.

[0142] In some optional embodiments, the battery management module may obtain the health factors of one or more charging and discharging time periods from the charging and discharging time periods that the battery pack has experienced as the health factors of one or more first charging and discharging time periods; or, the battery management module may also obtain the health factors of one or more first charging and discharging time periods from the health factor set.

[0143] The following describes how the battery management module obtains the health factors in one or more first charge and discharge time periods in the above two ways respectively.

[0144] Method 1: The battery management module obtains health factors in one or more charge and discharge time periods from the charge and discharge time periods that the battery pack has experienced as health factors in one or more first charge and discharge time periods.

[0145] It should be noted that the first charge and discharge time period is the charge and discharge time period that the battery pack has experienced.

[0146] The battery management module obtains the operating data of each battery cell in one or more first charge and discharge time periods, and determines the health factor of each corresponding battery cell in each first charge and discharge time period based on the operating data of each battery cell in each first charge and discharge time period.

[0147] It should be noted that the operating data in the first charge and discharge time period includes but is not limited to discharge data collected at multiple sampling moments in the first charge and discharge time period.

[0148] Optionally, the battery management module obtains discharge data of each battery cell collected at multiple sampling moments in each of the one or more first charge and discharge time periods.

[0149] In the embodiment of the present application, the health factor may include a capacity increment curve peak, a power variance, and a power difference variance.

[0150] In some optional implementations, the battery management module may determine the health factor of each battery cell in each first charge and discharge time period through the following steps:

[0151] a1: The battery management module determines the power variance and power difference variance of each battery cell in each first charge and discharge time period based on the discharge data of each battery cell collected at multiple sampling moments in each first charge and discharge time period.

[0152] The discharge data includes a voltage value and a current value corresponding to the voltage value.

[0153] It should be noted that the voltage and current values ​​in the discharge data are the voltage and current values ​​output by the battery cell at the sampling moment.

[0154] Optional, such as Figure 4 As shown, the battery management module can determine the power variance and power difference variance of each battery cell in each first charge and discharge time period through the following steps:

[0155] S401: The battery management module divides the voltage value of each battery cell collected at multiple sampling moments in each first charge and discharge time period to obtain multiple voltage value sets in each first charge and discharge time period.

[0156] It should be noted that the difference between the maximum voltage value and the minimum voltage value in any voltage value set is equal to the set value.

[0157] For the convenience of the following description, taking one battery cell among the multiple battery cells as an example, obtaining multiple voltage value sets of the battery cell in a first charge and discharge time period is described.

[0158] The battery management module obtains the voltage values ​​in the discharge data of the battery cell collected at multiple sampling moments in the first charge and discharge time period from the working data of the battery cell, and then divides the voltage values ​​according to the order of the sampling moments according to the set value to obtain multiple voltage value sets.

[0159] For example, the set value is 4V, and according to the order of sampling time, the voltage values ​​are 10V, 8V, 6V, and 2V respectively. Then the voltage values ​​are divided, and the obtained voltage value sets are {10V, 8V, 6V} and {6V, 2V} respectively.

[0160] Optionally, when the voltage value needs to be divided into N voltage value sets, the set value can be determined by the following formula:

[0161] ΔV=(V max -V min ) / N

[0162] Among them, ΔV represents the set value, V max Indicates the maximum voltage value in the discharge data collected at multiple sampling moments, V min represents the minimum value of the voltage values ​​in the discharge data collected at multiple sampling moments, and N represents the number of voltage value sets.

[0163] S402: The battery management module determines, based on the multiple voltage value sets in each first charge and discharge time period, multiple current value sets corresponding one-to-one to the multiple voltage value sets in each first charge and discharge time period.

[0164] It should be noted that, for a corresponding voltage value set and a corresponding current value set, one or more current values ​​included in the current value set are in one-to-one correspondence with one or more voltage values ​​included in the corresponding voltage value set.

[0165] For a corresponding voltage value set and current value set, for each sampling moment corresponding to a voltage value included in the voltage value set, a corresponding current value can be found in the current value set at the same sampling moment as the voltage value. Furthermore, there is only one current value in the current value set that has the same sampling moment as the voltage value in the voltage value set.

[0166] S403: The battery management module determines the power value corresponding to each voltage value set in each first charge and discharge time period according to each voltage value set and the current value set corresponding to each voltage value set in each first charge and discharge time period.

[0167] Optionally, the battery management module may determine the power value corresponding to each voltage value set in each first charge and discharge time period by an ampere-hour integration method to obtain a power value sequence in each first charge and discharge time period.

[0168] For example, for a voltage value set within a first charging and discharging time period, the battery management module arranges the voltage value set from large to small according to the maximum voltage value in the voltage value set to obtain a voltage value set sequence; the battery management module determines the power value sequence corresponding to the voltage value set sequence based on the ampere-hour integration method.

[0169] S404 , the battery management module determines the power variance and power difference variance of each battery cell in each first charge and discharge time period according to the power value corresponding to each voltage value set in each first charge and discharge time period.

[0170] With respect to the power value in a first charge and discharge time period, the battery management module may determine the power variance and power difference variance of the battery cells in the first charge and discharge time period in the following manner.

[0171] Power value sequence Q i (V) = [Q1, Q2, ..., Q N ], the battery management module will Q i (V) Subtract the total charge Q1 to get dQ i (V) a sequence, wherein:

[0172] dQ i (V)=[Q1-Q1, Q2-Q1,···,Q N -Q1]

[0173] Where i represents the total number of charge and discharge time periods that the battery pack has experienced; Q represents the charge level.

[0174] The battery management module gets dQ i (V) sequence, dQ i (V) Calculate the variance of the sequence and obtain the variance of the electricity quantity.

[0175] The battery management module will dQ i (V) sequence and dQ obtained in the first charge and discharge time period 1 (V) sequence difference, to obtain the ΔdQ in the first charge and discharge time period i sequence;

[0176] Where ΔdQ i =dQ i -dQ 1

[0177] The battery management module obtains ΔdQ i After the sequence, ΔdQ i Calculate the variance of the sequence and obtain the variance of the power difference.

[0178] a2: The battery management module determines the battery capacity increment of each battery cell corresponding to the multiple sampling moments in each first charge and discharge time period based on the discharge data of each battery cell collected at the multiple sampling moments in each first charge and discharge time period.

[0179] It should be noted that there is no specific order in which step a1 and step a2 are performed.

[0180] Multiple sampling moments can be set within the first charge-discharge time period, and discharge data of multiple battery cells can be collected at each of the multiple sampling moments. Furthermore, the discharge data includes voltage and current values. That is, the voltage and current values ​​of the battery cells are present at each sampling moment within the first charge-discharge time period.

[0181] Optionally, the battery management module determines the battery capacity increment of each battery cell corresponding to multiple sampling moments in each first charge and discharge time period based on the voltage value and corresponding current value of each battery cell collected at multiple sampling moments in each first charge and discharge time period.

[0182] In some optional implementations, the battery management module may determine the battery capacity increment using the following formula:

[0183]

[0184] Among them, (f -1)' represents the battery capacity increment, Q represents the battery capacity, I represents the current value, V represents the voltage value, and t represents the sampling time.

[0185] a3: The battery management module uses the maximum value of the battery capacity increments corresponding to multiple moments of each battery cell in each first charge and discharge time period as the capacity increment curve peak value of the corresponding battery cell in the corresponding first charge and discharge time period.

[0186] Method 2: The battery management module may obtain one or more health factors within the first charge and discharge time period from the health factor set.

[0187] In some optional embodiments, the battery management module obtains a set of health factors for each battery cell, and selects one or more health factors within a charging and discharging time period that are adjacent to the second charging and discharging time period from the health factor set as health factors within one or more first charging and discharging time periods.

[0188] It should be noted that the health factor set includes the health factor of each battery cell in the charge and discharge time period that the battery pack has experienced, and the health factor of each battery cell in the predicted charge and discharge time period that the battery pack has not experienced.

[0189] For example, the health factor set includes health factors A, B, C, D, and E; wherein A, B, and C are health factors within the charge and discharge time period that the battery pack has experienced, wherein the order of A, B, and C is arranged according to the order in which the battery pack has experienced its corresponding charge and discharge time period; D and E are health factors of the battery cells within the predicted charge and discharge time period that the battery pack has not experienced; wherein the number of charges and discharges corresponding to the charge and discharge time period corresponding to E is lower than the number of charges and discharges corresponding to the charge and discharge time period corresponding to D. When the battery management module obtains a health factor within a first charge and discharge time period, since the charge and discharge time period corresponding to E is adjacent to the second charge and discharge time period, E is used as the health factor within the first charge and discharge time period. If the battery management module obtains health factors within two first charge and discharge time periods, and the charge and discharge time periods corresponding to D and E are adjacent to the second charge and discharge time period, D and E are used as the health factors within the first charge and discharge time period.

[0190] S302 : The battery management module predicts the health factor of each battery cell in a second charge and discharge time period based on the health factor of each battery cell in one or more first charge and discharge time periods.

