Battery life prediction method based on imaged single-cycle sampling quantity and related device

Abstract

The application relates to a battery life prediction method based on an imaged single-cycle sampling quantity and a related device. The method comprises the following steps: sampling a battery sampling quantity during a single charging and discharging cycle of the battery, obtaining a plurality of single-cycle sampling quantities of the battery, and constituting a single-cycle sampling quantity sequence; dividing the single-cycle sampling quantity sequence into a plurality of equal-length subsequences; calculating the distance between each subsequence to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence; and inputting the sampling quantity image into a battery life prediction model to obtain the expected cycle life of the battery. The application can accurately predict the remaining service life of the battery in the super-early stage of battery use.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of batteries, in particular to a battery life prediction method based on an image single cycle sampling quantity and a related device. BACKGROUND

[0002] With the increasing popularity of electric vehicles and renewable energy, lithium-ion batteries have gradually become the first choice for electrified transportation and energy storage systems due to their high energy density, low self-discharge rate, and long cycle life, among other advantages. The types of lithium-ion batteries include, but are not limited to, lithium iron phosphate batteries, ternary lithium batteries, lithium titanate batteries, etc. In addition, to meet the diverse needs of different application scenarios, other types of battery technologies such as sodium-ion batteries, lithium metal batteries, semi-solid state batteries, and solid state batteries are also continuously developing. However, as the performance of the battery gradually deteriorates over time, it leads to a series of safety hazards and economic losses. Therefore, how to accurately estimate the state of health (SOH) of the battery and predict its remaining useful life (RUL) has become a key technical problem to ensure operational efficiency and safety.

[0003] Currently, traditional battery life prediction methods mainly include physical model-based and / or data-driven methods. The physical model usually relies on a deep understanding of the battery degradation mechanism, but due to its large computational requirements, it cannot be efficiently applied to actual battery life prediction. Data-driven methods, especially machine learning and deep learning algorithms, require analyzing a large amount of historical data to achieve relatively accurate prediction results. Once the data volume is insufficient, it will seriously affect the accuracy of the prediction results. Therefore, a battery life prediction method based on an image single cycle sampling quantity and a related device are needed. SUMMARY

[0004] The present application provides a battery life prediction method based on an image single cycle sampling quantity and a related device, which improves the problem that the prior art cannot accurately and quickly predict the remaining useful life of the battery due to the need for a large amount of historical data or complex physical model to predict the battery life.

[0005] The present application provides a battery life prediction method based on an image single cycle sampling quantity, which comprises: sampling the battery sampling quantity during a single charge-discharge cycle of the battery, obtaining a plurality of single cycle sampling quantities of the battery, constituting a single cycle sampling quantity sequence; dividing the single cycle sampling quantity sequence into a plurality of equal-length subsequences; calculating the distance between each subsequence to generate a sampling quantity image corresponding to the single cycle sampling quantity sequence; inputting the sampling quantity image into a battery life prediction model to obtain the expected cycle life of the battery.

[0006] In an embodiment of the present application, the step of dividing the single-cycle sampling quantity sequence into multiple equal-length subsequences comprises: performing size normalization on the length of the single-cycle sampling quantity sequence according to a preset sequence length; performing data normalization on the single-cycle sampling quantity sequence after size normalization; and performing sliding sampling on the single-cycle sampling quantity sequence after data normalization according to a preset sampling length to obtain multiple equal-length subsequences.

[0007] In an embodiment of the present application, the step of performing size normalization on the length of the single-cycle sampling quantity sequence according to a preset sequence length comprises: counting the length of the single-cycle sampling quantity sequence; if the length of the single-cycle sampling quantity sequence is greater than the preset sequence length, extracting multiple single-cycle sampling quantities from the single-cycle sampling quantity sequence based on a downsampling method to form a single-cycle sampling quantity sequence with the preset sequence length; and if the length of the single-cycle sampling quantity sequence is less than the preset sequence length, expanding the single-cycle sampling quantity sequence based on an interpolation method to form a single-cycle sampling quantity sequence with the preset sequence length.

[0008] In an embodiment of the present application, the single-cycle sampling quantity is one kind, and the step of calculating the distance between each subsequence to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence comprises: calculating the distance between each subsequence to generate a distance matrix; and mapping each distance value in the distance matrix to a gray value to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence.

[0009] In an embodiment of the present application, the single-cycle sampling quantity is multiple kinds, and the step of calculating the distance between each subsequence to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence comprises: for each kind of single-cycle sampling quantity, calculating the distance between each subsequence to generate a distance matrix; mapping each distance value in the distance matrix to a gray value to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence; and stacking the sampling quantity images of all single-cycle sampling quantity sequences according to the channel dimension to form a multi-channel sampling quantity image and taking the multi-channel sampling quantity image as the sampling quantity image of all single-cycle sampling quantity sequences.

