CN116660756B - Battery capacity decay curve generation method based on conditional generative adversarial network - Google Patents
Battery capacity decay curve generation method based on conditional generative adversarial network Download PDFInfo
- Publication number
- CN116660756B CN116660756B CN202310772148.6A CN202310772148A CN116660756B CN 116660756 B CN116660756 B CN 116660756B CN 202310772148 A CN202310772148 A CN 202310772148A CN 116660756 B CN116660756 B CN 116660756B
- Authority
- CN
- China
- Prior art keywords
- curve
- generator
- capacity decay
- decay curve
- data set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/378—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC] specially adapted for the type of battery or accumulator
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/385—Arrangements for measuring battery or accumulator variables
- G01R31/387—Determining ampere-hour charge capacity or SoC
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Secondary Cells (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
本发明公开了一种基于条件生成对抗网络的电池容量衰减曲线生成方法，属于锂电池衰减分析的技术领域，获取锂电池的真实容量衰减曲线；对真实容量衰减曲线归一化处理，且将处理后的数据集划分为训练数据集和测试数据集；构建CGAN网络模型；将训练数据集输入到CGAN网络模型中进行训练，获取最优的生成器；通过最优的生成器批量生成锂电池的预测容量衰减曲线，并挑选出最具有代表性的预测容量衰减曲线；计算测试数据集与最具有代表性的预测容量衰减曲线之间的相似性，以评估生成器的性能，以解决锂电池容量数据因传统数据采集方法而面临的成本和时间限制，从而可有效地生成大量的数据，用于之后电池容量曲线预测、电池性能评估等方面的研究。
The present invention discloses a method for generating a battery capacity decay curve based on a conditional generative adversarial network, which belongs to the technical field of lithium battery decay analysis, and obtains a real capacity decay curve of a lithium battery; normalizes the real capacity decay curve, and divides the processed data set into a training data set and a test data set; constructs a CGAN network model; inputs the training data set into the CGAN network model for training to obtain an optimal generator; generates predicted capacity decay curves of lithium batteries in batches through the optimal generator, and selects the most representative predicted capacity decay curve; calculates the similarity between the test data set and the most representative predicted capacity decay curve to evaluate the performance of the generator, so as to solve the cost and time limitations faced by lithium battery capacity data due to traditional data collection methods, thereby effectively generating a large amount of data for subsequent research in battery capacity curve prediction, battery performance evaluation, etc.
Description
Technical Field
The invention belongs to the technical field of lithium battery attenuation analysis, and particularly relates to a battery capacity attenuation curve generation method based on a condition generation countermeasure network.
Background
As an important rechargeable battery, a lithium ion battery has been widely used in the fields of electric automobiles, mobile devices, energy storage and the like, and the performance of the lithium ion battery is often related to factors such as the number of charge and discharge cycles, the charge and discharge multiplying power and the like. Therefore, the method accurately predicts the capacity curve of the battery and has important significance for performance evaluation, equipment design and control of the battery.
In recent years, a data-driven method has achieved great achievement in battery capacity degradation track prediction, and deep learning technology stands out by virtue of accurate, stable and efficient prediction analysis and other characteristics. However, deep learning correlation algorithms typically require a large number of data sets for model training, and the data sets involved in the network model training need to have a certain diversity, thereby improving the accuracy and generalization of model predictive analysis. However, in practical application, the conventional data acquisition method needs to consume a lot of time and cost, and the acquisition of the battery degradation data is affected by various factors such as the battery working environment, load, acquisition method and the like, and the battery degradation data is not as many as thousands as financial and image data sets, so that the available data sets in the research are extremely precious. In order to acquire more and richer battery degradation data sets, research results and application scenes are further expanded, and the existing battery degradation data sets are expanded through a data generation technology, so that the number of samples and diversity of the data sets are improved, the accuracy and robustness of prediction analysis of a deep learning network model are improved, and research and development of battery capacity degradation track prediction are promoted.
At present, in the generation of battery capacity data, a mainstream method is a data generation method based on a traditional model, wherein the method is generally used for exploring the rule of battery capacity degradation through the establishment of a mathematical or physical model so as to simulate the generation of data, the method is higher in the quantity and quality of the generated data, but the generated data is single and poor in generalization capability, and based on a statistical method, the method is mainly used for analyzing and extracting the rule of battery capacity change along with time by adopting the statistical method and generating the data according to the extracted rule, and the method can accurately simulate the change of the battery capacity, but needs a large amount of experimental data to support and has the same disadvantages as the generation of the traditional model data. Therefore, it is urgent to find a data enhancement method that can generate a large number of data, has a high diversity, and generates data with high fidelity. Deep learning has emerged from birth to date as a vast array of superior models for generating countermeasure networks (GAN), recurrent Neural Networks (RNN), variational self-encoders (VAE), and the like. These models have also achieved very good results in the field of data enhancement.