[0191] Optionally, the battery management module inputs the health factor of each battery cell in one or more first charge and discharge time periods into the health factor attenuation model to obtain the corresponding health factor of each battery cell in the second charge and discharge time period.

[0192] Optionally, since the health factor may include charge difference variance, charge variance, and capacity increment curve peak, each health factor corresponds to a health factor attenuation model, then one battery cell corresponds to three health factor attenuation models, and each health factor attenuation model corresponds to one health factor.

[0193] In some optional embodiments, such as Figure 5 As shown, the present application provides a structural schematic diagram of a health factor attenuation model. The health factor attenuation model may include a double exponential sub-model and a neural network model.

[0194] Optionally, the battery management module may predict the health factor of each battery cell during the second charge and discharge time period through the following steps:

[0195] b1: The battery management module inputs the number of charge and discharge times corresponding to the second charge and discharge time period into the double exponential sub-model to obtain a first candidate health factor for each corresponding battery cell in the second charge and discharge time period.

[0196] Among them, the mathematical expression of the double exponential submodel is as follows:

[0197] y=ae bx +ce dx

[0198] Where y represents the health factor, a, b, c, and d represent the model parameters of the double exponential sub-model, and x represents the number of charge and discharge times.

[0199] b2: The battery management module inputs the health factor of each battery cell in one or more first charge and discharge time periods into the neural network model to obtain a second candidate health factor of each corresponding battery cell in a second charge and discharge time period.

[0200] Among them, the neural network model can include an LSTM layer and a fully connected layer.

[0201] Optionally, the battery management module inputs the health factor of the battery cell in one or more first charge and discharge time periods into the neural network model in the health factor attenuation model corresponding to the battery cell to obtain a second candidate health factor of the battery cell in the second charge and discharge time period.

[0202] In some optional embodiments, the battery management module inputs one or more health factors during the first charge and discharge time period into an LSTM layer in the neural network model. The LSTM layer calculates the health factors and obtains an output vector. The battery management module then inputs the obtained output vector into a fully connected layer to obtain a second candidate health factor. The output vector is the hidden state result corresponding to the health factor obtained by the LSTM layer.

[0203] For example, when the number of health factors in the first charge and discharge time period is 1, the window length of the LSTM layer is 1. The LSTM layer takes the previous hidden state result (health factor in the first charge and discharge time period) before the time step (the second charge and discharge time period) as input and inputs it into the LSTM layer to obtain the second candidate health factor of the battery cell in the second charge and discharge time period.

[0204] For example, when the window length of the LSTM layer is 30, the first 30 hidden state results before the time step are used as input to the LSTM layer. The battery management module uses the health factor power variance stdQ, the number of charge and discharge times m corresponding to the second charge and discharge time period, and [HI m-30 , HI m-29 ,···,HI m-1 ] is input into the LSTM layer of the neural network model, and the LSTM layer performs the calculation of stdQ and [HI m-30 , HI m-29 ,···,HI m-1 ] to calculate and get the output vector HI m At this time, the window [HI m-30 , HI m-29 ,···,HI m-1 ]Move right one position to get a new window [HI m-29 , HI m-28 ,···,HI m ], where the new window [HI m-29 , HI m-28 ,···,HI m ] is used to predict HI m+1 .

[0205] It should be noted that there is no particular order between step b1 and step b2.

[0206] b3: The battery management module determines the health factor of each battery cell in the second charge and discharge time period according to the first candidate health factor and the second candidate health factor of each battery cell in the second charge and discharge time period.

[0207] Optionally, after obtaining the first candidate health factor and the second candidate health factor, the battery management module determines whether the second candidate health factor is less than the product of the first candidate health factor and a set ratio. The set ratio is less than 1, for example, the set ratio can be set to 95%, 90%, or 80%.

[0208] In some optional embodiments, if the second candidate health factor is less than the product of the first candidate health factor and the set ratio, the first candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period.

[0209] In some other optional embodiments, if the second candidate health factor is greater than or equal to the product of the first candidate health factor and the set ratio, the second candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period.

[0210] In some optional embodiments, before the battery management module executes the remaining service life prediction process of the battery pack, the battery management module needs to update the health factor attenuation model.

[0211] Optional, such as Figure 6 As shown, an embodiment of the present application provides a method for updating a health factor attenuation model, wherein the battery management module can update the health factor attenuation model through, but is not limited to, the following steps:

[0212] S601: The battery management module obtains a health factor of each battery cell in a third charge and discharge time period.

[0213] It should be noted that, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the third charge and discharge time period is the last charge and discharge time period among the charge and discharge time periods experienced by the battery pack.

[0214] The process of the battery management module obtaining the health factor of each battery cell in the third charge and discharge time period is the same as the process of determining the health factor of each battery cell in the first charge and discharge time period, and will not be repeated here, wherein the first charge and discharge time period is the charge and discharge time period that the battery pack has experienced.

[0215] S602: The battery management module uses the health factor of each battery cell in the third charge and discharge time period as a first sample health factor.

[0216] The steps of the battery management module updating the corresponding health factor attenuation model according to the first sample health factor of each battery cell include:

[0217] S603: The battery management module adjusts the corresponding double-exponential sub-model according to the first sample health factor and the number of charge and discharge times corresponding to the third charge and discharge time period.

[0218] Optionally, after the battery management module inputs the first sample health factor and the number of charge and discharge times corresponding to the third charge and discharge time period into the double exponential sub-model, it can identify the model parameters of the double exponential sub-model through the least squares method and adjust the double exponential sub-model according to the obtained model parameters.

[0219] The model parameters can be determined by the following formula:

[0220]

[0221] Among them, x represents the number of charge and discharge times corresponding to the charge and discharge time period, y represents the health factor, and ω represents the model parameter.

[0222] S604: The battery management module inputs the first sample health factor into the fully connected layer of the corresponding neural network model, trains the fully connected layer, and updates the corresponding neural network model according to the training results.

[0223] It should be noted that S603 and S604 are executed simultaneously, regardless of the order.

[0224] Optionally, the battery management module freezes the LSTM layer in the neural network model, inputs the first sample health factor into the neural network model, retrains the fully connected layer in the neural network model, and updates the neural network model according to the training results.

[0225] The battery management module can directly update the health factor attenuation model corresponding to each battery cell based on the first sample health factor of each battery cell in the third charge and discharge time period, without the need to obtain the health factor in real time. While achieving the purpose of quickly updating the health factor attenuation model, it also improves the robustness of the health factor attenuation model.

[0226] S605 : The battery management module inputs the number of charge and discharge times corresponding to the second charge and discharge time period into the double exponential sub-model to obtain a first candidate health factor of each battery cell in the second charge and discharge time period.

[0227] S606: The battery management module inputs the health factor of each battery cell in one or more first charge and discharge time periods into the neural network model to obtain a second candidate health factor of each battery cell in a second charge and discharge time period.

[0228] S605 and S606 are executed simultaneously, regardless of the order.

[0229] S607 : The battery management module determines the health factor of each battery cell in the second charge and discharge time period according to the first candidate health factor and the second candidate health factor of each battery cell in the second charge and discharge time period.

[0230] In some optional embodiments, after the battery management module starts working, it is necessary to update the health factor attenuation model corresponding to the battery cell based on the first sample health factor of each battery cell, so as to increase the adaptability of the health factor attenuation model to the battery cell, thereby improving the prediction accuracy of the health factor.

[0231] Optionally, the battery management module may regularly update the health factor attenuation model according to a set update cycle, so that the health factor attenuation model is more suitable for the battery cell and the prediction accuracy of the health factor is ensured.

[0232] For example, the update cycle may be set to update the health factor attenuation model every time the battery pack has experienced 200 charge and discharge time periods. For example, when the number of charge and discharge time periods experienced by a battery cell is 200, the health factor of the battery cell in each of the 200 charge and discharge time periods is used as the first sample health factor, and the health factor attenuation model is updated based on the first sample health factor. For another example, when the number of charge and discharge time periods experienced by a battery cell is 400, the health factor of the battery cell in each of the 400 charge and discharge time periods is used as the first sample health factor, or the health factor of the battery cell in each of the 200-400 charge and discharge time periods is used as the first sample health factor; and the health factor attenuation model is updated based on the first sample health factor.

[0233] In some optional embodiments, when the type, quantity or connection mode of the battery cells in the battery pack changes, the battery management module needs to retrain the health factor attenuation model to improve the prediction accuracy of the health factor attenuation model.

[0234] like Figure 7 As shown, the present application provides a schematic diagram of a health factor attenuation model training process, including the following steps:

[0235] S701: The battery management module obtains sample data of a battery cell to be tested of the same type as a plurality of battery cells during a full life cycle period.

[0236] The sample data includes a first initial sample health factor and a first sample battery capacity corresponding to the battery cell to be tested. The sample data can be stored in the cloud.