[0010] In an embodiment of the present application, the battery life prediction model is AlexNet, the battery life prediction model is one, and the step of inputting the sampling quantity image into the battery life prediction model to obtain the expected cycle life of the battery comprises: inputting the sampling quantity image into a convolution network of the battery life prediction model to extract image features of the sampling quantity image layer by layer; wherein the convolution network comprises multiple cascaded convolution layers; and inputting the image features generated by the last convolution layer of the convolution network into a fully connected network of the battery life prediction model to obtain the expected cycle life of the battery.

[0011] In an embodiment of the present application, the battery life prediction model is AlexNet, the battery life prediction model is multiple, the network architectures of the multiple battery life prediction models are the same but the parameters are different, and the inputting of the sampling quantity image into the battery life prediction model to obtain the expected cycle life of the battery comprises: for each battery life prediction model, inputting the sampling quantity image into a convolution network of the battery life prediction model, and extracting image features of the sampling quantity image layer by layer; wherein the convolution network comprises multiple cascaded convolution layers; inputting the image features generated by the last convolution layer of the convolution network into a fully connected network of the battery life prediction model, and mapping the image features into one-dimensional feature vectors; flattening and splicing each feature vector, inputting the spliced feature vector into a fully connected layer, and obtaining the expected cycle life of the battery.

[0012] In an embodiment of the present application, a battery life prediction device based on an imaged single-cycle sampling quantity is also provided, which comprises: a data acquisition module configured to sample the sampling quantity of a battery during a single charging and discharging cycle of the battery, and obtain multiple single-cycle sampling quantities of the battery, thereby forming a single-cycle sampling quantity sequence; a sequence segmentation module configured to segment the single-cycle sampling quantity sequence into multiple equal-length sub-sequences; an imaging module configured to calculate the distance between each sub-sequence, and generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence; and a battery life prediction module configured to input the sampling quantity image into a battery life prediction model to obtain the expected cycle life of the battery.

[0013] In an embodiment of the present application, an electronic device is also provided, which comprises: one or more processors; and a storage device configured to store one or more programs, which, when executed by the one or more processors, cause the electronic device to implement the battery life prediction method based on an imaged single-cycle sampling quantity according to any one of the above embodiments.

[0014] In an embodiment of the present application, a computer-readable storage medium having a computer program stored thereon is also provided, which, when executed by a processor of a computer, causes the computer to execute the battery life prediction method based on an imaged single-cycle sampling quantity according to any one of the above embodiments.

[0015] As described above, the battery life prediction method based on the image single cycle sampling quantity and the related device provided by the present application have the following beneficial effects: the battery sampling quantity is collected during the single charging and discharging cycle of the battery, the single cycle sampling quantity sequence is constructed, and the single cycle sampling quantity sequence is divided into multiple equal-length subsequences. By calculating the distance between each subsequence, the sampling quantity image is generated, and the battery life prediction model is called to realize the prediction of the remaining cycle life of the battery. The present application does not need to rely on complex physical models or a large amount of historical data, and only needs to use the battery sampling quantity during the single charging and discharging cycle to realize the accurate prediction of the remaining service life of the battery at any time. Even the remaining service life of the battery can be accurately predicted in the super early stage of the battery use, which effectively improves the prediction accuracy and efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of the battery life prediction method based on the image single cycle sampling quantity provided by the embodiment of the present application is shown.

[0017] Figure 2 A current, voltage and capacity curve diagram provided by the embodiment of the present application is shown.

[0018] Figure 3a A visualization image of the distance matrix of the voltage provided by the embodiment of the present application is shown.

[0019] Figure 3b A visualization image of the distance matrix of the current provided by the embodiment of the present application is shown.

[0020] Figure 3c A visualization image of the distance matrix of the capacity provided by the embodiment of the present application is shown.

[0021] Figure 4 A general flowchart provided by the embodiment of the present application is shown.

[0022] Figure 5 A structure block diagram of the battery life prediction device based on the image single cycle sampling quantity provided by the embodiment of the present application is shown.

[0023] Figure 6 A structure diagram of the electronic device provided by the embodiment of the present application is shown. DETAILED DESCRIPTION

[0024] Following make the embodiments of the present application more clear through specific, concrete examples, and other advantages and effects of the present application can be easily understood by those skilled in the art from the disclosure of the specification. The present application can also be implemented or applied through other different specific embodiments, and various modifications or changes can be made to the details in the specification based on different views and applications without departing from the spirit of the present application. It should be noted that the following examples and features in the examples can be combined with each other without conflict.

[0025] It should be noted that the diagrams provided in the following examples only illustrate the basic concepts of the present application in a schematic manner, and only the components related to the present application are shown in the diagrams, not the number, shape and size of the components when actually implemented. The actual implementation of each component may be a random change in type, number and proportion, and the layout pattern of the components may also be more complex.

