Battery cell residual capacity prediction method and device, electronic equipment and storage medium

Abstract

The application discloses a kind of electric core residual capacity prediction method, device, electronic equipment and storage medium, belong to computer technical field.The method includes: obtaining the pressure sequence of target electric core in this round charge-discharge process;The pressure sequence includes multiple matrices;Each described matrix is used to indicate the pressure data of each area on the surface of the target electric core at a sampling time;Characteristic data is obtained by processing the pressure sequence based on variational mode decomposition algorithm and permutation entropy algorithm;The residual capacity of the target electric core is obtained by inputting the characteristic data into prediction model.The electric core residual capacity prediction method disclosed in the application can improve the accuracy of electric core residual capacity prediction.

Description

Technical Field

[0001] This application belongs to the field of computer technology, and in particular relates to a method, apparatus, electronic device and storage medium for predicting the remaining capacity of a battery cell. Background Technology

[0002] During battery use, the charging and discharging cycle process will cause adverse phenomena such as lithium deposition to gradually appear inside the battery, eventually leading to a sharp drop in the battery cell capacity. At this time, the prediction of battery cell capacity is particularly important for battery development and use.

[0003] Generally, the capacity of a battery cell is predicted based on information such as battery voltage and current. However, during the charging and discharging cycle of a battery, the graphite anode material expands and contracts as lithium ions insert and detach, and is accompanied by reactions such as gas production, which causes changes in the surface pressure of the cell. Therefore, the cell capacity predicted based on information such as battery voltage and current is not accurate. Summary of the Invention

[0004] This application aims to address at least one of the technical problems existing in the prior art. To this end, this application proposes a method, apparatus, electronic device, and storage medium for predicting the remaining capacity of battery cells, so as to improve the accuracy of predicting the remaining capacity of battery cells.

[0005] In a first aspect, this application provides a method for predicting the remaining capacity of a battery cell, the method comprising:

[0006] The pressure sequence of the target cell during this round of charging and discharging is obtained; the pressure sequence includes multiple matrices; each matrix is ​​used to indicate the pressure data of each region on the surface of the target cell at a sampling time.

[0007] The stress sequence is processed using variational mode decomposition algorithm and permutation entropy algorithm to obtain feature data;

[0008] The feature data is input into the prediction model to obtain the remaining capacity of the target battery cell.

[0009] According to the cell remaining capacity prediction method of this application, the pressure sequence of the target cell during the current charge and discharge process is obtained, and the pressure sequence is processed based on the variational mode decomposition algorithm and the permutation entropy algorithm to obtain feature data. The feature data is then input into the prediction model to obtain the remaining capacity of the target cell. Since the remaining capacity of the target cell is obtained by processing the pressure sequence, which includes the pressure data of each region on the surface of the target cell during the current charge and discharge process, and inputting it into the prediction model, the accuracy is higher than that of predicting the remaining capacity of the target cell by voltage and current.

[0010] According to one embodiment of this application, the prediction model includes: a gated recurrent unit, an encoding unit, and a transformation unit;

[0011] The gated loop unit includes an update gate and a reset gate; the reset gate is used to combine the data input to the gated loop unit at the current time step with the data already stored in the gated loop unit to update the output data of the gated loop unit; the update gate is used to determine the amount of stored data to be retained at the current time step.

[0012] According to the cell remaining capacity prediction method of this application, the prediction model includes a gated loop unit. The gated loop unit includes a reset gate that combines the data input to the gated loop unit at the current time step with the data already memorized by the gated loop unit, and an update gate that determines the amount of data to retain at the current time step. This enables the prediction model to understand the long-term dependencies in the data of the stress sequence and to pay attention to contextual information.

[0013] According to one embodiment of this application, a stress sequence is processed based on a variational mode decomposition algorithm and a permutation entropy algorithm to obtain feature data, including:

[0014] The pressure sequence is processed by variational mode decomposition algorithm to obtain multiple modes with different frequencies.

[0015] Multiple modalities are filtered using the permutation entropy algorithm to obtain feature data.

[0016] According to the cell remaining capacity prediction method of this application, after obtaining the pressure sequence of the target cell in this round of charging and discharging, the pressure sequence can be processed by variational mode decomposition algorithm to obtain multiple modes with different frequencies. This allows for the selection of multiple modes based on permutation entropy algorithm to obtain feature data, thereby achieving noise reduction and reconstruction of the pressure data in the pressure sequence. The feature data obtained from the noise reduction and reconstruction is then input into the prediction model, making the remaining capacity obtained by the prediction model more accurate.

[0017] According to one embodiment of this application, multiple modalities are filtered based on a permutation entropy algorithm to obtain feature data, including:

[0018] Based on the permutation entropy algorithm, the permutation entropy of each mode is obtained;

[0019] The modes whose permutation entropy is lower than a preset threshold are identified as feature data.

[0020] According to the battery cell remaining capacity prediction method provided in the embodiments of this application, after obtaining multiple modes with different frequencies, the multiple modes are screened based on the permutation entropy algorithm, and those modes with permutation entropy lower than a preset threshold are retained. In this way, modes with higher regularity and lower complexity are retained, and the signal components that contribute most to the analysis of battery cell capacity decay trend are selected to obtain feature data. This enables the denoising and reconstruction of pressure data in the pressure sequence, and then the feature data obtained by denoising and reconstruction is input into the prediction model, so that the remaining capacity obtained by the prediction model is more accurate.

[0021] According to one embodiment of this application, feature data is input into a prediction model to obtain the remaining capacity of the target battery cell, including:

[0022] Input the feature data into the gated loop unit and obtain the output data of the gated loop unit;

[0023] The output data of the gated loop unit is input into the encoding unit. Based on the position of the element in the output data of the gated loop unit, the output data of the gated loop unit is encoded to obtain the output data of the encoding unit.