[0237] In some optional embodiments, after obtaining the sample data, the battery management module performs data processing on the sample data; wherein the data processing operations include but are not limited to: deduplication and leakage filling, smoothing difference, outlier elimination and homogenization processing.

[0238] S702: The battery management module determines the correlation between the health factor of the first initial sample and the battery capacity of the first sample to obtain a correlation coefficient.

[0239] Optionally, the battery management module may determine the correlation coefficient using the following formula:

[0240]

[0241] Among them, ρ xi represents the correlation coefficient, x i represents the first initial sample health factor sequence, z i represents the first sample battery capacity sequence, represents the average value of the first initial sample health factor sequence, represents the average value of the first sample capacity sequence, and i represents the number of charge and discharge time periods that the battery pack has experienced.

[0242] S703: The battery management module uses the first initial sample health factor whose correlation coefficient is greater than the correlation threshold as the first target sample health factor.

[0243] S704: The battery management module determines the model parameters of the double exponential sub-model according to the first target sample health factor and the number of charge and discharge time periods that the battery pack corresponding to the first target sample health factor has experienced, and obtains a trained double exponential sub-model.

[0244] Optionally, the battery management module may determine the model parameters of the double exponential sub-model by a least squares method. The specific process is the same as the process of adjusting the model parameters of the double exponential sub-model, and will not be repeated here.

[0245] S705, the battery management module inputs the first target sample health factor and the number of charge and discharge time periods that the battery pack corresponding to the first target sample health factor has experienced into the neural network model, trains the neural network model, and obtains a trained neural network model.

[0246] In some optional embodiments, users can also construct an offline experimental scenario to obtain a sample data set to train the health factor attenuation model. The following describes the model training process using a specific embodiment, which specifically includes the following steps:

[0247] c1: The user performs initial capacity calibration on the sample battery cell.

[0248] It should be noted that the sample battery cell is of the same battery type as the battery cells in the battery pack, wherein the number of the sample battery cells may be 9.

[0249] c2: Users can divide the sample battery cells into three equal parts, perform constant current and constant voltage charge and discharge experiments on the sample battery cells under certain environmental conditions, and establish a working condition test experiment database.

[0250] The operating condition test database includes but is not limited to: current, voltage, temperature, and capacity.

[0251] Specifically, three sample battery cells were placed in environments with temperatures of 35° C., 45° C., and 50° C., respectively, and constant current and constant voltage charge and discharge experiments were performed.

[0252] For example, Figure 8 As shown, the present application provides a battery charge and discharge strategy schematic diagram, wherein: after the user leaves the sample battery cell at a constant temperature of 35°C for 2 hours, a charging test is performed under constant current and constant voltage conditions to obtain the voltage and current data of the sample battery cell; wherein, when the sample battery cell is in a charging state, the current value is +50A. When the sample battery cell is in a stationary state, the current value is 0. After the sample battery cell is left at rest for 30 minutes, a discharge test is performed under constant current and constant voltage conditions to obtain battery cell related data. At this time, the current value is -50A. wherein, the charging test is the same as the constant current and constant voltage conditions in the discharge test. The battery charge and discharge strategy is executed cyclically until it is determined that the battery capacity of the battery cell is less than 80% of the initial calibrated capacity of the battery cell, and a working condition test experiment database is established based on the acquired voltage and current data.

[0253] In other optional implementations, the user's experimental process for sample battery cells at different temperatures is the same, which will not be described again here.

[0254] c3: The battery management module extracts discharge data from the operating condition test database and preprocesses the discharge data. Preprocessing operations include, but are not limited to, deduplication and omission correction, smoothing interpolation, outlier removal, and homogenization.

[0255] c4: The battery management module determines the second initial sample health factor of the sample battery cell in each charge and discharge time period based on the extracted discharge data, and performs a correlation analysis between the determined second initial sample health factor and the sample capacity of the battery cell, screening out the second target sample health factor that has a strong correlation with the sample capacity of the sample battery cell.

[0256] In c4, the process of the battery management module determining the second initial sample health factor and screening the second target sample health factor is the same as the process of determining the first initial sample health factor and screening the first target sample health factor, which will not be repeated here.

[0257] c5: The battery management module averages and fuses the second target sample health factors of the sample battery cells under the same operating conditions to obtain a fused health factor.

[0258] The sample battery cells under the same operating conditions refer to sample battery cells at the same temperature, for example, sample battery cells 1, 2, and 3 that are subjected to a constant voltage and constant current charge and discharge experiment at a constant temperature of 35°C.

[0259] For example, Figure 9 As shown in the figure, a constant voltage and constant current charge and discharge experiment was conducted at a constant temperature of 35°C. The health factor length of sample battery cell 1 was 6, the health factor length of sample battery cell 2 was 3, and the health factor length of sample battery cell 3 was 4. Among them, the health factor length of sample battery cell 2 was the smallest. The battery management module intercepted the health factor length that matched the health factor length of sample battery cell 3 and averaged the health factors of sample battery cells 1, 2, and 3 for this length to obtain the fused health factor.

[0260] c6: The battery management module trains the double exponential sub-model in the health factor attenuation model based on the fused health factor.

[0261] c7: The battery management module connects in series the average values ​​of the second target sample health factors of the sample battery cells under different working conditions to obtain a health factor sequence.

[0262] Optionally, the battery management module first averages and fuses the second target sample health factors of sample battery cells under the same operating condition to obtain a fused health factor, and then concatenates the fused health factors under different operating conditions to obtain a health factor sequence.

[0263] For example, the second target health factors of sample battery cells 1, 2 and 3 at 35°C in different charge and discharge time periods are averaged and fused to obtain a fused health factor sequence A1; the second target health factors of sample battery cells 4, 5 and 6 at 40°C in different charge and discharge time periods are averaged and fused to obtain a fused health factor sequence A2; the second target health factors of sample battery cells 7, 8 and 9 at 50°C in different charge and discharge time periods are averaged and fused to obtain a fused health factor sequence A3; the fused health factor sequences A1, A2 and A3 at different temperatures are connected in series to obtain a health factor sequence.

[0264] c8, the battery management module trains the neural network model according to the health factor sequence.

[0265] After training the double exponential sub-model and the neural network model, the battery management module obtains the health factor attenuation model.

[0266] Step S303 : The battery management module determines the battery capacity of the battery pack in the second charge and discharge time period according to the health factors of the plurality of battery cells in the second charge and discharge time period.

[0267] In some optional embodiments, such as Figure 10 As shown, the embodiment of the present application provides a battery capacity prediction method; wherein, the battery management module can perform S303 by the following steps:

[0268] S1001: A battery management module inputs health factors of multiple battery cells in a second charge and discharge time period into a capacity estimation model to obtain candidate battery capacity values ​​of the battery pack in the second charge and discharge time period.

[0269] S1002, the battery management module determines whether the candidate battery capacity value is greater than the capacity threshold; if so, execute step S1001; if not, execute step S1003.

[0270] The capacity threshold may be 80% of the rated battery capacity of the battery pack, or may be other values ​​set by the user, which is not limited here.

[0271] S1003: The battery management module uses the candidate battery capacity value as the battery capacity of the battery pack in the second charge and discharge time period.

[0272] In some other optional embodiments, the battery management module inputs the health factors of the multiple battery cells in the second charge and discharge time period into the capacity estimation model, and directly outputs the obtained candidate battery capacity values ​​of the battery pack in the second charge and discharge time period.

[0273] In some optional embodiments, when the number of charge and discharge time periods experienced by the battery pack is equal to the charge and discharge threshold, the battery management module needs to train the capacity estimation model. That is, before executing S403, the battery management module needs to perform the following operations to train the capacity estimation model:

[0274] d1: The battery management module determines the battery capacity of the battery pack in the fourth charge and discharge time period and the health factor of each battery cell in the fourth charge and discharge time period based on the operating data of each battery cell in the fourth charge and discharge time period.

[0275] Wherein, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the fourth charge and discharge time period is the last charge and discharge time period among the charge and discharge time periods experienced by the battery pack.

[0276] For example, when the charge and discharge threshold is 100, the battery management module obtains the operating data of each battery cell in the plurality of battery cells in each of the previous 100 charge and discharge time periods experienced by the battery pack.

[0277] In d1, the process of the battery management module determining the health factor is the same as the process of the battery management module determining the health factor of each battery cell in the first charge and discharge time period when the first charge and discharge time period is the charge and discharge time period that the battery pack has experienced, and will not be repeated here.

[0278] In d1, the battery management module can determine the total current value generated by the battery pack during the discharge process based on the discharge data of multiple battery cells collected at multiple sampling moments in the fourth charge and discharge time period, and determine the battery capacity of the battery pack in the fourth charge and discharge time period based on the obtained total current value.

[0279] d2: the battery management module uses the health factor of each battery cell in the fourth charge and discharge time period as the second sample health factor; and uses the battery capacity of the battery pack in the fourth charge and discharge time period as the sample battery capacity.

