Battery life prediction method and device, vehicle and storage medium

Abstract

The invention relates to the technical field of batteries, in particular to a battery life prediction method and device, a vehicle and a storage medium, and the method comprises the steps: obtaining the adjacent historical data of a target battery at the current moment under the condition that the target battery is detected to be in a rapid attenuation stage based on the current capacity attenuation slope, and extracting health features of the target battery, obtaining an input feature vector to train a preset neural network, inputting the time sequence feature vector of the target battery into the trained neural network to carry out rolling prediction until a preset capacity limit value is reached, and obtaining a life prediction result. According to the embodiment of the invention, the rapid attenuation stage of the battery can be accurately identified, the attenuation stage approaching historical data is used for training the long-short-term memory neural network, the influence of early attenuation slow historical data of the battery on the capacity prediction precision is effectively avoided, and the residual life prediction accuracy is improved; the operation time and hardware cost of the model are reduced, and the method is safer and more efficient.

Description

Technical Field

[0001] The present application relates to the field of battery technology, and in particular to a battery life prediction method, device, vehicle and storage medium. Background Art

[0002] With the continuous development of new energy vehicles, in order to solve customers' mileage anxiety, fast charging with different rates has come into being. However, under fast charging conditions, the battery capacity decays slowly in the early stage and gradually accelerates in the later stage, which brings challenges to the prediction of the remaining battery life.

[0003] In the related technology, a data-driven model can be trained based on historical battery data, and predictions can be made through the model, thereby realizing the judgment of battery capacity decay and life calculation, which is more flexible than electrochemical models and equivalent circuit models.

[0004] However, in the related technology, the model training data includes historical data of slow capacity decay in the early stage of the battery and historical data of rapid capacity decay in the later stage, which affects the prediction accuracy of the trained model for the later rapid capacity decay, resulting in an increase in the computational load of the model's online prediction and an increase in operating costs, which reduces the accuracy and efficiency of battery life prediction and needs to be urgently resolved. Summary of the invention

[0005] The present application provides a battery life prediction method, device, vehicle and storage medium to solve the problem in the related art that model training data includes historical data of slow capacity decay in the early stage of the battery and historical data of rapid capacity decay in the later stage, which affects the prediction accuracy of the trained model for the later rapid capacity decay, resulting in an increase in the computational load of the model's online prediction, an increase in operating costs, and a reduction in the accuracy and efficiency of battery life prediction.

[0006] The first aspect of the present application provides a battery life prediction method, comprising the following steps: obtaining a current capacity decay slope of a target battery, and detecting whether the target battery is in a rapid decay stage according to the current capacity decay slope; when it is detected that the target battery is in the rapid decay stage, obtaining neighboring historical data of the target battery at the current moment, and extracting a first health feature and a second health feature of the target battery according to the neighboring historical data to obtain an input feature vector of a preset neural network; training the preset neural network based on the input feature vector to obtain a trained neural network, and inputting the time series feature vector of the target battery into the trained neural network for target prediction, until the output capacitance value of the trained neural network reaches a preset capacity limit, to obtain a life prediction result of the target battery.

[0007] Optionally, in one embodiment of the present application, detecting whether the target battery is in the rapid decay stage according to the current capacity decay slope includes: confirming the actual life type of the target battery using a target k-means algorithm; matching the capacity decay slope mean and capacity decay slope standard deviation of the target battery in a pre-constructed capacity decay slope data set based on the actual life type; if the current capacity decay slope is within one of the capacity decay slope standard deviations and is greater than the capacity decay slope mean, determining that the target battery is in the rapid decay stage.

[0008] Optionally, in one embodiment of the present application, before matching the capacity decay slope mean and capacity decay slope standard deviation of the target battery, it also includes: based on the sample battery data history set, obtaining multiple capacity decay slopes under the current life type, and using the multiple capacity decay slopes to calculate the capacity decay slope mean and capacity decay slope standard deviation corresponding to the current life type; traversing all life types to obtain the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types, and constructing the pre-constructed capacity decay slope data set according to the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types.