[0024] The output data of the encoding unit is input into the conversion unit to obtain the remaining capacity of the target cell.

[0025] According to the cell remaining capacity prediction method provided in the embodiments of this application, after obtaining the output data of the gated loop unit, the encoding unit encodes the output data of the gated loop unit, injects corresponding position information into the data of each position in the output data of the gated loop unit, so as to help the model capture the sequential relationship of the elements in the data, so that the prediction unit predicts the remaining capacity of the target cell based on the output data of the encoding unit obtained through the encoding process, and the prediction result is more accurate.

[0026] According to one embodiment of this application, after inputting feature data into a prediction model to obtain the remaining capacity of the target battery cell, the method further includes:

[0027] Based on the remaining capacity of the target battery cell obtained this time and in the past, determine the health status of the target battery cell;

[0028] A warning signal is issued when the target battery cell is in an unhealthy state.

[0029] According to the cell remaining capacity prediction method provided in the embodiments of this application, after obtaining the remaining capacity of the target cell based on the prediction model, the health status of the target cell can be judged based on the remaining capacity of the target cell obtained this time and in the past, so that the health status of the target cell obtained based on the pressure data of the target cell is more accurate.

[0030] According to one embodiment of this application, determining the health status of the target battery cell based on the remaining capacity obtained this time and historically includes:

[0031] If the difference between the remaining capacity of the target cell obtained this time and the remaining capacity of the target cell obtained last time is greater than the target threshold, the health status of the target cell is determined to be unhealthy.

[0032] According to the cell remaining capacity prediction method provided in the embodiments of this application, after obtaining the remaining capacity of the target cell based on the prediction model, if the difference between the remaining capacity of the target cell obtained this time and the remaining capacity of the target cell obtained last time is greater than the target threshold, the health status of the target cell is determined to be unhealthy. This achieves the purpose of alarming the unhealthy status of the target cell by predicting the remaining capacity of the target cell based on the pressure data of the target cell.

[0033] Secondly, this application provides a battery cell remaining capacity prediction device, the device comprising:

[0034] The acquisition module is used to acquire the pressure sequence of the target cell during this round of charging and discharging; the pressure sequence includes multiple matrices; each matrix is ​​used to indicate the pressure data of each region on the surface of the target cell at a sampling time;

[0035] The processing module is used to process the pressure sequence based on the variational mode decomposition algorithm and the permutation entropy algorithm to obtain feature data;

[0036] The prediction module is used to input the feature data into the prediction model to obtain the remaining capacity of the target battery cell.

[0037] According to the battery cell remaining capacity prediction device of this application, the remaining capacity of the target battery cell is obtained by acquiring the pressure sequence of the target battery cell during the current charge and discharge process, processing the pressure sequence based on the variational mode decomposition algorithm and the permutation entropy algorithm to obtain feature data, and then inputting the feature data into the prediction model. Since the remaining capacity of the target battery cell is obtained by processing the pressure sequence including the pressure data of each region on the surface of the target battery cell during the current charge and discharge process and inputting it into the prediction model, the accuracy is higher than that of predicting the remaining capacity of the target battery cell by voltage and current.

[0038] Thirdly, this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the cell remaining capacity prediction method as described in the first aspect above.

[0039] Fourthly, this application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the cell remaining capacity prediction method as described in the first aspect above.

[0040] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the cell remaining capacity prediction method as described in the first aspect above.

[0041] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0042] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0043] Figure 1 This is a flowchart illustrating the cell remaining capacity prediction method provided in the embodiments of this application;

[0044] Figure 2 This is a schematic diagram of the pressure sensing system provided in the embodiments of this application;

[0045] Figure 3 This is a schematic diagram of the target cell pressure data acquisition system provided in an embodiment of this application;

[0046] Figure 4 This is a schematic diagram of the structure of the prediction model provided in the embodiments of this application;

[0047] Figure 5 This is one of the schematic diagrams comparing the predicted remaining capacity and the actual remaining capacity of the target battery cell provided in the embodiments of this application;

[0048] Figure 6 This is the second schematic diagram comparing the predicted remaining capacity and the actual remaining capacity of the target battery cell provided in the embodiments of this application;

[0049] Figure 7 This is a schematic diagram of the structure of the battery cell remaining capacity prediction device provided in the embodiments of this application;

[0050] Figure 8 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0051] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0052] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0053] The following description, in conjunction with the accompanying drawings, details the battery cell remaining capacity prediction method, battery cell remaining capacity prediction device, electronic device, and readable storage medium provided in this application through specific embodiments and application scenarios.

[0054] The method for predicting the remaining capacity of battery cells can be applied to a terminal, specifically executed by hardware or software within the terminal. This terminal includes, but is not limited to, portable communication devices such as mobile phones or tablets with touch-sensitive surfaces (e.g., touchscreen displays and / or touchpads). It should also be understood that, in some embodiments, the terminal may not be a portable communication device, but rather a desktop computer with touch-sensitive surfaces (e.g., touchscreen displays and / or touchpads).

[0055] The following embodiments describe a terminal including a display and a touch-sensitive surface. However, it should be understood that the terminal may include one or more other physical user interface devices such as a physical keyboard, mouse, and joystick.

[0056] The battery cell remaining capacity prediction method provided in this application embodiment can be executed by an electronic device or a functional module or functional entity in an electronic device that can implement the battery cell remaining capacity prediction method. The electronic devices mentioned in this application embodiment include, but are not limited to, mobile phones, tablets, computers, cameras and wearable devices. The following uses an electronic device as the execution subject to illustrate the battery cell remaining capacity prediction method provided in this application embodiment.

[0057] like Figure 1 As shown, the method for predicting the remaining capacity of a battery cell includes steps 110, 120, and 130.