[0026] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application, however, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details, and in other embodiments, structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present application difficult to understand.

[0027] The present application provides a battery life prediction method based on imaging single-cycle sampling quantity, which collects battery sampling quantity during single charging and discharging cycle of the battery, constructs single-cycle sampling quantity sequence, and divides it into multiple equal-length subsequences. By calculating the distance between each subsequence, a sampling quantity image is generated, and a battery life prediction model is called to predict the remaining cycle life of the battery. The present application does not need to rely on complex physical models or a large amount of historical data, but only needs to use the battery sampling quantity during single charging and discharging cycle, to accurately predict the remaining service life of the battery in the ultra-early stage of battery use, effectively improving the accuracy and efficiency of prediction.

[0028] Please refer to Figure 1 The battery life prediction method based on imaging single-cycle sampling quantity includes the following steps:

[0029] S1, sampling the battery sampling quantity during single charging and discharging cycle of the battery, obtaining multiple single-cycle sampling quantities of the battery, and constituting a single-cycle sampling quantity sequence.

[0030] A single charge-discharge cycle refers to the complete cycle of a battery going from a fully discharged state (0% capacity) to a fully charged state (100% capacity) through a charging process, and then discharging back to a fully discharged state. Correspondingly, a single-cycle sampling quantity refers to the battery sampling quantity during a single charge-discharge cycle. This sampling quantity includes, but is not limited to, current, voltage, and capacity. Capacity refers to the cumulative charge-discharge capacity, which gradually increases during charging and gradually decreases during discharging. During a single charge-discharge cycle, the battery sampling quantity is recorded by a sampling device according to a preset sampling interval. The battery sampling quantities are then arranged sequentially according to the sampling time to generate a time-series single-cycle sampling quantity sequence. As an example, if the battery sampling quantity is current, then the single-cycle current sequence can be I = {I1, I2, I3, ..., I...} n}, where I n Let be the current at the nth sampling point. It can be understood that the start and end times of a single charge-discharge cycle and the corresponding sample values ​​can be determined based on the battery sampling data. For example, taking battery sampling data as voltage, the charging start point is the minimum voltage value, and the discharging end point is when the voltage drops back to the minimum voltage. The time interval between two adjacent minimum voltage values ​​and their corresponding sample data are used as the sampling data recording range for a single charge-discharge cycle. As an example, such as... Figure 2 As shown, a single charge-discharge cycle begins at time 0 and ends at the 50th minute. During this period, charging occurs first, with the voltage gradually rising from a minimum of 2V to a peak. Then, discharging occurs, with the voltage gradually decreasing from the peak back to a minimum of 2V. The trends of current and capacity are similar to those of voltage and will not be detailed here.

[0031] It should be noted that the battery types described in this application include, but are not limited to, lithium-ion batteries, sodium-ion batteries, lithium metal batteries, semi-solid-state batteries, and solid-state batteries. Lithium-ion battery types include, but are not limited to, lithium iron phosphate batteries, ternary lithium batteries, and lithium titanate batteries. Preferably, this application uses lithium-ion batteries. Furthermore, considering that sampling points for different sampling amounts may differ during the sampling period, to facilitate subsequent processing, after constructing the single-cycle sampling sequence, the method further includes aligning the sampling times of different types of single-cycle sampling amounts.

[0032] S2. Divide the single-cycle sampling sequence into multiple subsequences of equal length.

[0033] For each type of single-cycle sampling, starting from the beginning of the sequence, a sliding window of preset length W is used to move along the sequence according to a preset step size S, dividing it into multiple subsequences of length W, as shown in formula (1):

[0034]

[0035] Where N is the total number of subsequences, L is the length of the single-cycle sampling quantity sequence, W is the size of the sliding window (i.e., the number of data points included in each subsequence), S is the step size (i.e., the size by which the sliding window moves backward each time), represents rounding down. It should be noted that for the same single-cycle sampling quantity sequence, the division method of the subsequences in the present invention can be any of the following: 1) There is no sampling quantity overlap between the subsequences; 2) There is a sampling quantity interval between the subsequences; 3) There is a sampling quantity overlap between the subsequences. Those skilled in the art can adaptively select the specific division method based on the actual performance of the battery, and it is not limited here.

[0036] Specifically, for the first division method: when S = W, that is, when the step size is the same as the size of the sliding window, the subsequences are adjacent to each other, and there is no sampling quantity overlap or interval between them. For example, if the length L of the single-cycle sampling quantity sequence is 9, the size W of the sliding window and the step size S are both 3, then the divided subsequences are [a1, a2, a3], [a4, a5, a6], [a7, a8, a9], where a i is the i-th single-cycle sampling quantity.