[0009] Optionally, in one embodiment of the present application, before using the target k-means algorithm to confirm the actual life type of the target battery, it also includes: based on the sample battery data history set, obtaining historical charging data of multiple batteries, and extracting first health characteristics and second health characteristics of the multiple batteries based on the historical charging data of the multiple batteries; using the first health characteristics and the second health characteristics of the multiple batteries to train a preset k-means algorithm to obtain the target k-means algorithm.

[0010] Optionally, in one embodiment of the present application, the time series feature vector of the target battery is input into the trained neural network for target prediction until the output capacitance value of the trained neural network reaches a preset capacity limit, thereby obtaining a life prediction result of the target battery, including: detecting whether the current remaining life of the target battery is less than a preset life threshold based on the life prediction result; in the case of detecting that the current remaining life is less than the preset life threshold, generating life warning information of the target battery, and sending the life warning information to at least one user interaction device.

[0011] A second aspect of the present application provides a battery life prediction device, comprising: a detection module, configured to obtain a current capacity decay slope of a target battery, and detect whether the target battery is in a rapid decay stage according to the current capacity decay slope;

[0012] an extraction module, configured to, when detecting that the target battery is in the rapid decay stage, obtain neighboring historical data of the target battery at a current moment, and extract a first health feature and a second health feature of the target battery according to the neighboring historical data to obtain an input feature vector of a preset neural network;

[0013] A prediction module is used to train the preset neural network based on the input feature vector to obtain a trained neural network, input the time series feature vector of the target battery into the trained neural network for target prediction, until the output capacitance value of the trained neural network reaches a preset capacity limit, and obtain the life prediction result of the target battery.

[0014] Optionally, in one embodiment of the present application, the detection module includes: a confirmation unit, used to confirm the actual life type of the target battery using a target k-means algorithm; a matching unit, used to match the capacity decay slope mean and capacity decay slope standard deviation of the target battery in a pre-constructed capacity decay slope data set based on the actual life type; and a determination unit, used to determine that the target battery is in the rapid decay stage when the current capacity decay slope is within one of the capacity decay slope standard deviations and is greater than the capacity decay slope mean.

[0015] Optionally, in one embodiment of the present application, the detection module further includes: a calculation unit, used to obtain multiple capacity decay slopes under the current life type based on the sample battery data history set before matching the capacity decay slope mean and capacity decay slope standard deviation of the target battery, and use the multiple capacity decay slopes to calculate the capacity decay slope mean and capacity decay slope standard deviation corresponding to the current life type; a traversal unit, used to traverse all life types, obtain the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types, and construct the pre-constructed capacity decay slope data set according to the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types.

[0016] Optionally, in one embodiment of the present application, the detection module also includes: an extraction unit, used to obtain historical charging data of multiple batteries based on the sample battery data history set before using the target k-means algorithm to confirm the actual life type of the target battery, and extract the first health feature and the second health feature of the multiple batteries based on the historical charging data of the multiple batteries; a training unit, used to train a preset k-means algorithm using the first health feature and the second health feature of the multiple batteries to obtain the target k-means algorithm.

[0017] Optionally, in one embodiment of the present application, the prediction module includes: a detection unit, used to detect whether the current remaining life of the target battery is less than a preset life threshold based on the life prediction result; a generation unit, used to generate life warning information of the target battery when it is detected that the current remaining life is less than the preset life threshold, and send the life warning information to at least one user interaction device.

[0018] A third aspect of the present application provides a vehicle, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the battery life prediction method as described in the above embodiment.

[0019] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned battery life prediction method.

[0020] A fifth aspect of the present application provides a computer program, which, when executed, implements the above battery life prediction method.