[0058] As the number of charge-discharge cycles of a battery increases, the remaining capacity of the battery cell gradually decreases due to factors such as lithium insertion / extraction, which affects the health of the cell to some extent. Therefore, predicting the remaining cell capacity is particularly important for the development and use of batteries.

[0059] Step 110: Obtain the pressure sequence of the target cell during this round of charging and discharging; the pressure sequence includes multiple matrices; each matrix is ​​used to indicate the pressure data of each region on the surface of the target cell at a sampling time.

[0060] In actual execution, the target cell can be any battery cell, and the pressure sequence can be the pressure sequence collected by the target cell during any charging and discharging process.

[0061] In some embodiments, the sampling period can be any length, and multiple sampling moments can be determined during the charging and discharging process. For example, during the current charging and discharging process, the pressure data on the surface of the target cell can be sampled every 30 seconds with a sampling period of 30 seconds to obtain a matrix of sampling moments.

[0062] In practical implementation, a pressure sensing system can be used to acquire the pressure sequence of the target battery cell during this round of charging and discharging. The pressure sensing system can include a pressure sensor and the target battery cell; the pressure sensor can be a thin-film pressure sensor. The pressure sensor is placed in close contact with the target battery cell to collect pressure data during the charging and discharging process.

[0063] In some embodiments, the pressure sequence of the target battery cell during the current charge-discharge process can be obtained based on a thin-film pressure sensor. In some embodiments, the thin-film pressure sensor can be divided into multiple regions, and pressure data of different regions of the target battery cell at a sampling time can be obtained by multiple detection points on the thin-film pressure sensor, so as to obtain a matrix of pressure data of each region on the surface of the target battery cell collected at that sampling time. In this way, the matrix of the target battery cell at multiple sampling times during the current charge-discharge process can be obtained, so as to obtain the pressure sequence of the target battery cell during the current charge-discharge process.

[0064] In practical implementation, the pressure sequence can include a matrix, a list, a data string, or any theoretically feasible form. Taking a matrix as an example, each element in the matrix represents the pressure data of a region on the target cell at a sampling time. In some embodiments, the target cell can be divided into multiple regions, and the pressure signal of each region on the target cell can be acquired by various detection points on a thin-film pressure sensor. The pressure signal is then converted into an electrical signal through the piezoresistive effect, and the electrical signal is converted into an element at the corresponding position in the matrix through analog-to-digital conversion.

[0065] Step 120: Process the pressure sequence based on variational mode decomposition algorithm and permutation entropy algorithm to obtain feature data.

[0066] In actual implementation, the pressure sequence can be processed by variational mode decomposition algorithm to obtain the decomposition result, which can be multiple modes with different frequencies obtained by decomposing the pressure sequence.

[0067] In practice, after obtaining the decomposition results based on the variational mode decomposition algorithm, the decomposition results can be filtered based on the permutation entropy algorithm to select feature data that can be used as input to the prediction model.

[0068] In some embodiments, after obtaining multiple modes with different frequencies based on the variational mode decomposition algorithm, the multiple modes with different frequencies can be filtered based on the permutation entropy algorithm to obtain feature data.

[0069] Step 130: Input the feature data into the prediction model to obtain the remaining capacity of the target cell.

[0070] In practice, the prediction model can be trained using pressure sequences collected during the historical charging and discharging process of the target battery cell as samples, and the remaining capacity of the target battery cell after the end of the historical charging and discharging process as a label. This model is then used to predict the remaining capacity of the target battery cell. The prediction model can be based on a multi-head attention mechanism.

[0071] According to the cell remaining capacity prediction method of this application, the pressure sequence of the target cell during the current charge and discharge process is obtained, and the pressure sequence is processed based on the variational mode decomposition algorithm and the permutation entropy algorithm to obtain feature data. The feature data is then input into the prediction model to obtain the remaining capacity of the target cell. Since the remaining capacity of the target cell is obtained by processing the pressure sequence, which includes the pressure data of each region on the surface of the target cell during the current charge and discharge process, and inputting it into the prediction model, the accuracy is higher than that of predicting the remaining capacity of the target cell by voltage and current.

[0072] In some embodiments, the prediction model may include: a gated recurrent unit (GRU), an encoding unit, and a transformation unit; the gated recurrent unit includes an update gate and a reset gate; the reset gate is used to combine the data input to the gated recurrent unit at the current time step with the data already memorized by the gated recurrent unit to update the output data of the gated recurrent unit; the update gate is used to determine the amount of data to retain at the current time step.

[0073] In actual implementation, a transformation unit can be a unit that executes the Transformer algorithm.

[0074] In actual execution, the data input to the gating loop unit at the current time step can be the feature data obtained after processing the pressure sequence based on the variational mode decomposition algorithm and permutation entropy algorithm mentioned above. The reset gate can be used to combine the data input to the gating loop unit at the current time step with the memorized data to obtain the output data of the gating loop unit. The update gate can be used to determine the amount of memorized data to be retained at the current time step.

[0075] In some embodiments, after obtaining the output data of the gating loop unit based on the data input to the gating loop unit at the current time step, the output data of the gating loop unit can be encoded based on the encoding unit to obtain the encoding result, and then the encoding result can be input to the conversion unit to obtain the remaining capacity of the target cell.

[0076] According to the cell remaining capacity prediction method of this application, the prediction model includes a gated loop unit. The gated loop unit includes a reset gate that combines the data input to the gated loop unit at the current time step with the data already memorized by the gated loop unit, and an update gate that determines the amount of data to retain at the current time step. This enables the prediction model to understand the long-term dependencies in the data of the stress sequence and to pay attention to contextual information.

[0077] In some embodiments, after obtaining the pressure sequence of the target cell during this round of charging and discharging, the pressure sequence can be processed by variational mode decomposition algorithm to obtain multiple modes with different frequencies; then, the multiple modes can be filtered by permutation entropy algorithm to obtain feature data, and then the feature data can be input into the prediction model to obtain the remaining capacity of the target cell.