[0037] For the second division method: when S > W, that is, when the step size is greater than the size of the sliding window, there is an interval between the subsequences. For example, if the length L of the single-cycle sampling quantity sequence is 11, the size W of the sliding window and the step size S are 3 and 4 respectively, then the divided subsequences are [a1, a2, a3], [a5, a6, a7], [a9, a 10 , a 11 ], where a i is the i-th single-cycle sampling quantity.

[0038] For the third division method: when S < W, that is, when the step size is less than the size of the sliding window, there is a sampling quantity overlap between the subsequences, that is, some sampling quantities will exist in two adjacent subsequences at the same time. For example, if the length L of the single-cycle sampling quantity sequence is 7, the size W of the sliding window and the step size S are 3 and 2 respectively, then the divided subsequences are [a1, a2, a3], [a3, a4, a5], [a5, a6, a7], where a i is the i-th single-cycle sampling quantity.

[0039] In an embodiment of the present invention, the dividing the single-cycle sampling quantity sequence into multiple equal-length subsequences includes:

[0040] According to a preset sequence length, performing size normalization on the length of the single-cycle sampling quantity sequence;

[0041] Performing data normalization on the single-cycle sampling quantity sequence after size normalization;

[0042] According to the preset sampling length, the single-cycle sampling quantity sequence after data normalization is slidingly sampled to obtain a plurality of equal-length subsequences.

[0043] Considering different types of single-cycle sampling quantities, due to different sampling points, the length of the generated single-cycle sampling quantity sequence may differ. In order to facilitate unified processing, each single-cycle sampling quantity sequence needs to be normalized in size to adjust each single-cycle sampling quantity sequence to the same sequence length L. For each single-cycle sampling quantity sequence after size normalization: each single-cycle sampling quantity in it is normalized to a preset range, thereby eliminating dimensional differences to improve the accuracy of subsequent model processing. It can be understood that the data normalization method includes but is not limited to standardization, minimum-maximum normalization, etc., which are not limited here. The single-cycle sampling quantity sequence after data normalization is slidingly sampled according to a preset sliding window length and a preset step, thereby dividing a single-cycle sampling quantity sequence into several subsequences. It should be noted that when the type of single-cycle sampling quantity is only one, the above size normalization processing is also required to make the corresponding single-cycle sampling quantity sequence suitable for subsequent model processing.

[0044] In an embodiment of the present application, the size normalization processing of the length of the single-cycle sampling quantity sequence according to the preset sequence length comprises:

[0045] The length of the single-cycle sampling quantity sequence is counted:

[0046] If the length of the single-cycle sampling quantity sequence is greater than the preset sequence length, a plurality of single-cycle sampling quantities are extracted from the single-cycle sampling quantity sequence based on a downsampling method to form a single-cycle sampling quantity sequence of the sequence length;

[0047] If the length of the single-cycle sampling quantity sequence is less than the preset sequence length, the single-cycle sampling quantity sequence is expanded based on an interpolation method to form a single-cycle sampling quantity sequence of the sequence length.

[0048] When the single-cycle sampling quantity sequence is subjected to size normalization processing, different sizes correspond to different processing manners, and thus the length of the single-cycle sampling quantity sequence needs to be counted first. The length of the single-cycle sampling quantity sequence refers to the total amount of single-cycle sampling quantities contained in the sequence. If the length is exactly equal to the preset sequence length, the sequence does not need to be subjected to size normalization processing. If the length is greater than the preset sequence length, the sequence needs to be cropped through downsampling, so that the length of the cropped sequence is equal to the preset sequence length. If the length is less than the preset sequence length, new data needs to be added to the original single-cycle sampling quantity sequence through an interpolation method, so that the sequence length meets the requirement. The interpolation method includes, but is not limited to, bilinear interpolation, Lagrange interpolation, etc.

[0049] S3, calculate the distance between each sub-sequence to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence.

[0050] For each single-cycle sampling quantity, each sub-sequence in the single-cycle sampling quantity is regarded as a vector, a distance matrix is generated by calculating the distance between each sub-sequence, and the distance matrix is regarded as the sampling quantity image of the single-cycle sampling quantity sequence. The distance matrix is input into the battery life prediction model to predict the remaining cycle life of the battery. It can be understood that, in addition to the above-mentioned manner, Gram angle field, Markov transition field, etc. can also be used for time series data imaging processing, which is not limited herein. For example, Figure 3a 、 Figure 3b and Figure 3c as shown, which shows an example of distance matrix imaging. In the above three figures, the deeper the color, the greater the distance between the corresponding two sub-sequences. Among them Figure 3a is the sampling quantity image of voltage, and the horizontal and vertical coordinates are the coordinate positions in the distance matrix. The right color bar is the distance value of the corresponding position in the distance matrix of voltage. The distance matrix is visually displayed in the form of an image, so as to clearly reflect the similarity between the voltage sub-sequences. Figure 3b and Figure 3c are the sampling quantity images of current and capacity, respectively. The specific information is similar to that of the voltage sampling quantity image, which is not described in detail herein.