[0021] The embodiment of the present application can accurately identify the rapid decay stage of the battery, use the historical data of the decay stage to train the long short-term memory neural network, effectively avoid the impact of the historical data of the slow decay of the battery in the early stage on the accuracy of capacity prediction, improve the accuracy of the remaining life prediction, reduce the running time and hardware cost of the model, and be safer and more efficient. Thus, the problem in the related technology that the model training data includes the historical data of the slow decay of the battery's early capacity and the historical data of the rapid decay of the later capacity affects the prediction accuracy of the trained model for the later rapid decay capacity, resulting in an increase in the computational load of the model's online prediction, an increase in the operating cost, and a reduction in the accuracy and efficiency of the battery life prediction is solved.

[0022] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0024] Figure 1 A flowchart of a battery life prediction method provided according to an embodiment of the present application;

[0025] Figure 2 This is a schematic diagram of the architecture of a remaining life warning method based on a battery capacity rapid decay stage according to an embodiment of the present application;

[0026] Figure 3 is a schematic structural diagram of a battery life prediction device according to an embodiment of the present application;

[0027] Figure 4 It is a schematic diagram of the structure of a vehicle according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0029] The following describes the battery life prediction method, device, vehicle and storage medium of the embodiment of the present application with reference to the accompanying drawings. In view of the related art mentioned in the above background technology, the model training data includes historical data of slow capacity decay in the early stage of the battery and historical data of rapid capacity decay in the later stage, which affects the prediction accuracy of the model for the later rapid decay capacity after training, resulting in an increase in the calculation load of the model online prediction, an increase in the operating cost, and a reduction in the accuracy and efficiency of the battery life prediction. The present application provides a battery life prediction method, in which the rapid decay stage of the battery is accurately identified, and the long short-term memory neural network is trained using the historical data of the decay stage nearby, which effectively avoids the influence of the slow historical data of the early decay of the battery on the capacity prediction accuracy, improves the accuracy of the remaining life prediction, reduces the running time and hardware cost of the model, and is safer and more efficient. Thus, the problem that the model training data includes historical data of slow capacity decay in the early stage of the battery and historical data of rapid capacity decay in the later stage in the related art, affects the prediction accuracy of the model for the later rapid decay capacity after training, resulting in an increase in the calculation load of the model online prediction, an increase in the operating cost, and a reduction in the accuracy and efficiency of the battery life prediction is solved.

[0030] Specifically, Figure 1 A flowchart of a battery life prediction method provided in an embodiment of the present application.

[0031] like Figure 1 As shown, the battery life prediction method includes the following steps:

[0032] In step S101, the current capacity decay slope of the target battery is obtained, and whether the target battery is in a rapid decay stage is detected according to the current capacity decay slope.

[0033] It can be understood that in the embodiments of the present application, a capacity decay slope threshold is obtained for determining whether a battery has entered a rapid decay stage. The current capacity decay slope of the battery to be estimated, i.e., the target battery, is calculated in real time and compared with the threshold value range to confirm whether the target battery has entered a rapid decay stage.

[0034] Optionally, in one embodiment of the present application, detecting whether the target battery is in a rapid decay stage is performed based on the current capacity decay slope, including: confirming the actual life type of the target battery using a target k-means algorithm; matching the capacity decay slope mean and capacity decay slope standard deviation of the target battery in a pre-constructed capacity decay slope data set based on the actual life type; if the current capacity decay slope is within a capacity decay slope standard deviation and is greater than the capacity decay slope mean, determining that the target battery is in a rapid decay stage.

[0035] In the actual execution process, the target k-means algorithm can be pre-built to obtain the early historical data features of the target battery to extract the first health feature and the second health feature, and the target k-means algorithm can be input to calculate the Euclidean distance between the target battery and multiple initial clustering center points. The corresponding category is matched by the minimum Euclidean distance value to confirm the life type of the target battery, and the capacity decay slope mean and standard deviation are obtained by further matching the life type with the pre-built capacity decay slope data set. If the current capacity decay slope is greater than the capacity decay slope mean within the range covered by a capacity decay slope standard deviation, it is determined that the battery has entered the rapid decay stage.