[0078] In some embodiments, the pressure sequence can be subjected to variational mode decomposition based on a variational mode decomposition model to obtain multiple modes with different frequencies.

[0079] In actual implementation, the variational mode decomposition model can be as follows:

[0080]

[0081] st∑ k u k (t) = c(t);

[0082] in, Let c(t) represent the square of the Euclidean norm, and c(t) represent the pressure sequence. ∑ k u k (t) represents the sum of modal components at different frequencies, where t represents time. Let δ(t) represent the partial derivative with respect to time, and let ω represent the Dirac distribution function. kThe center frequency of each modal component is represented by , and j represents the imaginary unit.

[0083] In some embodiments, during the variational mode decomposition of the pressure sequence based on the variational mode decomposition model, a penalty factor α and a Lagrange multiplier operator λ can be introduced. The alternating direction multiplier method (ADMM) is used to solve the unconstrained problem. By fixing variables, the optimal solution is approximated through iterative optimization. The iterative process is as follows:

[0084]

[0085] Where τ represents the noise margin.

[0086] According to the battery cell remaining capacity prediction method provided in the embodiments of this application, after obtaining the pressure sequence of the target battery cell in this round of charging and discharging, the pressure sequence can be processed by variational mode decomposition algorithm to obtain multiple modes with different frequencies. This allows for the selection of multiple modes based on permutation entropy algorithm to obtain feature data, thereby achieving noise reduction and reconstruction of the pressure data in the pressure sequence. The feature data obtained from the noise reduction and reconstruction is then input into the prediction model, making the remaining capacity obtained by the prediction model more accurate.

[0087] In some embodiments, after performing variational mode decomposition on the pressure sequence based on the variational mode decomposition algorithm to obtain multiple modes with different frequencies, the permutation entropy of each mode can be obtained based on the permutation entropy algorithm; and the modes with permutation entropy lower than a preset threshold are identified as feature data.

[0088] In actual implementation, after performing variational mode decomposition on the pressure sequence based on the variational mode decomposition algorithm to obtain multiple modes with different frequencies, each mode among these multiple modes with different frequencies is reconstructed to obtain the reconstructed components. Each mode among the multiple modes with different frequencies can be in the form of a sequence.

[0089] In actual execution, for each mode among multiple modes with different frequencies, after obtaining the reconstructed components of that mode, the reconstructed components can be arranged to obtain the arrangement number, and then the probability of the arrangement number appearing can be calculated. Based on the probability of the arrangement number appearing, the arrangement entropy of that mode can be obtained.

[0090] In some embodiments, the modal f(i), i = 1, 2, ..., n can be reconstructed as:

[0091] Fj=[f(j),f(j+t),…,f(j+(m-1)t)];

[0092] Where m is the number of sampling points for the mode, and t is the sampling time interval.

[0093] Sort Fj in ascending order, and we get:

[0094] F j =[f(j+(j1-1)t)≤f(j+(j2-1)t)≤…≤f(j+(j m -1)t)];

[0095] Calculate the probability of different permutations, where the probability of each sequence occurring is P. g The entropy of the mode arrangement is calculated as follows:

[0096]

[0097] In actual implementation, after obtaining the permutation entropy of each mode among multiple modes with different frequencies, modes with permutation entropy below a preset threshold are used as feature data. For example, modes with permutation entropy below 0.7 among multiple modes with different frequencies are used as feature data.

[0098] According to the battery cell remaining capacity prediction method provided in the embodiments of this application, after obtaining multiple modes with different frequencies, the multiple modes are screened based on the permutation entropy algorithm, and those modes with permutation entropy lower than a preset threshold are retained. In this way, modes with higher regularity and lower complexity are retained, and the signal components that contribute most to the analysis of battery cell capacity decay trend are selected to obtain feature data. This enables the denoising and reconstruction of pressure data in the pressure sequence, and then the feature data obtained by denoising and reconstruction is input into the prediction model, so that the remaining capacity obtained by the prediction model is more accurate.

[0099] In some embodiments, after acquiring feature data, the feature data can be input into a gated loop unit to acquire the output data of the gated loop unit; the output data of the gated loop unit can be input into an encoding unit, and the output data of the gated loop unit can be encoded based on the position of the element in the output data of the gated loop unit to acquire the output data of the encoding unit; the output data of the encoding unit can be input into a conversion unit to acquire the remaining capacity of the target cell.

[0100] In actual execution, after inputting feature data into the gated loop unit and obtaining its output data, the output data can be input into the encoding unit. Based on the position of each element in the gated loop unit's output data, the output data is encoded. Encoding can be performed using a position encoder, which generates a unique code for the position and dimension of each data point in the first set of data using a combination of sine and cosine functions.

[0101] In actual execution, the following formula can be used to apply it to each unit in the first data (e.g.) Perform position encoding:

[0102]

[0103] Among them, PE (pos,2i) Represents even-numbered position encoding, PE (pos,2i+1) This represents the encoding for odd-numbered positions, where pos is the position sequence number, i represents the sequence dimension, and d... model Indicates the encoder dimension. It is the position-encoded input.

[0104] According to the cell remaining capacity prediction method provided in the embodiments of this application, after obtaining the output data of the gated loop unit, the encoding unit encodes the output data of the gated loop unit, injects corresponding position information into each element in the output data of the gated loop unit, so as to help the model capture the sequential relationship of the elements in the data, so that the prediction unit predicts the remaining capacity of the target cell based on the output data of the encoding unit obtained through the encoding process, and the prediction result is more accurate.

[0105] In some embodiments, after obtaining the remaining capacity of the target cell, the health status of the target cell can be determined based on the remaining capacity of the target cell obtained this time and in the past. If the health status of the target cell is unhealthy, an early warning signal can be issued.