[0280] d3: The battery management module builds a capacity estimation model based on the second sample health factor and the sample battery capacity.

[0281] In d3, the battery management module can establish a mapping relationship between the second sample health factor and the sample battery capacity through machine learning, and build a capacity estimation model.

[0282] For example, when the battery pack includes 5 battery cells, the health factors are three health factors: charge variance, charge difference variance, and capacity increment curve peak. Therefore, the mapping relationship between the second sample health factor and the sample battery capacity established by the battery management module is the mapping relationship between the 15-dimensional health factor matrix and the battery capacity.

[0283] In some optional embodiments, the battery management module inputs the second sample health factor and the sample battery capacity into the capacity estimation model, and can train the capacity estimation model through the GPR algorithm to obtain a mapping relationship between the health factor and the battery capacity, thereby obtaining a trained capacity estimation model.

[0284] Optionally, a capacity estimation model is usually defined as:

[0285] f(x)~GP(m(x), k(x, x'))

[0286] Where f(x) represents the battery capacity and obeys Gaussian distribution; x represents the health factor; m(x) is the mean function; and k(x, x') is the covariance function (kernel function).

[0287] Among them, the specific expressions of m(x) and k(x, x') are as follows:

[0288]

[0289] In this model, the mean function m(x) is taken as zero mean, and the square exponential covariance is used as the kernel function. The expression of the kernel function is as follows:

[0290]

[0291] Where M = diag(l 2 ), l is the variance scale, is the signal variance; where the hyperparameters in the kernel function are

[0292] Next, the battery management module constructs the negative log-likelihood function L(θ) based on Bayesian inference and maximum likelihood. It then uses the gradient ascent method to find the partial derivative of L(θ) to optimize the hyperparameter θ, completing the training of the capacity estimation model. The expressions for L(θ) and its partial derivative with respect to the hyperparameter θ are as follows:

[0293]

[0294] in,

[0295] Then, the battery management module determines the posterior distribution of the sample battery capacity y and the battery capacity y* of the battery pack during the second charge and discharge time period, using the following formula:

[0296]

[0297] in,

[0298]

[0299] Indicates the battery capacity of the battery pack during the second charge and discharge time period; σ 2 (y*) represents the predicted covariance, which is used to characterize the confidence of the capacity estimation model. For example, a 95% confidence interval can be expressed as

[0300] In this application, the battery management module constructs a capacity estimation model based on the second sample health factor and sample battery capacity, and only uses the health factor and battery capacity within the fourth charging and discharging time period. There is no need to obtain the health factor in real time to train the capacity estimation model. While realizing the rapid training of the capacity estimation model, the purpose of more accurate and faster prediction of battery capacity is achieved.

[0301] In some optional implementations, the battery management module may regularly train the capacity estimation model according to a set training cycle, so that the capacity estimation model can better fit the battery pack and ensure the prediction accuracy of the capacity estimation model.

[0302] Step S304 : The battery management module determines the remaining service life of the battery pack according to the battery capacity of the battery pack in the second charge and discharge time period.

[0303] In some optional embodiments, the battery management module determines the remaining service life of the battery pack through the following steps:

[0304] e1: The battery management module obtains the target charge and discharge times.

[0305] It should be noted that the target charge and discharge times are used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge.

[0306] e2: The battery management module determines a difference between the calibrated battery capacity of the battery pack and the battery capacity of the battery pack during the second charge and discharge time period.

[0307] e3: The battery management module determines the ratio of the target charge and discharge times to the difference.

[0308] e5: The battery management module uses the product of the ratio and the battery capacity of the battery pack in the second charge and discharge time period as the remaining service life of the battery pack.

[0309] For example, the battery management module determines that the target number of charge and discharge times is 100 times, the calibrated battery capacity is 100%, and the battery capacity of the battery pack in the second charge and discharge time period is 80%. The remaining service life of the battery pack is 100*80% / (100%-80%)=400 full charge and discharge time periods, wherein the minimum capacity of the battery pack is set to 0%.

[0310] In some other optional implementations, the battery management module may further determine the remaining service life of the battery pack through the following steps:

[0311] f1: The battery management module obtains the target charge and discharge times.

[0312] It should be noted that the target charge and discharge times are used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge.

[0313] f2: The battery management module determines the number of charge and discharge times corresponding to the second charge and discharge time period when the battery capacity of the battery pack in the second charge and discharge time period is equal to the minimum capacity threshold; and uses the number of charge and discharge times as the actual service life of the battery pack.

[0314] f3: The battery management module uses the difference between the actual service life and the target charge and discharge times as the remaining service life of the battery pack.

[0315] For example, the minimum capacity threshold is 78% of the battery pack's calibrated battery capacity, and the battery pack determines a target charge and discharge cycle of 100 cycles. The battery pack determines whether the predicted battery capacity during the second charge and discharge time period reaches 78% of the calibrated battery capacity. If it reaches 78%, the battery pack determines that the number of charge and discharge cycles corresponding to the second charge and discharge time period is 500 cycles. 500 cycles are then used as the actual service life of the battery pack. The battery management module calculates the difference (400 cycles) between the actual service life of 500 cycles and the target charge and discharge cycle of 100 cycles as the remaining service life of the battery pack.

[0316] Optionally, after the battery management module determines the remaining service life of the battery pack, when determining the remaining service life of the battery pack during subsequent use, the difference between the remaining service life and the number of full charge and discharge time periods that the battery pack subsequently experiences can be directly used as the remaining service life corresponding to the subsequent use of the battery pack.

[0317] In other optional embodiments, after outputting the battery capacity of the battery pack in the second charge and discharge time period, the battery management module counts the total number of full charge and discharge charge and discharge time periods that the battery pack has experienced; the battery management module uses the difference between the calibrated service life of the battery pack and the total number of full charge and discharge charge and discharge time periods that the battery pack has experienced as the remaining service life of the battery pack.

[0318] For example, the calibrated service life of the battery pack is that the total number of full-charge-and-discharge charge and discharge time periods that the battery pack can experience is 300 times. After the battery management module determines the battery capacity of the battery pack in the second charge and discharge time period, the total number of full-charge-and-discharge charge and discharge time periods that the battery pack has experienced is 100. Then the remaining service life of the battery pack is, and the total number of full-charge-and-discharge charge and discharge time periods that the battery pack can experience is 300 times.

[0319] In some optional embodiments, after determining the remaining service life, the battery management module may send a battery abnormality message to the vehicle terminal and the energy storage display device when the remaining service life is close to the lower limit of the battery pack. After receiving the battery abnormality message, the vehicle terminal and the energy storage display device display the battery abnormality message on the corresponding display interface, so that the user can perform maintenance after receiving the battery abnormality message.

[0320] Figure 11 A complete flow chart of a battery pack life prediction method provided in an embodiment of the present application is applied to a battery pack, which includes multiple battery cells; wherein, the battery management module is used as the execution body, and the battery pack life prediction process is executed as an example, such as Figure 11 As shown, the method includes:

[0321] S1101: A battery management module obtains operating data of each of a plurality of battery cells in one or more first charge and discharge time periods.

[0322] S1102: The battery management module determines the health factor of each corresponding battery cell in each first charge and discharge time period according to the working data of each battery cell in each first charge and discharge time period.

[0323] It should be noted that the working data of each battery cell in each first charge and discharge time period includes: the discharge data of each battery cell collected at multiple sampling moments in each first charge and discharge time period; the health factor includes the capacity increment curve peak, power variance, and power difference variance.

[0324] In some optional embodiments, the battery management module may determine the health factor in the following manner:

[0325] Determine the power variance and power difference variance of each battery cell in each first charge and discharge time period based on the discharge data of each battery cell collected at multiple sampling moments in each first charge and discharge time period;

[0326] Determine, based on the discharge data of each battery cell collected at multiple sampling moments in each first charge and discharge time period, a battery capacity increment of each battery cell corresponding to the multiple sampling moments in each first charge and discharge time period;

[0327] The maximum value of the battery capacity increments of each battery cell corresponding to multiple sampling moments in each first charge and discharge time period is taken as the capacity increment curve peak value of the corresponding battery cell in the corresponding first charge and discharge time period.

[0328] S1103, the battery management module determines whether the total number of charge and discharge time periods experienced by the battery pack is greater than the charge and discharge threshold; if not, execute step S1104; if so, execute step S1110.

[0329] S1104: The battery management module obtains the health factor of each battery cell in the third charge and discharge time period.

[0330] Wherein, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the third charge and discharge time period is the last charge and discharge time period among the charge and discharge time periods experienced by the battery pack.

[0331] S1105 : The battery management module uses the health factor of each battery cell in the third charge and discharge time period as a first sample health factor, and updates the corresponding health factor attenuation model according to the first sample health factor of each battery cell.

[0332] It should be noted that the health factor attenuation model includes a double exponential model and a neural network model.