[0036] Among them, the first health feature can be obtained by intercepting the battery charging data and extracting the time required for the battery to reach the [3.25-3.55V] voltage range during the charging process according to the formula as the first health feature:

[0037]

[0038] Among them, t1 represents the time when the battery voltage reaches 3.25V, t2 represents the time when the battery voltage reaches 3.55V, the subscript n represents the number of cycles, and the superscript 1 represents the first health feature. The second health feature can be extracted by the formula to extract the average voltage of the battery in the voltage range of [2.75-3.55] during the charging process as the second health feature:

[0039]

[0040] Among them, V i is a sequence of voltage measurement values ​​greater than 2.75 V and less than 3.55 V, k represents the number of voltage measurement values ​​greater than 2.75 V and less than 3.55 V, the subscript n represents the number of cycles, and the superscript 2 represents the second health feature.

[0041] Optionally, in one embodiment of the present application, before matching the capacity decay slope mean and capacity decay slope standard deviation of the target battery, it also includes: based on the sample battery data history set, obtaining multiple capacity decay slopes under the current life type, and using the multiple capacity decay slopes to calculate the capacity decay slope mean and capacity decay slope standard deviation corresponding to the current life type; traversing all life types to obtain the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types, and constructing a pre-constructed capacity decay slope data set according to the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types.

[0042] In the actual implementation process, first for a certain life type, based on the big data platform, the battery history data of this life type is extracted to determine the capacity average of a certain number of charge and discharge cycles (such as the first 50 charge and discharge cycles) in the initial stage of the battery, and for each cycle from the initial stage of the next cycle (such as 51 cycles) to 80% of the battery capacity, solve its decay slope relative to the initial capacity, and obtain multiple decay slopes. Using multiple decay slopes, calculate the mean of the battery capacity decay slope and the standard deviation of the battery capacity decay slope. The mean represents the overall battery decay level, which is defined within a standard deviation range. If the capacity decay slope is greater than the mean, the battery enters a rapid decay stage. By traversing the battery history data of all life types, a pre-constructed capacity decay slope data set is obtained.

[0043] Optionally, in one embodiment of the present application, before using the target k-means algorithm to confirm the actual life type of the target battery, it also includes: based on the sample battery data history set, obtaining historical charging data of multiple batteries, and extracting first health characteristics and second health characteristics of multiple batteries based on the historical charging data of the multiple batteries; using the first health characteristics and second health characteristics of the multiple batteries to train a preset k-means algorithm to obtain a target k-means algorithm.

[0044] It should be noted that the preset k-means algorithm can be set by those skilled in the art according to actual conditions and is not specifically limited here.

[0045] In the actual implementation process, the preset k-means algorithm is used as the initial algorithm for training, so that the algorithm can classify the early data in combination with the extracted health features and output the battery life type. Specifically, based on the sample battery data history set of the big data platform, the first health feature and the second health feature of multiple batteries can be input, k cluster centers can be selected (for example, set to 3), the battery history data can be divided into short-life, medium-life and long-life batteries, cluster training can be performed, and the target k-means algorithm can be obtained.

[0046] The embodiment of the present application can apply the k-means algorithm, utilize the early health characteristics of the battery, match the battery to be estimated with the corresponding class, and use the mean of the corresponding class of the battery to be estimated as the threshold to quickly identify the rapid decay stage of the battery capacity, thereby predicting the battery capacity only in the rapid decay stage, estimating the remaining life, and reducing the model running time cost.

[0047] In step S102, when it is detected that the target battery is in a rapid decay stage, the neighboring historical data of the target battery at the current moment is obtained, and the first health feature and the second health feature of the target battery are extracted based on the neighboring historical data to obtain an input feature vector of the preset neural network.