[0106] In actual implementation, the remaining capacity of the target battery cell obtained in the past can be the remaining capacity of the target battery cell obtained before this time, or it can be the remaining capacity of the target battery cell obtained multiple times before this time.

[0107] In some embodiments, a remaining capacity curve can be obtained based on the remaining capacity of the target cell obtained in the past. If the remaining capacity of the target cell obtained this time deviates significantly from the remaining capacity curve, the health status of the target cell is determined to be unhealthy, and an early warning signal is issued.

[0108] According to the cell remaining capacity prediction method provided in the embodiments of this application, after obtaining the remaining capacity of the target cell based on the prediction model, the health status of the target cell can be judged based on the remaining capacity of the target cell obtained this time and in the past, so that the health status of the target cell obtained based on the pressure data of the target cell is more accurate.

[0109] In some embodiments, if the difference between the remaining capacity of the target cell obtained this time and the remaining capacity of the target cell obtained last time is greater than a target threshold, the health status of the target cell is determined to be unhealthy.

[0110] In actual implementation, if the difference between the remaining capacity of the target cell obtained this time and the remaining capacity of the target cell obtained last time is less than or equal to the target threshold, the health status of the target cell is determined to be healthy.

[0111] According to the cell remaining capacity prediction method provided in the embodiments of this application, after obtaining the remaining capacity of the target cell based on the prediction model, if the difference between the remaining capacity of the target cell obtained this time and the remaining capacity of the target cell obtained last time is greater than the target threshold, the health status of the target cell is determined to be unhealthy. This achieves the purpose of alarming the unhealthy status of the target cell by predicting the remaining capacity of the target cell based on the pressure data of the target cell.

[0112] To better understand the cell remaining capacity prediction method provided in the embodiments of this application, further explanation is provided below. It should be understood that the following discussion is merely exemplary.

[0113] This application provides a method for predicting the remaining capacity of a battery cell, the specific steps of which are as follows:

[0114] S1: Obtain the pressure sequence of the target cell during this round of charging and discharging; the pressure sequence includes multiple matrices; each matrix is ​​used to indicate the pressure data of each region on the surface of the target cell at a sampling time.

[0115] In actual execution, the target cell can be any battery cell, and the pressure sequence can be the pressure sequence collected by the target cell during any charging and discharging process.

[0116] In some embodiments, the sampling period can be any length, and multiple sampling moments can be determined during the charging and discharging process. For example, during the current charging and discharging process, the pressure data on the surface of the target cell can be sampled every 30 seconds with a sampling period of 30 seconds to obtain a matrix of sampling moments.

[0117] In practical implementation, a pressure sensing system can be used to acquire the pressure sequence of the target battery cell during this round of charging and discharging. The pressure sensing system can include a pressure sensor and the target battery cell; the pressure sensor can be a thin-film pressure sensor. The pressure sensor is placed in close contact with the target battery cell to collect pressure data during the charging and discharging process.

[0118] In actual implementation, the pressure sensing system can be as follows: Figure 2As shown, the pressure sensing system may include an upper steel plate, a middle steel plate, a lower steel plate, a load cell 210, a thin-film pressure sensor 220, and a target battery cell 230. The target battery cell is located on the lower steel plate. The thin-film pressure sensor 220 is uniformly and flatly fixed on the surface of the target battery cell 230. The middle steel plate is located above the thin-film pressure sensor 220, and the load cell 210 is located above the middle steel plate, used to check whether the preload between the upper and middle steel plates meets the requirements. The upper steel plate is located above the load cell 210. The pressure sensing system may be located in a charge / discharge machine, and the target battery cell is charged and discharged by connecting to its two tabs.

[0119] In practical implementation, the pressure sequence of the target battery cell during this charge-discharge cycle can be obtained based on a thin-film pressure sensor. In some embodiments, the target pressure sensor can be divided into multiple regions, and pressure data of different regions of the target battery cell at a sampling time can be obtained by multiple detection points on the thin-film pressure sensor. This results in a matrix indicating the pressure data of each region on the surface of the target battery cell at that sampling time. In this way, the matrix of the target battery cell at multiple sampling times during this charge-discharge cycle can be obtained, thus obtaining the pressure sequence of the target battery cell during this charge-discharge cycle.

[0120] In actual implementation, such as Figure 3 As shown, several thin-film pressure sensors are connected to a data acquisition unit (such as a first-channel data acquisition unit and a second-channel data acquisition unit), and then connected to a target device via a USB expansion dock. The pressure data acquired by the thin-film pressure sensors is transmitted to the target device through the data acquisition unit and the USB expansion dock. The target device can be a PC.

[0121] In actual implementation, each element in the matrix represents the pressure data of a region on the target cell at a sampling time. In some embodiments, the target cell can be divided into multiple regions, and the pressure signal of each region on the target cell can be acquired by each detection point on the thin-film pressure sensor. The pressure signal is then converted into an electrical signal through the piezoresistive effect, and the electrical signal is converted into an element at the corresponding position in the matrix through analog-to-digital conversion.

[0122] S2: The pressure sequence is processed by variational mode decomposition algorithm to obtain multiple modes with different frequencies.

[0123] In some embodiments, the pressure sequence can be subjected to variational mode decomposition based on a variational mode decomposition model to obtain multiple modes with different frequencies.

[0124] In actual implementation, the variational mode decomposition model can be as follows:

[0125]

[0126] st∑ k u k (t) = c(t);

[0127] in, Let c(t) represent the square of the Euclidean norm, and c(t) represent the pressure sequence. ∑ k u k (t) represents the sum of modal components at different frequencies, where t represents time. Let δ(t) represent the partial derivative with respect to time, and let ω represent the Dirac distribution function. k The center frequency of each modal component is represented by , and j represents the imaginary unit.