[0051] Specifically, in an embodiment of the present application, the single-cycle sampling quantity is one kind, and the calculation of the distance between each sub-sequence to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence comprises:

[0052] calculating the distance between each sub-sequence and generating a distance matrix;

[0053] mapping each distance value in the distance matrix to a gray value to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence.

[0054] For each single-cycle sampling quantity sequence, first, divide it into multiple subsequences according to the foregoing manner, regard each subsequence as a W-dimensional vector (W is the total sampling quantity contained in the subsequence), and calculate the distance between each subsequence and the remaining subsequences, as shown in formula (2):

[0055] F i,j =‖p i -p j ‖ (2)

[0056] where p i and p j are the i-th and j-th subsequences, and F i,j is the distance between the i-th subsequence and the j-th subsequence. After the distance between all the subsequences is calculated in the foregoing manner, a distance matrix is formed. The rows and columns of the distance matrix correspond to the respective subsequences, and the elements in the matrix represent the distance between the corresponding row and column subsequences. It should be noted that the distance metric has symmetry, so the distance matrix is a symmetric matrix, and the elements on the diagonal of the matrix represent the distance between the subsequence and itself, so the value is zero. Normalize each distance value in the distance matrix to map it to the range of gray values (i.e., 0 to 255). Each normalized distance value in the normalized distance matrix is a pixel point in the image, so the distance matrix can be regarded as a sampling quantity image after gray scaling. As an example, the distance value can be mapped to the gray value in the following manner: the minimum distance value and the maximum distance value in the distance matrix are respectively mapped to the gray value 0 and the gray value 255, and then the other distance values are linearly mapped to obtain the corresponding gray values.

[0057] It can be understood that in the present application, the distance calculation method between two subsequences can use the Euclidean distance, or other distance calculation methods such as Manhattan distance, Chebyshev distance, Minkowski distance, etc., which can be adaptively selected by those skilled in the art based on the characteristics of the battery sampling quantity, which is not limited herein.

[0058] In an embodiment of the present application, there are multiple single-cycle sampling quantities, and the method for calculating the distance between each subsequence and generating a sampling quantity image corresponding to the single-cycle sampling quantity sequence comprises:

[0059] For each single-cycle sampling quantity:

[0060] Calculate the distance between each subsequence and generate a distance matrix;

[0061] Map each distance value in the distance matrix to a gray value to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence;

[0062] Stacking the sampling quantity images of all single-cycle sampling quantity sequences according to the channel dimension to form a multi-channel sampling quantity image and taking the multi-channel sampling quantity image as the sampling quantity images of all single-cycle sampling quantity sequences.

[0063] If the type of the single-cycle sampling quantity is multiple, for each single-cycle sampling quantity, after generating the corresponding sampling quantity image according to the foregoing method, each sampling quantity image corresponds to one channel based on a preset rule, and multiple single-channel grayscale images are stacked together to form a multi-channel comprehensive sampling quantity image. Through this multi-channel mode, the battery life prediction model can learn the correlation between different types of sampling quantities, thereby improving the accuracy of subsequent life prediction.

[0064] S4, inputting the sampling quantity image into a battery life prediction model to obtain the expected cycle life of the battery.

[0065] The sampling quantity image is input into the battery life prediction model, features in the sampling quantity image that are highly correlated with the cycle life of the battery are identified through feature extraction on the sampling quantity image, and the remaining cycle service life of the battery is obtained. The battery life prediction model is pre-trained, and the model is a CNN architecture, including but not limited to AlexNet, VGG, ResNet, etc.

[0066] In an embodiment of the present application, the battery life prediction model is AlexNet, the battery life prediction model is one, and the input of the sampling quantity image into the battery life prediction model to obtain the expected cycle life of the battery includes:

[0067] The image features of the sampling quantity image are extracted layer by layer by inputting the sampling quantity image into the convolutional network of the battery life prediction model; wherein the convolutional network includes a plurality of cascaded convolutional layers.

[0068] The image features generated by the last convolutional layer of the convolutional network are input into the fully connected network of the battery life prediction model to obtain the expected cycle life of the battery.

[0069] The generated sampling quantity image is input into an AlexNet model, the AlexNet has five cascaded convolution layers and three fully connected layers, and the five cascaded convolution layers form a convolution network. Specifically, the sampling quantity image is input into the convolution network, shallow image features are extracted in the first convolution layer, and subsequent convolution layers integrate and extract deep image features by reducing the size of the convolution kernel and integrating the features extracted by the previous convolution layer. The deep image features generated by the last convolution layer of the convolution network are input into the fully connected layer, the deep image features are weighted combined and nonlinearly mapped, so as to be mapped into a one-dimensional feature vector, and the expected cycle life of the battery is obtained through the output layer. It can be understood that the sampling quantity image described in the embodiment can be a single-channel sampling quantity image or a multi-channel sampling quantity image, which is not limited here.