[0048] It should be noted that the preset neural network can be set by those skilled in the art according to actual conditions and is not specifically limited here.

[0049] In an embodiment of the present application, the neighboring historical data is the battery failure data within a short time range before the starting point, based on the moment when the target battery enters the rapid decay stage. The specific time range is set by a technician in this field and is not specifically limited here.

[0050] It can be understood that in an embodiment of the present application, when it is detected that the target battery is in a rapid decay stage, the neighboring historical data of the target battery at the current moment can be obtained, and the moment when the target battery enters the rapid decay stage is taken as the starting point. The battery failure data within a short time range before the starting point is obtained as the neighboring historical data, and the first health feature and the second health feature of the target battery are calculated based on the neighboring historical data, thereby generating an input feature vector for training a preset neural network.

[0051] In step S103, a preset neural network is trained based on the input feature vector to obtain a trained neural network, and the time series feature vector of the target battery is input into the trained neural network for target prediction until the output capacitance value of the trained neural network reaches the preset capacity limit, thereby obtaining a life prediction result of the target battery.

[0052] It should be noted that the preset capacity limit can be set by those skilled in the art according to actual conditions and is not specifically limited here.

[0053] It can be understood that in the embodiment of the present application, a double-layer long short-term memory neural network model can be built as a preset neural network, the obtained input feature vector can be used as the training data of the preset neural network, the training battery capacity can be used as the target value, and the error between the predicted value and the target value can be calculated by the loss function. In the network model training process, the loss function is defined as:

[0054]

[0055] Where S is the length of the training data sequence, R is the number of responses, and C ij is the target battery capacity value, y ij is the network prediction value. The error between the prediction value and the target value is calculated through the loss function. The error is back-propagated according to the gradient guidance of the loss function, and the weights and bias coefficients of the preset neural network are continuously corrected. When the error reaches the expected value, the training learning process ends, and the weight and bias coefficient matrix is ​​determined, thus obtaining the trained neural network model.

[0056] Furthermore, the trained neural network model is input with a time series feature vector, i.e., the historical charge and discharge cycle performance data of the target battery, and prediction is performed in a rolling manner until the output battery capacity reaches 80% of the initial capacity, thereby obtaining the remaining life RUL of the target battery, i.e., the number of charge and discharge cycles experienced from the current moment to the capacity dropping to 80%.

[0057] Based on the identification of the rapid decay stage, the present application can apply a long short-term memory neural network to use part of the historical data before the prediction starting point instead of all the historical data for model training, thereby avoiding the influence of the historical data of slow decay of the early battery on the accuracy of capacity prediction in the rapid decay stage, improving the prediction accuracy, and reducing the computing load of the big data platform.

[0058] Optionally, in one embodiment of the present application, the time series feature vector of the target battery is input into the trained neural network for target prediction until the output capacitance value of the trained neural network reaches a preset capacity limit, and a life prediction result of the target battery is obtained, including: detecting whether the current remaining life of the target battery is less than a preset life threshold based on the life prediction result; when it is detected that the current remaining life is less than the preset life threshold, generating life warning information of the target battery, and sending the life warning information to at least one user interaction device.

[0059] It should be noted that the preset lifespan threshold can be set by those skilled in the art according to actual conditions and is not specifically limited here.

[0060] During the actual implementation process, a certain number of charge and discharge cycles can be set as a preset life threshold, for example, it can be set to 60 charge and discharge cycles. When the remaining battery life calculated in real time is less than the preset remaining life threshold, the remaining life warning is triggered, and the life warning information of the target battery is generated. The life warning information is sent to at least one user interaction device, and the user is reminded to replace the battery in time to avoid safety hazards caused by rapid battery life decay.

[0061] like Figure 2As shown, the working content of the embodiment of the present application is described in detail below with a specific embodiment.