[0128] In some embodiments, during the variational mode decomposition of the pressure sequence based on the variational mode decomposition model, a penalty factor α and a Lagrange multiplier operator λ can be introduced. The alternating direction multiplier method (ADMM) is used to solve the unconstrained problem. By fixing variables, the optimal solution is approximated through iterative optimization. The iterative process is as follows:

[0129]

[0130] Where τ represents the noise margin.

[0131] S3: Based on the permutation entropy algorithm, obtain the permutation entropy of each mode; determine the modes whose permutation entropy is lower than a preset threshold, and use them as feature data.

[0132] In actual implementation, after performing variational mode decomposition on the pressure sequence based on the variational mode decomposition algorithm to obtain multiple modes with different frequencies, each mode among these multiple modes with different frequencies is reconstructed to obtain the reconstructed components. Each mode among the multiple modes with different frequencies can be in the form of a sequence.

[0133] In actual execution, for each mode among multiple modes with different frequencies, after obtaining the reconstructed components of that mode, the reconstructed components can be arranged to obtain the arrangement number, and then the probability of the arrangement number appearing can be calculated. Based on the probability of the arrangement number appearing, the arrangement entropy of that mode can be obtained.

[0134] In some embodiments, the modal f(i), i = 1, 2, ..., n can be reconstructed as:

[0135] Fj=[f(j),f(j+t),...,f(j+(m-1)t)];

[0136] Where m is the number of sampling points for the mode, and t is the sampling time interval.

[0137] Sort Fj in ascending order, and we get:

[0138] F j =[f(j+(j1-1)t)≤f(j+(j2-1)t)≤…≤f(j+(j m -1)t)];

[0139] Calculate the probability of different permutations, where the probability of each sequence occurring is P. g The entropy of the mode arrangement is calculated as follows:

[0140]

[0141] In actual implementation, after obtaining the permutation entropy of each mode among multiple modes with different frequencies, modes with permutation entropy below a preset threshold are used as feature data. For example, modes with permutation entropy below 0.7 among multiple modes with different frequencies are used as feature data.

[0142] S4: Based on the improved Transformer model, establish a cell capacity prediction model.

[0143] In some embodiments, after acquiring the stress sequence, variational mode decomposition is performed on the stress sequence based on a variational mode decomposition model to obtain multiple modes with different frequencies. Then, the modes are filtered using a permutation entropy algorithm, and the filtered modes are used as feature data input. Figure 4 In the cell capacity prediction model shown.

[0144] In some embodiments, establishing a cell capacity prediction model may specifically include the following steps:

[0145] S4-1: For the feature data x input to the model t The output data of the gated loop unit is obtained after passing through the gated loop unit.

[0146] In actual execution, the gated loop unit consists of an update gate and a reset gate. The function of the reset gate is to input the data (which can be feature data x) into the gated loop unit at the current time step. t The update gate combines the data already stored in the gated recurrent unit (ROU) with the data stored in the RAU, updating the RAU's output data to prevent information loss in the deep network. The update gate is used to determine the amount of stored data to retain at the current time step.

[0147] S4-2: Perform position encoding on the output data of the gated loop unit to obtain the output data of the encoding unit.

[0148] In actual execution, after inputting feature data into the gated loop unit and obtaining its output data, the output data can be input into the encoding unit. Based on the position of each element in the gated loop unit's output data, the output data is encoded. Encoding can be performed using a position encoder, which generates a unique code for the position and dimension of each data point in the first set of data using a combination of sine and cosine functions.

[0149] In actual execution, the following formula can be used to apply it to each unit in the first data (e.g.) Perform position encoding:

[0150]

[0151] Among them, PE (pos,2i) Represents even-numbered position encoding, PE (pos,2i+1) This represents the encoding for odd-numbered positions, where pos is the position sequence number, i represents the sequence dimension, and d... model Indicates the encoder dimension. It is the position-encoded input.

[0152] S4-3: Construct a lifetime prediction model based on deep networks.

[0153] The Transformer model consists of two main parts: an encoder and a decoder. The encoder maps the input sequence into a series of continuous vector representations, while the decoder generates the target sequence based on the encoder's output. The complete Transformer architecture is formed by stacking multiple such encoder and decoder layers. In the encoder, each layer contains two key sublayers: a multi-head self-attention sublayer to capture complex dependencies within the input sequence, and a positionally fully connected feedforward network sublayer that performs non-linear transformations on the input features. The decoder has a similar structure to the encoder, but it additionally introduces a multi-head attention sublayer specifically designed to focus on the relationship between the encoder output and the decoder's own input, enhancing the model's understanding of the connections between sequences. Each attention sublayer is followed by a fully connected feedforward network, which uses the ReLU activation function to enrich the model's expressive power.

[0154] The second data, after positional encoding, is first fed into the multi-head attention module. The multi-head attention module consists of multiple parallel single-head attention units. Each single-head attention unit generates an attention score matrix representing the relationships between elements in the sequence by calculating the dot product between the query vector and the key vector. Subsequently, these scores are standardized using the Softmax function, thus transforming them into an attention weight distribution for each query vector. Finally, a weighted average is calculated based on these weights to generate the output representation of the self-attention mechanism. This process can be performed using the following formula:

[0155]

[0156] Q = X f W Q ;

[0157] K = X f W K ;

[0158] V = X f W v ;

[0159] Where Q represents the query matrix, K represents the key matrix, V represents the value matrix, and X represents the value matrix. f Denotes the characteristic matrix, W Q W represents the weight corresponding to the query matrix. K W represents the weights corresponding to the key matrix. V dk represents the weights corresponding to the value matrix, and dk is the scaling factor.

[0160] After each self-attention head computes its result, multi-head attention concatenates these results, thereby improving the representational power, stability, and robustness of the prediction model. The concatenation of the results from each self-attention head using multi-head attention can be represented as follows:

[0161] MultiHead(Q,K,V)=Concat(head1,...,head H )·W

[0162] Where W is the multi-head attention weight matrix, and headi is the calculation result of the i-th self-attention head.