[0070] In an embodiment of the application, the battery life prediction model is AlexNet, there are multiple battery life prediction models, the network architectures of the multiple battery life prediction models are the same but the parameters are different, and the input of the sampling quantity image into the battery life prediction model to obtain the expected cycle life of the battery includes:

[0071] For each battery life prediction model:

[0072] The sampling quantity image is input into the convolution network of the battery life prediction model, and the image features of the sampling quantity image are extracted layer by layer; wherein the convolution network includes multiple cascaded convolution layers;

[0073] The image features generated by the last convolution layer of the convolution network are input into the fully connected network of the battery life prediction model, and the image features are mapped into a one-dimensional feature vector;

[0074] Each feature vector is flattened and spliced, and the spliced feature vector is input into a fully connected layer to obtain the expected cycle life of the battery.

[0075] In order to improve the accuracy of the prediction, in the embodiment, a plurality of AlexNet battery life prediction models with the same network architecture but different sampling quantities are used to process the sampling quantity image to obtain the expected cycle life of the battery. Specifically, the same sampling quantity image is input into each battery life prediction model. For each battery life prediction model: the features of the sampling quantity image are extracted layer by layer through the convolution network composed of a plurality of cascaded convolution layers, and the image features generated by the last convolution layer are input into the fully connected network of the model to generate a one-dimensional feature vector. After all the battery life prediction models are processed in the above manner, the feature vectors output by all the battery life prediction models are flattened and spliced together to form a comprehensive feature vector, and the comprehensive feature vector formed by splicing is input into a fully connected layer to obtain the expected cycle life of the battery. Since the spliced feature vector contains feature information extracted by multiple models, it can more comprehensively and accurately reflect the relationship between the sampling quantity image and the battery life.

[0076] As a specific example, as Figure 4As shown, the categories of the single-cycle sampling quantities are current, voltage and capacity, so the sampling quantity image input to the model is a three-channel image, and the battery life prediction model is AlexNet. The current, voltage and capacity during a single charge-discharge cycle of the battery are collected, and after constructing the single-cycle voltage sequence, single-cycle current sequence and single-cycle capacity sequence, the preprocessed single-cycle voltage sequence, single-cycle current sequence and single-cycle capacity sequence are obtained through preprocessing processes such as size normalization and data standardization processing. According to the preset sliding window size, each time series is sampled by a preset step size, and each sequence is divided into multiple equal-length subsequences. For each single-cycle sampling quantity: by calculating the distance between each pair of subsequences, an N*N distance matrix corresponding to the single-cycle sampling quantity sequence is generated (N is the total number of subsequences in the single-cycle sampling quantity sequence), and the distance matrix is regarded as the corresponding sampling quantity image. For current, voltage and capacity, current images, voltage images and capacity images are generated correspondingly. The three images are stacked into corresponding channels to form a three-channel image, and the size is adjusted to a uniform size (such as 224). The adjusted three-channel image is input into three trained battery life prediction models (such as AlexNet) respectively, and a feature vector of a preset scale (such as 1000) is generated by the first battery life prediction model AlexNet, the second battery life prediction model AlexNet2 and the third battery life prediction model AlexNet3 respectively. The three feature vectors are spliced together to form a comprehensive feature vector with a scale of 3000, and the remaining cycle life of the battery is obtained accordingly. Experiments show that, in the case of using only single-cycle data, the average absolute percentage error of the cycle life prediction of the public battery dataset used for model evaluation is within 7.1%. Therefore, using this way of the present application can achieve excellent results in the early stage of battery life prediction, not only can realize high-precision prediction of battery life, but also significantly reduces the dependence on large-scale historical data and complex physical models, so this way of the present application has strong popularization and practicality.

[0077] See Figure 5 The battery life prediction device 100 based on the image of the single-cycle sampling quantity includes a data acquisition module 110, a sequence segmentation module 120, an imaging module 130 and a battery life prediction module 140. The data acquisition module 110 is used to sample the battery sampling quantity during a single charge-discharge cycle of the battery, obtain a plurality of single-cycle sampling quantities of the battery, and form a single-cycle sampling quantity sequence. The sequence segmentation module 120 is used to divide the single-cycle sampling quantity sequence into a plurality of equal-length subsequences. The imaging module 130 is used to calculate the distance between each subsequence to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence. The battery life prediction module 140 is used to input the sampling quantity image into a battery life prediction model to obtain the expected cycle life of the battery.

[0078] The specific limitations of the battery life prediction device based on the imaged single-cycle sampling quantity can refer to the limitations of the battery life prediction method based on the imaged single-cycle sampling quantity described above, which will not be repeated here. Each module in the above battery life prediction device based on the imaged single-cycle sampling quantity can be realized by software, hardware, and combinations thereof, in whole or in part. The above modules can be embedded in or independent of the processor in the computer device in hardware format, or stored in the memory in the computer device in software format, so that the processor calls the corresponding operations of the above modules.