[0062] In the pre-construction stage, based on multiple battery historical data sets on the big data platform as samples, the mean and standard deviation of the capacity decay slope of historical data of batteries with different lifespans whose capacity drops to 80% of the initial capacity can be calculated to obtain the mean and standard deviation set of capacity decay slopes of batteries with different lifespans; and feature extraction is performed on the battery data on the big data platform, and based on the k-means algorithm, the early health characteristics of historical data of batteries with different lifespans are used to establish a preliminary life recognition model library.

[0063] In the actual application stage, the cloud data set of the big data platform can be used to intercept the early charging history data of the battery to be estimated, extract the health characteristics, and make a preliminary judgment on the estimated battery life through the life recognition model library, match the corresponding life category, and thus confirm the capacity decay slope mean and standard deviation for identifying the rapid decay stage of the battery capacity to be estimated; based on the rapid capacity decay stage, the long short-term memory neural network is applied to predict the historical data before the starting point, train the neural network, input the feature vector of the time series to be predicted into the model, predict the future capacity of the battery to 80% of the initial capacity, and calculate its remaining life; compare the remaining life with the set remaining life threshold, and when the remaining life is less than the threshold, trigger the life warning.

[0064] According to the battery life prediction method proposed in the embodiment of the present application, the rapid decay stage of the battery can be accurately identified, and the long short-term memory neural network can be trained using the historical data of the decay stage, which can effectively avoid the influence of the historical data of the slow decay of the battery in the early stage on the capacity prediction accuracy, improve the accuracy of the remaining life prediction, reduce the running time and hardware cost of the model, and be safer and more efficient. Thus, the problem in the related technology that the model training data includes the historical data of the slow decay of the battery's early capacity and the historical data of the rapid decay of the later capacity, which affects the prediction accuracy of the trained model for the later rapid decay capacity, resulting in an increase in the computational load of the model's online prediction, an increase in the operating cost, and a reduction in the accuracy and efficiency of the battery life prediction is solved.

[0065] Next, a battery life prediction device according to an embodiment of the present application is described with reference to the accompanying drawings.

[0066] Figure 3 It is a schematic diagram of the structure of a battery life prediction device according to an embodiment of the present application.

[0067] like Figure 3 As shown, the battery life prediction device 10 includes: a detection module 100 , an extraction module 200 and a prediction module 300 .

[0068] The detection module 100 is used to obtain the current capacity decay slope of the target battery, and detect whether the target battery is in a rapid decay stage according to the current capacity decay slope.

[0069] The extraction module 200 is used to obtain the neighboring historical data of the target battery at the current moment when it is detected that the target battery is in a rapid decay stage, and extract the first health feature and the second health feature of the target battery based on the neighboring historical data to obtain an input feature vector of a preset neural network.

[0070] The prediction module 300 is used to train a preset neural network based on an input feature vector to obtain a trained neural network, input the time series feature vector of the target battery into the trained neural network for target prediction, until the output capacitance value of the trained neural network reaches a preset capacity limit, and obtain the life prediction result of the target battery.

[0071] Optionally, in one embodiment of the present application, the detection module 100 includes: a confirmation unit and a matching unit.

[0072] The confirmation unit is used to confirm the actual life type of the target battery using the target k-means algorithm.

[0073] The matching unit is used to match the capacity fade slope mean and the capacity fade slope standard deviation of the target battery in a pre-built capacity fade slope data set based on the actual life type.

[0074] The determination unit is used to determine that the target battery is in a rapid decay stage when the current capacity decay slope is within a capacity decay slope standard deviation and is greater than the capacity decay slope mean.

[0075] Optionally, in one embodiment of the present application, the detection module 100 further includes: a calculation unit and a traversal unit.

[0076] Among them, the calculation unit is used to obtain multiple capacity decay slopes under the current life type based on the sample battery data history set before matching the capacity decay slope mean and capacity decay slope standard deviation of the target battery, and use the multiple capacity decay slopes to calculate the capacity decay slope mean and capacity decay slope standard deviation corresponding to the current life type.