[0163] The output of the multi-head attention mechanism, after layer normalization, passes through a feedforward neural network consisting of two fully connected layers and a ReLU activation function, and then undergoes another layer normalization process. Its output is:

[0164] FFN(MSA(θ))=LayerNorm(ReLU(MSA(θ)W1+b1)W2+b2)

[0165] This completes the output of a Transformer Encoder. After being processed by multiple Transformer Encoders, the output data is input into a three-layer fully connected layer for fitting to obtain the predicted value of the remaining capacity of the battery cell.

[0166] S5: Training and validating the prediction model, the specific steps are as follows:

[0167] S5-1: Use the training set iterator to load data from the training dataset in batches. Each batch contains features and corresponding labels for a specified number of samples, processing large amounts of data in batches to reduce memory consumption and improve computational efficiency.

[0168] S5-2: Input the loaded feature data into the model, and the model will generate predicted values ​​based on its learned parameters (weights and biases). These predicted values ​​are the regression results of the model on the input data.

[0169] S5-3: Calculate the loss function to measure the difference between the model's predicted value sequence and the actual label sequence. The loss functions used are: relative error (RE), mean absolute error (MAE), and root mean square error (RMSE).

[0170]

[0171] Among them, RUL pr RUL represents the predicted capacity value. tr y represents the actual capacity value. i This represents the predicted capacity value. This indicates the actual capacity value.

[0172] S5-5: Calculate the gradient by taking the partial derivative of the loss with respect to the weights, and use the Adam optimizer to update the model weights to implement the backpropagation process and minimize the loss.

[0173] S5-6: Iterate continuously until the accuracy of the training set meets the requirements or the set number of rounds is reached. Find the best-performing model, save the parameters, and use it as the final prediction model.

[0174] S6: Use a predictive model to predict the remaining capacity of the target battery cell and determine whether to issue a warning signal. The specific steps are as follows:

[0175] S6-1: Collect pressure distribution data of the target battery cell and calculate the feature data fed into the model using the methods in S2 and S3.

[0176] S6-2: Input the feature data into the prediction model and output the remaining capacity of the target cell.

[0177] In some embodiments, after obtaining the remaining capacity of the target cell, the health status of the target cell can be determined based on the remaining capacity of the target cell obtained this time and in the past. If the health status of the target cell is unhealthy, an early warning signal can be issued.

[0178] In practice, the remaining capacity of the target battery cell obtained historically can be the remaining capacity of the target battery cell obtained previously, or it can be the remaining capacity of the target battery cell obtained multiple times before this time. Based on the remaining capacity of the target battery cell obtained historically, a remaining capacity curve can be obtained. If the remaining capacity of the target battery cell obtained this time deviates significantly from the remaining capacity curve, the health status of the target battery cell is determined to be unhealthy, and an early warning signal is issued.

[0179] In some embodiments, the remaining capacity of the target battery cell obtained from multiple consecutive predictions can be compared with the actual remaining capacity of the target battery cell to obtain the difference in the remaining capacity of the target battery cell, so as to further update the prediction model and reduce the prediction error of the prediction model. Figure 5 , Figure 6 As shown, the relatively flat curve is the actual remaining capacity curve of the target cell, the curve with a larger variation is the predicted remaining capacity curve of the target cell, and the bar chart is the difference between the predicted remaining capacity and the actual remaining capacity of the target cell.

[0180] This application also provides a battery cell remaining capacity prediction device.

[0181] like Figure 7 As shown, the battery cell remaining capacity prediction device 700 includes: an acquisition module 710, a processing module 720, and a prediction module 730.

[0182] The acquisition module 710 is used to acquire the pressure sequence of the target cell during this round of charging and discharging; the pressure sequence includes multiple matrices; each matrix is ​​used to indicate the pressure data of each region on the surface of the target cell at a sampling time.

[0183] Processing module 720 is used to process the pressure sequence based on variational mode decomposition algorithm and permutation entropy algorithm to obtain feature data;

[0184] The prediction module 730 is used to input the feature data into the prediction model to obtain the remaining capacity of the target battery cell.

[0185] According to the battery cell remaining capacity prediction device of this application, the pressure sequence of the target battery cell during the current charge and discharge process is obtained. The pressure sequence is processed based on the variational mode decomposition algorithm and the permutation entropy algorithm to obtain feature data. The feature data is input into the prediction model to obtain the remaining capacity of the target battery cell. Since the remaining capacity of the target battery cell is obtained by processing the pressure sequence including the pressure data of each region on the surface of the target battery cell during the current charge and discharge process and inputting it into the prediction model, the accuracy is higher than that of predicting the remaining capacity of the target battery cell by voltage and current.

[0186] In some embodiments, according to one embodiment of this application, the prediction model includes: a gated loop unit, an encoding unit, and a transformation unit;

[0187] The gated loop unit includes an update gate and a reset gate; the reset gate is used to combine the data input to the gated loop unit at the current time step with the data already stored in the gated loop unit to update the output data of the gated loop unit; the update gate is used to determine the amount of stored data to be retained at the current time step.

[0188] In some embodiments, the processing module 720 includes:

[0189] The processing unit is used to perform variational mode decomposition on the pressure sequence based on the variational mode decomposition algorithm to obtain multiple modes with different frequencies.

[0190] The filtering unit is used to filter multiple modalities based on the permutation entropy algorithm to obtain feature data.

[0191] In some embodiments, the filtering unit is specifically used for:

[0192] Based on the permutation entropy algorithm, the permutation entropy of each mode is obtained;

[0193] The modes whose permutation entropy is lower than a preset threshold are identified as feature data.