The present study uses a condition generation countermeasure network (CGAN) in the GAN family to enhance the battery capacity degradation dataset. CGAN is a more suitable option in the present invention than RNN, VAE or other models. Firstly, the RNN can only generate a sequence similar to the historical data, but when generating a battery capacity fading curve, the condition information must be considered, and the RNN cannot effectively use the condition information to generate a capacity fading curve which meets the condition, and secondly, the VAE can generate data similar to the training data, but cannot control the characteristics of the generated data, and in the task of generating the battery capacity fading curve, we need to generate a capacity fading curve with specific requirements according to different condition information, and the VAE cannot effectively realize the problem. In contrast CGAN is a model dedicated to generating conditional constraints, so it is suitable for generating battery capacity decay curves. In addition, CGAN has the greatest advantage over the standard GAN in that the battery capacity fading curve can be generated in a multi-dimensional space by inputting noise and condition information due to the introduction of the condition vector, so that the generated curve has more diversity.
Disclosure of Invention
In view of the above, in order to solve the above problems of the prior art, an object of the present invention is to provide a battery capacity fading curve generation method based on a condition generation countermeasure network to solve the cost and time limitation faced by the conventional data acquisition method for lithium battery capacity data, so that a large amount of data can be effectively generated for later study in battery capacity curve prediction, battery performance evaluation, and the like.
The invention adopts the technical scheme that the battery capacity attenuation curve generating method based on the condition generating countermeasure network comprises the following steps:
s1, acquiring a real capacity attenuation curve of a lithium battery;
S2, carrying out normalization processing on a real capacity attenuation curve of the lithium battery, and dividing a processed data set into a training data set and a test data set;
s3, constructing CGAN a network model, wherein the CGAN network model comprises a generator and a discriminator;
s4, inputting the training data set into a CGAN network model for training to obtain an optimal generator;
s5, generating predicted capacity attenuation curves of the lithium batteries in batches through an optimal generator, and selecting the most representative predicted capacity attenuation curves by using a maximum diversity generation curve selection method;
And S6, calculating a similarity score between the test data set and the most representative predicted capacity fading curve so as to evaluate the performance of the generator, obtaining a group of good CGAN network models if the performance of the generator is good, and returning to S4 to continue training if the performance of the generator is not good.
Further, in S1, the method for obtaining the real capacity fading curve data of the lithium battery is as follows:
s101, dividing a lithium battery into three types according to the service life, namely short service life, medium service life and long service life, wherein the short service life is less than 700 cycles, the medium service life is 700-1500 cycles, and the long service life is more than 1500 cycles;
And S102, clustering by adopting a k-means clustering algorithm to obtain a clustering center curve corresponding to the real capacity attenuation curve of each type of lithium battery.
Further, in the S2, it includes:
S201, defining one-hot codes of various real capacity attenuation curves as 000,010,100 respectively;
S202, reading clustering center curve data of various real capacity attenuation curves, adding artificial control noise n1 to the x axis and adding artificial control noise n2 to the y axis;
s203, storing each cluster center curve and each corresponding one-hot code expansion m times into a set object;
s204, carrying out normalization processing on the acquired collection objects;
s205, packaging the normalized aggregate object into a data loader serving as a training data set.
Further, after removing the cluster center curve from each type of real capacity fade curve, the bar data is randomly chosen as the test dataset.
Further, in S3, the built CGAN network model includes a generator and a arbiter;
the generator comprises BatchNorm linear layers and comprises a BatchNorm d function layer, wherein the input of the generator is random noise and one-hot coding of each type of real capacity attenuation curve, and the output is a predicted capacity attenuation curve generated after five-layer linear layer processing;
The input of the discriminator comprises 4 linear layers, the input of the discriminator is a real capacity fading curve, one-hot codes of each type of real capacity fading curve and the output of the generator, and the output is a predicted capacity fading curve with higher reality and the probability of judging the type of the current predicted capacity fading curve to be true.