[0077] The traversal unit is used to traverse all life types, obtain the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types, and construct a pre-constructed capacity decay slope data set according to the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types.

[0078] Optionally, in one embodiment of the present application, the detection module 100 further includes: an extraction unit and a training unit.

[0079] Among them, the extraction unit is used to obtain historical charging data of multiple batteries based on a sample battery data history set before using the target k-means algorithm to confirm the actual life type of the target battery, and extract the first health feature and the second health feature of the multiple batteries based on the historical charging data of the multiple batteries.

[0080] The training unit is used to train a preset k-means algorithm using the first health characteristics and the second health characteristics of the plurality of batteries to obtain a target k-means algorithm.

[0081] Optionally, in one embodiment of the present application, the prediction module 300 includes: a detection unit and a generation unit.

[0082] The detection unit is used to detect whether the current remaining life of the target battery is less than a preset life threshold based on the life prediction result;

[0083] The generating unit is used to generate life warning information of the target battery and send the life warning information to at least one user interaction device when detecting that the current remaining life is less than a preset life threshold.

[0084] It should be noted that the above explanation of the battery life prediction method embodiment is also applicable to the battery life prediction device of this embodiment, and will not be repeated here.

[0085] According to the battery life prediction device proposed in the embodiment of the present application, the rapid decay stage of the battery can be accurately identified, and the long short-term memory neural network can be trained using the historical data of the decay stage, which effectively avoids the influence of the historical data of the slow decay of the battery in the early stage on the capacity prediction accuracy, improves the accuracy of the remaining life prediction, reduces the running time and hardware cost of the model, and is safer and more efficient. Thus, the problem in the related technology that the model training data includes the historical data of the slow decay of the battery's early capacity and the historical data of the rapid decay of the later capacity affects the prediction accuracy of the trained model for the later rapid decay capacity, resulting in an increase in the computational load of the model's online prediction, an increase in the operating cost, and a reduction in the accuracy and efficiency of the battery life prediction is solved.

[0086] Figure 4 A schematic diagram of the structure of a vehicle provided in an embodiment of the present application. The vehicle may include:

[0087] Memory 401 , processor 402 , and a computer program stored in the memory 401 and executable on the processor 402 .

[0088] When the processor 402 executes the program, the battery life prediction method provided in the above embodiment is implemented.

[0089] Furthermore, the vehicle also includes:

[0090] The communication interface 403 is used for communication between the memory 401 and the processor 402 .

[0091] The memory 401 is used to store computer programs that can be executed on the processor 402 .

[0092] The memory 401 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0093] If the memory 401, the processor 402 and the communication interface 403 are implemented independently, the communication interface 403, the memory 401 and the processor 402 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0094] Optionally, in a specific implementation, if the memory 401, the processor 402 and the communication interface 403 are integrated on a chip, the memory 401, the processor 402 and the communication interface 403 can communicate with each other through an internal interface.

[0095] The processor 402 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0096] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the battery life prediction method as described above is implemented.

[0097] This embodiment also provides a computer program, which implements the above battery life prediction method when executed.

[0098] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0099] In addition, 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 technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0100] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0101] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or N wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways as necessary and then storing it in a computer memory.

[0102] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0103] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0104] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0105] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for predicting battery life, characterized in that: The following steps are involved: Acquire a current capacity decay slope of a target battery, and detect whether the target battery is in a rapid decay stage according to the current capacity decay slope; When it is detected that the target battery is in the rapid decay stage, obtaining neighboring historical data of the target battery at the current moment, and extracting a first health feature and a second health feature of the target battery according to the neighboring historical data to obtain an input feature vector of a preset neural network; The preset neural network is trained based on the input feature vector to obtain a trained neural network, and the time series feature vector of the target battery is input into the trained neural network for target prediction until the output capacitance value of the trained neural network reaches a preset capacity limit, thereby obtaining a life prediction result of the target battery.