[0194] In some embodiments, the prediction module 730 includes:

[0195] The first acquisition unit is used to input feature data into the gated loop unit and acquire the output data of the gated loop unit;

[0196] The second acquisition unit is used to input the output data of the gated loop unit into the encoding unit, and encode the output data of the gated loop unit based on the position of the element in the output data of the gated loop unit, thereby acquiring the output data of the encoding unit.

[0197] The third acquisition unit is used to input the output data of the encoding unit into the conversion unit to obtain the remaining capacity of the target cell.

[0198] In some embodiments, the cell remaining capacity prediction device 700 may further include:

[0199] The judgment module is used to determine the health status of the target battery cell based on the remaining capacity obtained this time and in the past.

[0200] The transmitting module is used to issue a warning signal when the target battery cell is in an unhealthy state.

[0201] In some embodiments, the determination module is specifically used for:

[0202] If the difference between the remaining capacity of the target cell obtained this time and the remaining capacity of the target cell obtained last time is greater than the target threshold, the health status of the target cell is determined to be unhealthy.

[0203] The battery cell remaining capacity prediction device in this application embodiment can be an electronic device or a component in an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, handheld computer, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook or personal digital assistant (PDA), etc. It can also be a server, network attached storage (NAS), personal computer (PC), television (TV), ATM or self-service machine, etc. The embodiments of this application do not specifically limit it.

[0204] The battery cell remaining capacity prediction device in this application embodiment can be a device with an operating system. This operating system can be a Microsoft (Windows) operating system, an Android operating system, an iOS operating system, or other possible operating systems; this application embodiment does not specifically limit it.

[0205] The battery cell remaining capacity prediction device 700 provided in this application embodiment can achieve... Figures 1 to 6 The various processes implemented in the method implementation examples will not be described again here to avoid repetition.

[0206] In some embodiments, such as Figure 8As shown, this application embodiment also provides an electronic device 800, including a processor 801, a memory 802, and a computer program stored in the memory 802 and executable on the processor 801. When the program is executed by the processor 801, it implements the various processes of the above-described battery cell remaining capacity prediction method embodiment and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0207] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0208] This application also provides a non-transitory computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described battery cell remaining capacity prediction method embodiment and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0209] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0210] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method for predicting the remaining capacity of battery cells.

[0211] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0212] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-described battery cell remaining capacity prediction method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0213] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0214] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0215] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0216] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0217] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0218] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for predicting the remaining capacity of a battery cell, characterized in that, include: The pressure sequence of the target battery cell during this round of charging and discharging is obtained; the pressure sequence includes multiple matrices; each matrix is ​​used to indicate the pressure data of each region on the surface of the target battery cell at a sampling time. The pressure sequence is processed using variational mode decomposition algorithm and permutation entropy algorithm to obtain feature data; The feature data is input into the prediction model to obtain the remaining capacity of the target battery cell.

2. The method for predicting the remaining capacity of a battery cell according to claim 1, characterized in that, The prediction model includes: a gated recurrent unit, an encoding unit, and a transformation unit; The gated loop unit includes an update gate and a reset gate; the reset gate is used to combine the data input to the gated loop unit at the current time step with the data already stored in the gated loop unit to update the output data of the gated loop unit; the update gate is used to determine the amount of stored data to be retained at the current time step.

3. The method for predicting the remaining capacity of a battery cell according to claim 1, characterized in that, The stress sequence is processed using the variational mode decomposition algorithm and the permutation entropy algorithm to obtain feature data, including: The pressure sequence is processed by variational mode decomposition algorithm to obtain multiple modes with different frequencies. The feature data is obtained by filtering multiple modes based on the permutation entropy algorithm.

4. The method for predicting the remaining capacity of a battery cell according to claim 3, characterized in that, The step of filtering multiple modalities based on the permutation entropy algorithm to obtain the feature data includes: Based on the permutation entropy algorithm, the permutation entropy of each mode is obtained; The modes whose permutation entropy is lower than a preset threshold are identified as the feature data.

5. The method for predicting the remaining capacity of a battery cell according to claim 2, characterized in that, The step of inputting the feature data into the prediction model to obtain the remaining capacity of the target battery cell includes: The feature data is input into the gated loop unit, and the output data of the gated loop unit is obtained. The output data of the gated loop unit is input into the encoding unit. Based on the position of the element in the output data of the gated loop unit, the output data of the gated loop unit is encoded to obtain the output data of the encoding unit. The output data of the encoding unit is input into the conversion unit to obtain the remaining capacity of the target battery cell.

6. The method for predicting the remaining capacity of a battery cell according to any one of claims 1 to 5, characterized in that, After inputting the feature data into the prediction model to obtain the remaining capacity of the target battery cell, the method further includes: Based on the remaining capacity of the target battery cell obtained this time and in the past, the health status of the target battery cell is determined; If the target battery cell is in an unhealthy state, a warning signal will be issued.

7. The method for predicting the remaining capacity of a battery cell according to claim 6, characterized in that, The determination of the health status of the target battery cell based on the remaining capacity obtained this time and historically includes: If the difference between the remaining capacity of the target battery cell obtained this time and the remaining capacity of the target battery cell obtained last time is greater than the target threshold, the health status of the target battery cell is determined to be unhealthy.

8. A battery cell remaining capacity prediction device, characterized in that, include: The acquisition module is used to acquire the pressure sequence of the target cell during this round of charging and discharging; the pressure sequence includes multiple matrices; each matrix is ​​used to indicate the pressure data of each region on the surface of the target cell at a sampling time; The processing module is used to process the pressure sequence based on the variational mode decomposition algorithm and the permutation entropy algorithm to obtain feature data; The prediction module is used to input the feature data into the prediction model to obtain the remaining capacity of the target battery cell.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the cell remaining capacity prediction method as described in any one of claims 1-7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the cell remaining capacity prediction method as described in any one of claims 1-7.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the cell remaining capacity prediction method as described in any one of claims 1-7.

Images