[0333] Optionally, the battery management module can update the health factor decay model in the following ways:

[0334] The battery management module adjusts the double exponential sub-model according to the first sample health factor and the number of charge and discharge times corresponding to the third charge and discharge time period; the battery management module inputs the first sample health factor into the fully connected layer of the neural network model, trains the fully connected layer, and updates the neural network model according to the training results.

[0335] S1106, the battery management module determines the battery capacity of the battery pack in the fourth charge and discharge time period and the health factor of each battery cell in the fourth charge and discharge time period based on the operating data of each battery cell in the fourth charge and discharge time period.

[0336] It should be noted that, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the fourth charge and discharge time period is the last charge and discharge time period among the charge and discharge time periods experienced by the battery pack.

[0337] Among them, S1104 and S1106 are executed simultaneously, regardless of the order.

[0338] S1107: The battery management module uses the health factor of each battery cell in the fourth charge and discharge time period as a second sample health factor.

[0339] S1108: The battery management module uses the battery capacity of the battery pack in the fourth charge and discharge time period as a sample battery capacity.

[0340] S1109: The battery management module constructs a capacity estimation model according to the second sample health factor and the sample battery capacity.

[0341] S1110 , the battery management module inputs the health factor of each battery cell in one or more first charge and discharge time periods into a health factor attenuation model to obtain the corresponding health factor of each battery cell in a second charge and discharge time period.

[0342] The one or more first charge and discharge time periods and the second charge and discharge time period are adjacent time periods.

[0343] S1111 , the battery management module inputs health factors of the plurality of battery cells in the second charge and discharge time period into a capacity estimation model to obtain candidate battery capacity values ​​of the battery pack in the second charge and discharge time period.

[0344] S1112, the battery management module determines whether the candidate battery capacity value is greater than the capacity threshold; if so, execute S1101; if not, execute S1113.

[0345] S1113 , the battery management module uses the candidate battery capacity value as the battery capacity of the battery pack in the second charge and discharge time period.

[0346] S1114: The battery management module obtains the target number of charge and discharge times.

[0347] It should be noted that the target charge and discharge times are used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge.

[0348] S1115: The battery management module determines a difference between the calibrated battery capacity of the battery pack and the battery capacity of the battery pack during the second charge and discharge time period.

[0349] S1116, the battery management module determines the ratio of the target charge and discharge times to the difference.

[0350] S1117: The battery management module uses the product of the ratio and the battery capacity of the battery pack in the second charge and discharge time period as the remaining service life of the battery pack.

[0351] S1118, the battery management module determines whether the remaining service life reaches the lower limit of the service life; if so, execute S1119; if not, execute S1101.

[0352] S1119, the battery management module sends battery abnormality information to the vehicle terminal, so that the vehicle terminal displays the battery abnormality information to the user.

[0353] Figure 12 A complete flow chart of another battery pack life prediction method provided in an embodiment of the present application is applied to a battery pack, which includes multiple battery cells; wherein, the battery management module is used as the execution body, and the battery pack life prediction process is executed as an example, such as Figure 12 As shown, the method includes:

[0354] S1201: The battery management module obtains a health factor set for each battery cell.

[0355] The health factor set includes the health factor of each battery cell in the charge and discharge time period that the battery pack has experienced, and the health factor of each battery cell in the predicted charge and discharge time period that the battery pack has not experienced.

[0356] In some optional implementations, before acquiring the health factor set, the battery management module has already updated the health factor attenuation model corresponding to each battery cell and completed the training of the capacity estimation model.

[0357] S1202: The battery management module selects, from the health factor set, health factors in one or more charging and discharging time periods that are adjacent to the second charging and discharging time period as health factors in one or more first charging and discharging time periods.

[0358] S1203: The battery management module inputs the health factor of each battery cell in one or more first charge and discharge time periods into a health factor attenuation model to obtain the corresponding health factor of each battery cell in a second charge and discharge time period.

[0359] S1204, the battery management module determines whether the number of charge and discharge times corresponding to the second charge and discharge time period reaches a set value, if not, execute step S1205; if yes, execute step S1206.

[0360] The set value is used to represent the number of full charge and discharge time periods that the battery pack can experience, which is set by the user.

[0361] S1205: The battery management module updates the health factor set of the corresponding battery cell according to the health factor of each battery cell in the second charge and discharge time period.

[0362] S1206: The battery management module inputs the health factors of the multiple battery cells in each of the multiple second charge and discharge time periods into the capacity estimation model to obtain the battery capacity of the battery pack in the corresponding second charge and discharge time period.

[0363] S1207: The battery management module obtains the target charge and discharge times.

[0364] The target charge and discharge times are used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge.

[0365] S1208: The battery management module determines the number of charge and discharge times corresponding to the second charge and discharge time period when the battery capacity of the battery pack in the second charge and discharge time period is equal to the minimum capacity threshold.

[0366] S1209: The battery management module uses the number of charge and discharge times as the actual service life of the battery pack.

[0367] S1210: The battery management module uses the difference between the actual service life and the target charge and discharge times as the remaining service life of the battery pack.

[0368] S1211, the battery management module determines whether the remaining service life reaches the lower limit of the service life; if so, execute S1212; if not, execute S1201.

[0369] S1212, the battery management module sends battery abnormality information to the vehicle terminal, so that the vehicle terminal displays the battery abnormality information to the user.

[0370] Based on the same inventive concept, the embodiment of the present application also provides a battery pack life prediction system, wherein the battery pack life prediction system can be located in the battery management module or in other processing modules, and is applied to a battery pack including a plurality of battery cells. Figure 13 As shown, the battery pack life prediction system includes: an offline battery cell health factor unit, a battery data acquisition unit, a health factor prediction unit, a capacity estimation unit, and an abnormality warning unit.

[0371] The offline battery cell health factor unit is used to store the health factor of the battery cell of the same type as the battery cell in the battery pack over the entire life period, as well as the health factor and battery capacity of the first 10% of the battery pack life period. The offline battery cell health factor unit can be a cloud server.

[0372] The battery data acquisition unit is configured to acquire the health factors stored in the offline battery cell health factor unit and to acquire the operating data of each battery cell in the battery pack during a first charge and discharge time period. The battery data acquisition unit determines the health factor for the first charge and discharge time period based on the operating data of the battery cells during the first charge and discharge time period. Furthermore, the battery data acquisition unit may also acquire a set of health factors corresponding to each battery cell and select one or more health factors from the set as the health factors for one or more first charge and discharge time periods.

[0373] The health factor prediction unit is used to predict the health factor in the second charging and discharging time period based on the health factors obtained in one or more first charging and discharging time periods.

[0374] The capacity estimation unit is configured to determine the battery capacity of the battery pack in the second charge and discharge time period according to the health factors of the plurality of battery cells in the second charge and discharge time period output by the health factor prediction unit.

[0375] The abnormality warning unit is configured to obtain a target number of charge and discharge cycles and determine the remaining useful life of the battery pack based on the battery capacity of the battery pack during the second charge and discharge time period and the target number of charge and discharge cycles. Furthermore, upon determining that the remaining useful life of the battery pack has reached a lower limit, the abnormality warning unit issues an alarm to maintenance personnel.

[0376] In some optional embodiments, when the battery data acquisition unit determines that the total number of charge and discharge time periods that the battery pack has experienced is greater than the charge and discharge threshold, the health factor and capacity of the battery pack in the set charge and discharge time period are obtained from the offline battery cell health factor unit. The health factor prediction unit updates the health factor attenuation models corresponding to multiple battery cells in the battery pack based on the health factors of the first ten percent of the battery pack's life period. After the update is completed, the health factor prediction unit predicts the health factor of the corresponding battery cell in the second charge and discharge time period based on the health factor of each battery cell in one or more first charge and discharge time periods. In addition, the capacity estimation unit can train the capacity estimation model based on the health factors and battery capacity of the first ten percent of the battery pack's life period. After completing the model training, the capacity estimation model predicts the battery capacity of the battery pack in the second charge and discharge time period based on the health factors of multiple battery cells in the battery pack in the second charge and discharge time period.

[0377] Based on the same technical concept, the embodiment of the present application also provides a battery pack life prediction device, which is applied to a battery pack, wherein the battery pack includes a plurality of battery cells, such as Figure 14 As shown, the battery life prediction device 1400 may include:

[0378] An acquiring unit 1401 is configured to acquire a health factor of each of the plurality of battery cells in one or more first charge and discharge time periods; the health factor is used to characterize a capacity decay characteristic of the battery cell;

[0379] A first prediction unit 1402 is configured to predict a health factor of each battery cell in a second charge and discharge time period based on the health factor of each battery cell in one or more of the first charge and discharge time periods;

[0380] A second prediction unit 1403 is configured to determine a battery capacity of the battery pack in the second charge and discharge time period according to health factors of the plurality of battery cells in the second charge and discharge time period;

[0381] The determining unit 1404 is configured to determine the remaining service life of the battery pack according to the battery capacity of the battery pack in the second charge and discharge time period.