2. The method according to claim 1, characterized in that The detecting, according to the current capacity decay slope, whether the target battery is in a fast decay stage comprises: Determining the actual life type of the target battery using a target k-means algorithm; Based on the actual life type, matching the capacity fade slope mean and the capacity fade slope standard deviation of the target battery in a pre-constructed capacity fade slope data set; If the current capacity decay slope is within one of the capacity decay slope standard deviations and is greater than the capacity decay slope mean, it is determined that the target battery is in the rapid decay stage.

3. The method according to claim 2, characterized in that Before matching the capacity decay slope mean and capacity decay slope standard deviation of the target battery, the method further includes: Based on the sample battery data history set, multiple capacity decay slopes under the current life type are obtained, and the capacity decay slope mean and capacity decay slope standard deviation corresponding to the current life type are calculated using the multiple capacity decay slopes; Traverse all life types to obtain the capacity decay slope means and capacity decay slope standard deviations corresponding to all life types, and construct the pre-constructed capacity decay slope data set according to the capacity decay slope means and capacity decay slope standard deviations corresponding to all life types.

4. The method according to claim 3, characterized in that Before using the target k-means algorithm to confirm the actual life type of the target battery, it also includes: Based on the sample battery data history set, acquiring historical charging data of a plurality of batteries, and extracting first health features and second health features of the plurality of batteries according to the historical charging data of the plurality of batteries; The preset k-means algorithm is trained using the first health characteristics and the second health characteristics of the plurality of batteries to obtain the target k-means algorithm.

5. The method according to claim 1, characterized in that The step of inputting the time series feature vector of the target battery into the trained neural network for target prediction until the output capacitance value of the trained neural network reaches a preset capacity limit value, thereby obtaining a life prediction result of the target battery, comprises: Detecting whether the current remaining life of the target battery is less than a preset life threshold based on the life prediction result; When it is detected that the current remaining life is less than the preset life threshold, life warning information of the target battery is generated, and the life warning information is sent to at least one user interaction device.

6. A battery life prediction device, characterized in that: include: A detection module, used to obtain a current capacity decay slope of a target battery, and detect whether the target battery is in a rapid decay stage according to the current capacity decay slope; an extraction module, configured to, when detecting that the target battery is in the rapid decay stage, obtain neighboring historical data of the target battery at a current moment, and extract a first health feature and a second health feature of the target battery according to the neighboring historical data to obtain an input feature vector of a preset neural network; A prediction module is used to train the preset neural network based on the input feature vector to obtain a trained neural network, input the time series feature vector of the target battery into the trained neural network for target prediction, until the output capacitance value of the trained neural network reaches a preset capacity limit, and obtain the life prediction result of the target battery.

7. The device according to claim 6, characterized in that The detection module comprises: A confirmation unit, used to confirm the actual life type of the target battery using a target k-means algorithm; A matching unit, configured to match the capacity decay slope mean and the capacity decay slope standard deviation of the target battery in a pre-constructed capacity decay slope data set based on the actual life type; A determination unit is used to determine that the target battery is in the rapid decay stage when the current capacity decay slope is within one of the capacity decay slope standard deviations and is greater than the capacity decay slope mean.

8. The device according to claim 7, characterized in that The detection module also includes: A calculation unit, configured to obtain a plurality of capacity decay slopes under a current life type based on a sample battery data history set before matching the capacity decay slope mean and capacity decay slope standard deviation of the target battery, and calculate the capacity decay slope mean and capacity decay slope standard deviation corresponding to the current life type using the plurality of capacity decay slopes; A traversal unit is used to traverse all life types, obtain the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types, and construct the pre-constructed capacity decay slope data set according to the capacity decay slope mean and capacity decay slope standard deviation corresponding to all life types.

9. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the battery life prediction method according to any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the battery life prediction method as described in any one of claims 1 to 5.

Images