[0079] It should be noted that, in order to highlight the innovative part of the present application, modules not closely related to solving the technical problems proposed in the present application are not introduced in this embodiment, but this does not mean that there are no other modules in this embodiment.

[0080] Please refer to Figure 6 , the electronic device 1 can include a memory 12, a processor 13 and a bus, and can also include a computer program stored in the memory 12 and executable on the processor 13, such as a battery life prediction program based on the imaged single-cycle sampling quantity.

[0081] Among them, the memory 12 at least includes one type of readable storage medium, which includes flash memory, mobile hard disk, multimedia card, card type memory (such as SD or DX memory, etc.), magnetic memory, disk, optical disk, etc. The memory 12 can be an internal storage unit of the electronic device 1 in some embodiments, such as the mobile hard disk of the electronic device 1. The memory 12 can also be an external storage device of the electronic device 1 in other embodiments, such as the plug-in mobile hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on the electronic device 1. Further, the memory 12 can include both the internal storage unit and the external storage device of the electronic device 1. The memory 12 can be used not only to store application software and various data installed on the electronic device 1, such as codes for battery life prediction based on the imaged single-cycle sampling quantity, but also to temporarily store data that has been output or will be output.

[0082] The processor 13 may, in some embodiments, be composed of integrated circuits, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits of the same function or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, combinations of various control chips, etc. The processor 13 is the control unit of the electronic device 1, and connects various components of the electronic device 1 through various interfaces and lines, executes programs or modules stored in the memory 12 (for example, a battery life prediction program based on image-based single-cycle sampling amount, etc.), and calls data stored in the memory 12, to perform various functions and process data of the electronic device 1.

[0083] The processor 13 executes the operating system of the electronic device 1 and various installed application programs. The processor 13 executes the application programs to implement the steps in the above-described battery life prediction method based on image-based single-cycle sampling amount.

[0084] By way of example, the computer program can be divided into one or more modules, which are stored in the memory 12 and executed by the processor 13 to complete the present application. The one or more modules can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the electronic device 1. For example, the computer program can be divided into a data acquisition module 110, a sequence segmentation module 120, an imaging module 130, and a battery life prediction module 140.

[0085] The integrated units implemented in the form of software function modules described above can be stored in a computer-readable storage medium, which can be non-volatile or volatile. The software function modules described above are stored in a storage medium, including a number of instructions for causing a computer device (which can be a personal computer, a computer device, or a network device, etc.) or a processor to perform part of the battery life prediction method based on image-based single-cycle sampling amount described in various embodiments of the present application.

[0086] In summary, the battery life prediction method based on the image of single cycle sampling quantity and the related device, the battery sampling quantity is collected during the single charging and discharging cycle of the battery, the single cycle sampling quantity sequence is constructed, and the single cycle sampling quantity sequence is divided into multiple equal-length subsequences. By calculating the distance between each subsequence, a sampling quantity image is generated, and a battery life prediction model is called to predict the remaining cycle life of the battery. The present application does not need to rely on complex physical models or a large amount of historical data, and only needs to use the battery sampling quantity during the single charging and discharging cycle, so as to accurately predict the remaining service life of the battery in the super early stage of the battery use, and effectively improve the prediction accuracy and efficiency. Compared with the traditional method, the present application greatly simplifies the data acquisition and processing process, and ingeniously establishes the connection between the battery sampling quantity and the image, analyzes the image of the battery sampling quantity through the deep learning model, and realizes the accurate prediction of the expected cycle life of the battery. Therefore, the present application effectively overcomes the various shortcomings in the prior art and has high industrial utilization value.

[0087] The above embodiments only exemplarily illustrate the principles and effects of the present application, and are not used to limit the present application. Any person skilled in the art can modify or change the above embodiments without departing from the spirit and scope of the present application. Therefore, all equivalent modifications or changes completed by those skilled in the art without departing from the spirit and technical idea disclosed by the present application should be covered by the claims of the present application.