[0382] In one embodiment, the first charge and discharge time period is a charge and discharge time period that the battery pack has experienced;

[0383] The acquiring unit 1401 is specifically configured to:

[0384] Acquire operating data of each battery cell during one or more of the first charge and discharge time periods;

[0385] According to the working data of each battery cell in each first charge and discharge time period, a health factor of each corresponding battery cell in each first charge and discharge time period is determined.

[0386] In one embodiment, the health factor includes a capacity increment curve peak, a power variance, and a power difference variance; the acquisition unit 1401 is specifically configured to:

[0387] Obtaining discharge data of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods within one or more of the first charge and discharge time periods;

[0388] determining, based on the discharge data of each battery cell collected at a plurality of sampling moments in each of the first charge and discharge time periods, a battery capacity increment of each battery cell corresponding to the plurality of sampling moments in each of the first charge and discharge time periods;

[0389] taking the maximum value of the battery capacity increments of each battery cell corresponding to the multiple moments in each of the first charge and discharge time periods as the peak value of the capacity increment curve of the corresponding battery cell in the corresponding first charge and discharge time period; and

[0390] According to the discharge data of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods, the power variance and power difference variance of each battery cell in each of the first charge and discharge time periods are determined.

[0391] In one embodiment, the discharge data includes a voltage value and a current value corresponding to the voltage value.

[0392] In one embodiment, the acquiring unit 1401 is specifically configured to:

[0393] The voltage values ​​of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods are divided to obtain multiple voltage value sets in each of the first charge and discharge time periods; the difference between the maximum voltage value and the minimum voltage value in any voltage value set is equal to the set value;

[0394] Determining, based on the multiple voltage value sets within each of the first charging and discharging time periods, multiple current value sets corresponding one-to-one to the multiple voltage value sets within each of the first charging and discharging time periods, wherein for a corresponding voltage value set and a corresponding current value set, one or more current values ​​included in the current value set are in one-to-one correspondence with one or more voltage values ​​included in the corresponding voltage value set;

[0395] Determining, according to each voltage value set and the current value set corresponding to each voltage value set in each first charging and discharging time period, a power value corresponding to each voltage value set in each first charging and discharging time period;

[0396] According to the power values ​​corresponding to each voltage value set in each first charge and discharge time period, the power variance and power difference variance of each battery cell in each first charge and discharge time period are determined.

[0397] In one embodiment, the acquiring unit 1401 is specifically configured to:

[0398] According to the voltage value and corresponding current value of each battery cell collected at multiple sampling moments in each first charge and discharge time period, the battery capacity increment of each battery cell corresponding to the multiple sampling moments in each first charge and discharge time period is determined.

[0399] In one embodiment, the acquiring unit 1401 is specifically configured to:

[0400] Obtaining a health factor set for each battery cell; the health factor set includes a health factor of each battery cell during a charge and discharge time period that the battery pack has experienced, and a health factor of each battery cell during a predicted charge and discharge time period that the battery pack has not experienced;

[0401] Health factors in one or more charging and discharging time periods that are adjacent to the second charging and discharging time period are selected from the health factor set as health factors in one or more first charging and discharging time periods.

[0402] In one embodiment, the first prediction unit 1402 is specifically configured to:

[0403] The health factor of each battery cell in one or more of the first charge and discharge time periods is input into a health factor attenuation model to obtain the corresponding health factor of each battery cell in the second charge and discharge time period; the one or more first charge and discharge time periods and the second charge and discharge time period are adjacent time periods.

[0404] In one embodiment, the health factor attenuation model includes a double exponential sub-model and a neural network model; the first prediction unit 1402 is specifically configured to:

[0405] Inputting the number of charge and discharge times corresponding to the second charge and discharge time period into the double exponential sub-model to obtain a first candidate health factor for each battery cell in the second charge and discharge time period; the number of charge and discharge times is the number of charge and discharge time periods experienced by the battery pack;

[0406] Inputting the health factor of each battery cell in one or more of the first charge and discharge time periods into the neural network model to obtain a second candidate health factor of each corresponding battery cell in the second charge and discharge time period;

[0407] The health factor of each battery cell in the second charge and discharge time period is determined according to the first candidate health factor and the second candidate health factor of each battery cell in the second charge and discharge time period.

[0408] In one embodiment, the first prediction unit 1402 is specifically configured to:

[0409] If the second candidate health factor is less than the product of the first candidate health factor and the set ratio, the first candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period;

[0410] If the second candidate health factor is greater than or equal to the product of the first candidate health factor and the set ratio, the second candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period.

[0411] In one embodiment, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the third charge and discharge time period is the last charge and discharge time period experienced by the battery pack;

[0412] The acquiring unit 1401 is further configured to: acquire the health factor of each battery cell in the third charge and discharge time period;

[0413] The first prediction unit 1402 is further configured to:

[0414] The health factor of each battery cell in the third charge and discharge time period is used as a first sample health factor, and the corresponding health factor attenuation model is updated according to the first sample health factor of each battery cell.

[0415] In one embodiment, the first prediction unit 1402 is specifically configured to:

[0416] adjusting the double exponential sub-model according to the first sample health factor and the number of charge and discharge times corresponding to the third charge and discharge time period;

[0417] The first sample health factor is input into the neural network model, the fully connected layer of the neural network model is trained, and the neural network model is updated according to the training results.

[0418] In one implementation, the second prediction unit 1403 is specifically configured to:

[0419] inputting the health factors of the plurality of battery cells in the second charge and discharge time period into a capacity estimation model to obtain a candidate value of the battery capacity of the battery pack in the second charge and discharge time period;

[0420] If the candidate battery capacity value is less than or equal to the capacity threshold, using the candidate battery capacity value as the battery capacity of the battery pack in the second charge and discharge time period;

[0421] If the battery capacity candidate value is greater than the capacity threshold, the health factors of the multiple battery cells in the second charge and discharge time period are continuously input into the capacity estimation model until the obtained battery capacity candidate value is less than or equal to the capacity threshold.

[0422] In one embodiment, when the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the fourth charge and discharge time period is the last charge and discharge time period experienced by the battery pack;

[0423] The acquiring unit 1401 is further configured to determine, based on the operating data of each battery cell in the fourth charge and discharge time period, the battery capacity of the battery pack in the fourth charge and discharge time period and the health factor of each battery cell in the fourth charge and discharge time period;

[0424] The second prediction unit 1403 is further configured to: use the health factor of each battery cell in the fourth charge and discharge time period as a second sample health factor; and use the battery capacity of the battery pack in the fourth charge and discharge time period as a sample battery capacity;

[0425] The capacity estimation model is constructed according to the second sample health factor and the sample battery capacity.

[0426] In one implementation, the determining unit 1404 is specifically configured to:

[0427] Obtaining a target charge and discharge number; the target charge and discharge number is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge;

[0428] determining a difference between a calibrated battery capacity of the battery pack and a battery capacity of the battery pack during the second charge and discharge time period;

[0429] Determining a ratio of the target charge and discharge times to the difference;

[0430] The product of the ratio and the battery capacity of the battery pack in the second charge and discharge time period is used as the remaining service life of the battery pack.

[0431] In one implementation, the determining unit 1404 is specifically configured to:

[0432] Obtaining a target charge and discharge number; the target charge and discharge number is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge;

[0433] determining the number of charge and discharge times corresponding to the second charge and discharge time period when the battery capacity of the battery pack in the second charge and discharge time period is equal to a minimum capacity threshold; and using the number of charge and discharge times as the actual service life of the battery pack;

[0434] The difference between the actual service life and the target charge and discharge times is used as the remaining service life of the battery pack.

[0435] like Figure 15 As shown, the embodiment of the present application provides a possible structural diagram of a battery module; wherein the structure of the battery module is as follows Figure 15 As shown, it includes a battery pack 1501 and a battery pack life prediction device 1502. When the battery pack is in operation, the battery pack life prediction device executes the above embodiments and the battery pack life prediction method provided in the embodiments.

[0436] like Figure 16 As shown, the embodiment of the present application provides a possible schematic diagram of the architecture of an energy storage system; wherein the structure of the energy storage system is as follows Figure 16 As shown, it at least includes a battery module 1601 and a controller 1602. Figure 15 The battery modules shown are the same and will not be described again here.

[0437] In some optional implementations, after receiving the operation instruction, the controller 1602 controls the operation of the battery module 1601. Specifically, the controller 1602 is used to control the charging or discharging of the battery module 1601.

[0438] When the battery module 1601 is in operation, it can provide power to other electronic devices connected to the energy storage system, and can also charge the battery pack in the battery module 1601 .

[0439] When battery module 1601 is operating, the battery life prediction device in battery module 1601 is configured to execute the method described in the above-described method embodiment to predict the remaining useful life of the battery pack. When battery module 1601 determines that the predicted remaining useful life has reached a lower limit, it sends an alarm signal to the controller, which in turn alerts maintenance personnel.