Further, in S4, the optimal generator is obtained by the following method:
S401, after a training data set is obtained, randomly generating a batch of random noise which accords with 0-1 positive-fit distribution and has the same size, and inputting the random noise and one-hot codes of various real capacity attenuation curves in the current training data set into a generator together for forward propagation;
S402, inputting the output of the generator, the category to which the current real capacity attenuation curve belongs and the real capacity attenuation curve into a discriminator;
S403, calculating a loss value through the output of the discriminator and the real capacity attenuation curve, carrying out back propagation and feeding back to the generator, and obtaining the optimal generator by the mutual countermeasure between the generator and the discriminator.
Further, in S5, the method of choosing the most representative predicted capacity fade curve is as follows:
s501, generating predicted capacity fading curves in batches by using the trained optimal generator and defining the curves as a set C;
S502, defining an empty set C *, and randomly selecting a curve C r in the generated set C;
S503, finding a curve C with the largest distance curve C r from the set C, then putting the curve C into the set C *, and removing the curve C from the set C;
And repeating S501-S503 until the number of curves in the set C * is equal to the input selection number n, and terminating the cycle to obtain the most representative n curves in the set C.
Further, in S6, the performance of the evaluation generator adopts the following method:
S601, calculating the DTW distance from a curve in the test data set to each predicted capacity attenuation curve in the set C in sequence;
S602, acquiring a predicted capacity fading curve C1 with the smallest distance in a set C, wherein the predicted capacity fading curve has data with the largest similarity;
s603, outputting the smallest predicted capacity fading curve c1 and the corresponding distance index value.
The beneficial effects of the invention are as follows:
1. The battery capacity attenuation curve generation method based on the condition generation countermeasure network solves the problems that the battery capacity attenuation curve data set in the existing battery capacity attenuation curve prediction field is less, the existing data enhancement method in the field is high in difficulty, poor in generalization performance and the like.
2. According to the battery capacity degradation curve generation method based on the condition generation countermeasure network, the existing battery degradation data set is expanded through the CGAN network model, so that the number of samples and the diversity of the data set can be increased, the accuracy and the robustness of prediction analysis of the deep learning network model are improved, and the research and the development of lithium battery capacity degradation track prediction are promoted to a certain extent.
Drawings
FIG. 1 is a flow chart of a method of generating a battery capacity fade profile based on a condition generating countermeasure network provided by the present invention;
FIG. 2 is a diagram of a network model of CGAN of a method for generating a battery capacity decay curve based on a condition generation countermeasure network provided by the present invention;
FIG. 3 is a graph of predicted capacity fade curve results generated in a method for generating a battery capacity fade curve based on a condition generation countermeasure network provided by the present invention;
FIG. 4 is a flow chart of a method for selecting a maximum diversity generation curve in a method for generating a battery capacity fading curve based on a condition generation countermeasure network according to the present invention;
FIG. 5 is a graph of the results of the maximum diversity generation curve selection algorithm in the battery capacity fading curve generation method based on the condition generation countermeasure network provided by the invention;
FIG. 6 is a flow chart of a generator evaluation algorithm designed in the method for generating a battery capacity decay curve based on a condition generation countermeasure network provided by the invention;
fig. 7 is a graph of the result of the evaluation algorithm of the generator in the battery capacity fading curve generation method based on the condition generation countermeasure network provided by the invention.
Detailed Description
Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar modules or modules having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the application. On the contrary, the embodiments of the application include all alternatives, modifications and equivalents as may be included within the spirit and scope of the appended claims.
Example 1
In an embodiment, a method for generating a battery capacity attenuation curve of an countermeasure network based on a condition is specifically provided, as shown in fig. 1, the method includes the following steps:
s1, acquiring a real capacity attenuation curve of a lithium battery, wherein in S1, the method for acquiring the real capacity attenuation curve data of the lithium battery comprises the following steps:
S101, dividing a lithium battery data set into three categories according to the service life length, namely short service life, medium service life and long service life, wherein the short service life is less than 700 cycles, the medium service life is 700-1500 cycles, and the long service life is more than 1500 cycles;
And S102, clustering by adopting a k-means clustering algorithm to obtain a clustering center curve corresponding to the real capacity attenuation curve of each type of lithium battery. The K-means clustering algorithm (K-means clustering algorithm) is an iterative solution clustering analysis algorithm, and the method comprises the steps of pre-dividing data into K groups, randomly selecting K objects as initial clustering centers, calculating the distance between each object and each seed clustering center, and distributing each object to the closest clustering center.