Claims

1. A battery life prediction method based on an imaged single-cycle sampling amount, characterized by, The method comprises: sampling the battery sampling quantity during a single charging and discharging cycle of the battery, obtaining a plurality of single-cycle sampling quantities of the battery, constituting a single-cycle sampling quantity sequence; wherein the single-cycle sampling quantity sequence is a time series sequence; dividing the single-cycle sampling quantity sequence into a plurality of equal-length subsequences; calculating the distance between each subsequence to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence; inputting the sampling quantity image into a battery life prediction model to obtain the expected cycle life of the battery; wherein the battery sampling quantity is current, voltage and capacity, and the sampling quantity image is obtained from the distance matrix between each subsequence; the battery life prediction model is AlexNet, the battery life prediction model is multiple, the network architecture of multiple battery life prediction models is the same but the parameters are different, and the battery life prediction model is obtained by inputting the sampling quantity image into the battery life prediction model to obtain the expected cycle life of the battery, comprising: for each battery life prediction model input the sampling quantity image into the convolutional network of the battery life prediction model, and extract the image features of the sampling quantity image layer by layer; wherein the convolutional network comprises a plurality of cascaded convolutional layers; input the image features extracted by the last convolutional layer of the convolutional network into the fully connected network of the battery life prediction model, and map the image features to a one-dimensional feature vector; flatten and splice each feature vector, input the spliced feature vector into a fully connected layer, and obtain the expected cycle life of the battery; there are multiple single-cycle sampling quantities, and the distance between each subsequence is calculated to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence, comprising: for each single-cycle sampling quantity calculate the distance between each subsequence and generate a distance matrix; map each distance value in the distance matrix to a gray value to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence; stack the sampling quantity images of all single-cycle sampling quantity sequences according to the channel dimension to form a multi-channel sampling quantity image and use it as the sampling quantity image of all single-cycle sampling quantity sequences. 2.The method of claim 1, wherein, The single-cycle sampling quantity sequence is divided into a plurality of equal-length subsequences, comprising: performing size normalization processing on the length of the single-cycle sampling quantity sequence according to a preset sequence length; performing data normalization processing on the single-cycle sampling quantity sequence after size normalization; performing sliding sampling on the single-cycle sampling quantity sequence after data normalization according to a preset sampling length to obtain a plurality of equal-length subsequences. 3.The method of claim 2, wherein, The size normalization processing on the length of the single-cycle sampling quantity sequence according to a preset sequence length comprises: statistically analyzing the length of the single-cycle sampling quantity sequence: if the length of the single-cycle sampling quantity sequence is greater than the preset sequence length, extract a plurality of single-cycle sampling quantities from the single-cycle sampling quantity sequence based on down-sampling method to form a single-cycle sampling quantity sequence with the sequence length; if the length of the single-cycle sampling quantity sequence is less than the preset sequence length, expand the single-cycle sampling quantity sequence based on interpolation method to form a single-cycle sampling quantity sequence with the sequence length.

4. A battery life prediction device based on an imaged single-cycle sampling amount, characterized by, The device comprises: The data acquisition module is configured to sample battery sampling quantities during a single charging and discharging cycle of the battery, to obtain a plurality of single-cycle sampling quantities of the battery, and to form a single-cycle sampling quantity sequence; the single-cycle sampling quantity sequence is a time sequence; The sequence segmentation module is configured to segment the single-cycle sampling quantity sequence into a plurality of equal-length subsequences; The imaging module is configured to calculate distances between the subsequences, and to generate a sampling quantity image corresponding to the single-cycle sampling quantity sequence; The battery life prediction module is configured to input the sampling quantity image into a battery life prediction model to obtain an expected cycle life of the battery; The battery sampling quantities are currents, voltages, and capacities, and the sampling quantity image is obtained based on a distance matrix between the subsequences; The battery life prediction model is AlexNet, and there are a plurality of battery life prediction models, which have the same network architecture but different parameters; the inputting of the sampling quantity image into the battery life prediction model to obtain the expected cycle life of the battery includes: For each battery life prediction model inputting the sampling quantity image into a convolutional network of the battery life prediction model, and extracting image features of the sampling quantity image layer by layer; the convolutional network includes a plurality of cascaded convolutional layers; inputting image features extracted by a last convolutional layer of the convolutional network into a fully connected network of the battery life prediction model, and mapping the image features into one-dimensional feature vectors; flattening and splicing the feature vectors, inputting the spliced feature vectors into a fully connected layer, and obtaining the expected cycle life of the battery; The single-cycle sampling quantity is of a plurality of types, and the calculation of the distances between the subsequences and the generation of the sampling quantity image corresponding to the single-cycle sampling quantity sequence include: For each type of single-cycle sampling quantity calculating distances between the subsequences and generating a distance matrix; mapping each distance value in the distance matrix into a gray value, and generating a sampling quantity image corresponding to the single-cycle sampling quantity sequence; stacking the sampling quantity images of all single-cycle sampling quantity sequences according to channel dimensions, forming a multi-channel sampling quantity image, and taking the multi-channel sampling quantity image as the sampling quantity images of all single-cycle sampling quantity sequences.

5. An electronic device, comprising: The electronic device includes: one or more processors; a storage device configured to store one or more programs that, when executed by the one or more processors, cause the electronic device to implement the battery life prediction method based on the imaged single-cycle sampling quantity according to any one of claims 1 to 3.

6. A computer readable storage medium characterized by, A computer program is stored thereon, and when the computer program is executed by a processor of a computer, the computer executes the battery life prediction method based on the imaged single-cycle sampling quantity according to any one of claims 1 to 3.

Images