[0440] It should be noted that the above Figure 16 The energy storage system shown is only an exemplary description of the energy storage system applicable to the present application solution and does not limit the system architecture applicable to the present application solution. Other devices or modules can also be added to the above system architecture, or some devices or modules can be reduced or modified.

[0441] Based on the above content and the same concept, the present application provides a computer-readable storage medium having a computer program or instructions stored thereon. When the computer program or instructions are executed, the computing device executes the method in the above method embodiment.

[0442] Based on the above content and the same concept, the present application provides a computer program product. When a computer executes the computer program product, the computing device executes the method in the above method embodiment.

[0443] It should be understood that the division of modules in the embodiments of the present application is illustrative and is merely a logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in the various embodiments of the present application may be integrated into a single processor, or may exist physically as separate modules, or two or more modules may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or software functional modules.

[0444] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0445] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0446] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0447] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0448] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A battery pack life prediction method, applied to a battery pack, wherein the battery pack includes a plurality of battery cells, characterized in that: The method comprises: Obtaining a health factor of each battery cell in the plurality of battery cells during one or more first charge and discharge time periods; the health factor is used to characterize a capacity decay characteristic of the battery cell; Inputting the number of charge and discharge times corresponding to the second charge and discharge time period into the double exponential submodel in the health factor attenuation model to obtain a first candidate health factor for each battery cell in the second charge and discharge time period; the number of charge and discharge times is the number of charge and discharge time periods experienced by the battery pack, and the one or more first charge and discharge time periods and the second charge and discharge time period are adjacent time periods; Inputting the health factor of each battery cell in one or more of the first charge and discharge time periods into the neural network model in the health factor attenuation model to obtain a second candidate health factor of each corresponding battery cell in a second charge and discharge time period; determining a health factor of each battery cell in the second charge and discharge time period according to the first candidate health factor and the second candidate health factor of each battery cell in the second charge and discharge time period; determining a battery capacity of the battery pack in the second charge and discharge time period according to health factors of the plurality of battery cells in the second charge and discharge time period; The remaining service life of the battery pack is determined according to the battery capacity of the battery pack in the second charge and discharge time period.

2. The method according to claim 1, characterized in that The first charge and discharge time period is a charge and discharge time period that the battery pack has experienced; The obtaining of the health factor of each battery cell in the plurality of battery cells in one or more first charge and discharge time periods includes: Acquire operating data of each battery cell during one or more of the first charge and discharge time periods; According to the working data of each battery cell in each first charge and discharge time period, a health factor of each corresponding battery cell in each first charge and discharge time period is determined.

3. The method according to claim 2, characterized in that The obtaining of the operating data of each battery cell in one or more of the first charge and discharge time periods includes: obtaining discharge data of each battery cell collected at a plurality of sampling moments in each of the one or more first charge and discharge time periods; The health factors include capacity increment curve peak, power variance, and power difference variance; The determining, based on the operating data of each battery cell in each of the first charge and discharge time periods, a health factor of each corresponding battery cell in each of the first charge and discharge time periods includes: determining, based on the discharge data of each battery cell collected at a plurality of sampling moments in each of the first charge and discharge time periods, a battery capacity increment of each battery cell corresponding to the plurality of sampling moments in each of the first charge and discharge time periods; taking the maximum value of the battery capacity increments of each battery cell corresponding to the plurality of sampling moments in each of the first charge and discharge time periods as the peak value of the capacity increment curve of the corresponding battery cell in the corresponding first charge and discharge time period; and According to the discharge data of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods, the power variance and power difference variance of each battery cell in each of the first charge and discharge time periods are determined.

4. The method according to claim 3, characterized in that The discharge data includes a voltage value and a current value corresponding to the voltage value.

5. The method according to claim 4, characterized in that The determining, based on the discharge data of each battery cell collected at a plurality of sampling moments in each of the first charge and discharge time periods, the power variance and power difference variance of each battery cell in each of the first charge and discharge time periods includes: The voltage values ​​of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods are divided to obtain multiple voltage value sets in each of the first charge and discharge time periods; the difference between the maximum voltage value and the minimum voltage value in any voltage value set is equal to the set value; Determining, based on the multiple voltage value sets within each of the first charging and discharging time periods, multiple current value sets corresponding one-to-one to the multiple voltage value sets within each of the first charging and discharging time periods, wherein for a corresponding voltage value set and a corresponding current value set, one or more current values ​​included in the current value set are in one-to-one correspondence with one or more voltage values ​​included in the corresponding voltage value set; Determining, according to each voltage value set and the current value set corresponding to each voltage value set in each first charging and discharging time period, a power value corresponding to each voltage value set in each first charging and discharging time period; According to the power values ​​corresponding to each voltage value set in each first charge and discharge time period, the power variance and power difference variance of each battery cell in each first charge and discharge time period are determined.

6. The method according to claim 4, characterized in that The step of determining the battery capacity increment of each battery cell corresponding to the plurality of sampling moments in each of the first charge and discharge time periods based on the discharge data of each battery cell collected at the plurality of sampling moments in each of the first charge and discharge time periods includes: According to the voltage value and corresponding current value of each battery cell collected at multiple sampling moments in each first charge and discharge time period, the battery capacity increment of each battery cell corresponding to the multiple sampling moments in each first charge and discharge time period is determined.

7. The method according to any one of claims 1 to 6, characterized in that The obtaining of the health factor of each battery cell in the plurality of battery cells in one or more first charge and discharge time periods includes: Obtaining a health factor set for each battery cell; the health factor set includes a health factor of each battery cell during a charge and discharge time period that the battery pack has experienced, and a health factor of each battery cell during a predicted charge and discharge time period that the battery pack has not experienced; Health factors in one or more charging and discharging time periods that are adjacent to the second charging and discharging time period are selected from the health factor set as health factors in one or more first charging and discharging time periods.

8. The method according to claim 1, characterized in that The determining, based on the first candidate health factor and the second candidate health factor of each battery cell in the second charge and discharge time period, includes: If the second candidate health factor is less than the product of the first candidate health factor and the set ratio, the first candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period; If the second candidate health factor is greater than or equal to the product of the first candidate health factor and the set ratio, the second candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period.

9. The method according to any one of claims 1 to 8, characterized in that in, When the number of charge and discharge time periods experienced by the battery pack is less than or equal to a charge and discharge threshold, the third charge and discharge time period is the last charge and discharge time period experienced by the battery pack; and the method further includes: Obtaining a health factor of each battery cell during the third charge and discharge time period; The health factor of each battery cell in the third charge and discharge time period is used as a first sample health factor, and the corresponding health factor attenuation model is updated according to the first sample health factor of each battery cell.

10. The method according to claim 9, characterized in that Updating a corresponding health factor attenuation model according to the first sample health factor of each battery cell includes: adjusting the double exponential sub-model according to the first sample health factor and the number of charge and discharge times corresponding to the third charge and discharge time period; The first sample health factor is input into the neural network model, the fully connected layer of the neural network model is trained, and the neural network model is updated according to the training results.

11. The method according to any one of claims 1 to 8, characterized in that The determining, based on the health factors of the plurality of battery cells in the second charge and discharge time period, the battery capacity of the battery pack in the second charge and discharge time period includes: inputting the health factors of the plurality of battery cells in the second charge and discharge time period into a capacity estimation model to obtain a candidate value of the battery capacity of the battery pack in the second charge and discharge time period; If the candidate battery capacity value is less than or equal to the capacity threshold, using the candidate battery capacity value as the battery capacity of the battery pack in the second charge and discharge time period; If the battery capacity candidate value is greater than the capacity threshold, the health factors of the multiple battery cells in the second charge and discharge time period are continuously input into the capacity estimation model until the obtained battery capacity candidate value is less than or equal to the capacity threshold.

12. The method according to claim 11, characterized in that When the number of charge and discharge time periods experienced by the battery pack is less than or equal to a charge and discharge threshold, the fourth charge and discharge time period is the last charge and discharge time period experienced by the battery pack; and the method further includes: determining, based on operating data of each battery cell in the fourth charge and discharge time period, a battery capacity of the battery pack in the fourth charge and discharge time period and a health factor of each battery cell in the fourth charge and discharge time period; using the health factor of each battery cell in the fourth charge and discharge time period as a second sample health factor; and using the battery capacity of the battery pack in the fourth charge and discharge time period as a sample battery capacity; The capacity estimation model is constructed according to the second sample health factor and the sample battery capacity.

13. The method according to any one of claims 1 to 12, characterized in that The determining the remaining service life of the battery pack according to the battery capacity of the battery pack in the second charge and discharge time period includes: Obtaining a target charge and discharge number; the target charge and discharge number is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge; determining a difference between a calibrated battery capacity of the battery pack and a battery capacity of the battery pack during the second charge and discharge time period; Determining a ratio of the target charge and discharge times to the difference; The product of the ratio and the battery capacity of the battery pack in the second charge and discharge time period is used as the remaining service life of the battery pack.

14. The method according to any one of claims 1 to 12, characterized in that The determining the remaining service life of the battery pack according to the battery capacity of the battery pack in the second charge and discharge time period includes: Obtaining a target charge and discharge number; the target charge and discharge number is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge; determining the number of charge and discharge times corresponding to the second charge and discharge time period when the battery capacity of the battery pack in the second charge and discharge time period is equal to a minimum capacity threshold; and using the number of charge and discharge times as the actual service life of the battery pack; The difference between the actual service life and the target charge and discharge times is used as the remaining service life of the battery pack.

15. A battery pack life prediction device, applied to a battery pack, wherein the battery pack includes a plurality of battery cells, characterized in that: The device comprises: an acquiring unit, configured to acquire a health factor of each battery cell of the plurality of battery cells during one or more first charge and discharge time periods; the health factor being used to characterize a capacity decay characteristic of the battery cell; a first prediction unit, configured to input the number of charges and discharges corresponding to the second charge and discharge time period into a double exponential submodel in a health factor attenuation model, to obtain a first candidate health factor of each corresponding battery cell in the second charge and discharge time period; the number of charges and discharges being the number of charge and discharge time periods experienced by the battery pack, and the one or more first charge and discharge time periods being adjacent to the second charge and discharge time period; inputting the health factor of each battery cell in the one or more first charge and discharge time periods into a neural network model in the health factor attenuation model, to obtain a second candidate health factor of each corresponding battery cell in the second charge and discharge time period; and determining the health factor of each battery cell in the second charge and discharge time period based on the first candidate health factor and the second candidate health factor of each battery cell in the second charge and discharge time period; a second prediction unit, configured to determine a battery capacity of the battery pack in the second charge and discharge time period according to health factors of the plurality of battery cells in the second charge and discharge time period; A determining unit is configured to determine the remaining service life of the battery pack according to the battery capacity of the battery pack in the second charge and discharge time period.

16. The device according to claim 15, characterized in that The first charge and discharge time period is a charge and discharge time period that the battery pack has experienced; The acquisition unit is specifically configured to: Acquire operating data of each battery cell during one or more of the first charge and discharge time periods; According to the working data of each battery cell in each first charge and discharge time period, a health factor of each corresponding battery cell in each first charge and discharge time period is determined.

17. The device according to claim 16, characterized in that The health factor includes the capacity increment curve peak, power variance, and power difference variance; the acquisition unit is specifically used to: Obtaining discharge data of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods within one or more of the first charge and discharge time periods; determining, based on the discharge data of each battery cell collected at a plurality of sampling moments in each of the first charge and discharge time periods, a battery capacity increment of each battery cell corresponding to the plurality of sampling moments in each of the first charge and discharge time periods; taking the maximum value of the battery capacity increments of each battery cell corresponding to the multiple moments in each of the first charge and discharge time periods as the peak value of the capacity increment curve of the corresponding battery cell in the corresponding first charge and discharge time period; as well as According to the discharge data of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods, the power variance and power difference variance of each battery cell in each of the first charge and discharge time periods are determined.

18. The device according to claim 17, characterized in that The discharge data includes a voltage value and a current value corresponding to the voltage value.

19. The device according to claim 18, characterized in that The acquisition unit is specifically configured to: Dividing the voltage value of each battery cell collected at multiple sampling moments in each of the first charge and discharge time periods to obtain multiple voltage value sets in each of the first charge and discharge time periods; The difference between the maximum voltage value and the minimum voltage value in any voltage value set is equal to the set value; Determining, based on the multiple voltage value sets within each of the first charging and discharging time periods, multiple current value sets corresponding one-to-one to the multiple voltage value sets within each of the first charging and discharging time periods, wherein for a corresponding voltage value set and a corresponding current value set, one or more current values ​​included in the current value set are in one-to-one correspondence with one or more voltage values ​​included in the corresponding voltage value set; Determining, according to each voltage value set and the current value set corresponding to each voltage value set in each first charging and discharging time period, a power value corresponding to each voltage value set in each first charging and discharging time period; According to the power values ​​corresponding to each voltage value set in each first charge and discharge time period, the power variance and power difference variance of each battery cell in each first charge and discharge time period are determined.

20. The device according to claim 18, characterized in that The acquisition unit is specifically configured to: According to the voltage value and corresponding current value of each battery cell collected at multiple sampling moments in each first charge and discharge time period, the battery capacity increment of each battery cell corresponding to the multiple sampling moments in each first charge and discharge time period is determined.

21. The device according to any one of claims 15 to 20, characterized in that The acquisition unit is specifically configured to: Obtaining a health factor set for each battery cell; the health factor set includes a health factor of each battery cell during a charge and discharge time period that the battery pack has experienced, and a health factor of each battery cell during a predicted charge and discharge time period that the battery pack has not experienced; Health factors in one or more charging and discharging time periods that are adjacent to the second charging and discharging time period are selected from the health factor set as health factors in one or more first charging and discharging time periods.

22. The device according to claim 15, characterized in that The first prediction unit is specifically configured to: If the second candidate health factor is less than the product of the set ratios of the first candidate health factors, the first candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period; If the second candidate health factor is greater than or equal to the product of the first candidate health factor and the set ratio, the second candidate health factor is used as the health factor of each corresponding battery cell in the second charge and discharge time period.

23. The device according to any one of claims 15 to 22, characterized in that When the number of charge and discharge time periods experienced by the battery pack is less than or equal to the charge and discharge threshold, the third charge and discharge time period is the last charge and discharge time period experienced by the battery pack; and the acquiring unit is further configured to: Obtaining a health factor of each battery cell during the third charge and discharge time period; The first prediction unit is further configured to: The health factor of each battery cell in the third charge and discharge time period is used as a first sample health factor, and the corresponding health factor attenuation model is updated according to the first sample health factor of each battery cell.

24. The device according to claim 23, characterized in that The first prediction unit is specifically configured to: adjusting the double exponential sub-model according to the first sample health factor and the number of charge and discharge times corresponding to the third charge and discharge time period; The first sample health factor is input into the neural network model, the fully connected layer of the neural network model is trained, and the neural network model is updated according to the training results.

25. The device according to any one of claims 15 to 22, characterized in that The second prediction unit is specifically configured to: inputting the health factors of the plurality of battery cells in the second charge and discharge time period into a capacity estimation model to obtain a candidate value of the battery capacity of the battery pack in the second charge and discharge time period; If the candidate battery capacity value is less than or equal to the capacity threshold, using the candidate battery capacity value as the battery capacity of the battery pack in the second charge and discharge time period; If the battery capacity candidate value is greater than the capacity threshold, the health factors of the multiple battery cells in the second charge and discharge time period are continuously input into the capacity estimation model until the obtained battery capacity candidate value is less than or equal to the capacity threshold.

26. The device according to claim 25, characterized in that When the number of charge and discharge time periods experienced by the battery pack is less than or equal to a charge and discharge threshold, the fourth charge and discharge time period is the last charge and discharge time period experienced by the battery pack; and the acquiring unit is further configured to: determining, based on operating data of each battery cell in the fourth charge and discharge time period, a battery capacity of the battery pack in the fourth charge and discharge time period and a health factor of each battery cell in the fourth charge and discharge time period; The second prediction unit is further configured to: using the health factor of each battery cell in the fourth charge and discharge time period as a second sample health factor; and using the battery capacity of the battery pack in the fourth charge and discharge time period as a sample battery capacity; The capacity estimation model is constructed according to the second sample health factor and the sample battery capacity.

27. The device according to any one of claims 15 to 26, characterized in that The determining unit is specifically configured to: Obtaining a target charge and discharge number; the target charge and discharge number is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge; determining a difference between a calibrated battery capacity of the battery pack and a battery capacity of the battery pack during the second charge and discharge time period; Determining a ratio of the target charge and discharge times to the difference; The product of the ratio and the battery capacity of the battery pack in the second charge and discharge time period is used as the remaining service life of the battery pack.

28. The device according to any one of claims 15 to 26, characterized in that The determining unit is specifically configured to: Obtaining a target charge and discharge number; the target charge and discharge number is used to represent the total number of charge and discharge time periods in which the battery pack has undergone full charge and discharge; When determining that the battery capacity of the battery pack in the second charge and discharge time period is equal to a minimum capacity threshold, the number of charge and discharge times corresponding to the second charge and discharge time period; Taking the number of charge and discharge times as the actual service life of the battery pack; The difference between the actual service life and the target charge and discharge times is used as the remaining service life of the battery pack.

29. A battery module, characterized in that: The invention comprises a battery pack and a battery pack life prediction device as claimed in any one of claims 15 to 28, wherein when the battery pack is in operation, the battery pack life prediction device executes the method as claimed in any one of claims 1 to 14.

30. An energy storage system, characterized in that: The invention comprises a controller and the battery module as claimed in claim 29, wherein when the controller controls the operation of the battery module, the battery module executes the method according to any one of claims 1 to 14.

Images