S2, carrying out normalization processing on a real capacity attenuation curve of the lithium battery, and dividing the processed data set into a training data set and a test data set, wherein in the S2, the method for acquiring the training data set comprises the following steps:
S201, defining one-hot codes of various real capacity attenuation curves as 000,010,100 respectively;
S202, reading clustering center curve data of various real capacity attenuation curves, adding artificial control noise n1 to the x axis and adding artificial control noise n2 to the y axis;
s203, storing each cluster center curve and each corresponding one-hot code expansion m times into a set object;
s204, carrying out normalization processing on the acquired collection objects;
S205, packaging the normalized aggregate object into a data loader as a training data set, wherein the training data set is used for model training of a subsequent CGAN network model.
In this S2, after removing the cluster center curves from each type of real capacity fading curves, randomly selecting n pieces of data as a test dataset by a random function, the test dataset being used for model evaluation by a subsequent CGAN network model.
S3, determining CGAN super parameters required by the network model, wherein the super parameters comprise the size B of batch processing, the learning rate LR, the momentum parameter B1 and the momentum parameter B2.
A CGAN network model is constructed, as shown in FIG. 2, the CGAN network model comprises a generator and a discriminator, optimizers of the generator and the discriminator adopt Adam optimizers, and loss functions adopt MSE losses.
The generator comprises BatchNorm layers of functions of BatchNorm d and 5 layers of linear functions, the activation function is LeakyReLU, and finally a layer of Tanh activation function is added to convert the linear output into a value between-1 and 1. The input of the generator is random noise and one-hot coding of each type of real capacity attenuation curve, and the output is a predicted capacity attenuation curve generated after five-layer linear layer processing;
The arbiter comprises 4 linear layers, the activation function is LeakyReLU, meanwhile, in order to prevent the arbiter from generating the over-fitting phenomenon, the middle linear layer is followed by a Dropout layer, and data are discarded according to a fixed proportion, so that data regularization is realized. The input of the discriminator is a real capacity fading curve, one-hot codes of each type of real capacity fading curve and the output of the generator, and the output is a predicted capacity fading curve with higher reality and the probability of judging the category to which the current predicted capacity fading curve belongs as true.
The specific structure of each layer in CGAN network model can be further described by the following table.
Table 1CGAN model each layer concrete structure
Note that point_num represents the number of sampling points of the input battery capacity curve, and n_ classes represents the number of categories of the battery capacity curve.
S4, inputting the training data set into a CGAN network model for training to obtain an optimal generator, and in S4, obtaining the optimal generator by the following method:
S401, after a training data set is obtained, randomly generating a batch of random noise which accords with 0-1 positive-fit distribution and has the same size, and inputting the random noise and one-hot codes of various real capacity attenuation curves in the current training data set into a generator together for forward propagation;
S402, inputting the output of the generator, the category to which the current real capacity attenuation curve belongs and the real capacity attenuation curve into a discriminator;
S403, calculating a loss value loss through the output of the discriminator and the real capacity attenuation curve, feeding back the loss value loss to the generator, continuously adjusting own model parameters, namely parameter weights of each item in a generator network model and the like by the generator according to feedback, and continuously opposing the generator and the discriminator to generate an optimal generator so as to finally acquire the optimal generator.
S5, generating predicted capacity fading curves of lithium batteries in batches through an optimal generator, and selecting the most representative predicted capacity fading curve by using a maximum diversity generation curve selection method, wherein in S5, as shown in FIG. 4, the method for selecting the most representative predicted capacity fading curve is as follows:
s501, generating predicted capacity attenuation curves in batches by using the trained optimal generator and defining the predicted capacity attenuation curves as a set C, wherein the generated curve set is shown in FIG. 3;
S502, defining an empty set C *, and randomly selecting a curve C r in the generated set C;
S503, finding a curve C with the largest distance curve C r from the set C, then putting the curve C into the set C *, and removing the curve C from the set C;
s504, judging whether k is equal to n;
S505, if the two curves are equal, acquiring a selected set C * and ending, if the two curves are not equal, finding a curve C ' with the largest distance curve C r in the set C, then putting the curve C ' into the set C *, removing the curve C ' from the set C, and returning to S504 after k+1;
And repeating the steps until the number of curves in the set C * is equal to the input selection number n, and terminating the cycle to obtain the n curves in the set C, which are the most representative, as shown in FIG. 5.
And S6, calculating a similarity score between the test data set and the most representative predicted capacity fading curve so as to evaluate the performance of the generator, obtaining a group of good CGAN network models if the performance of the generator is good, and returning to S4 to continue training and adjusting model parameters if the performance of the generator is bad. In S6, as shown in fig. 6, the performance of the evaluation generator adopts the following method:
S601, calculating the DTW distance from a curve in the test data set to each predicted capacity attenuation curve in the set C in sequence;
S602, acquiring a predicted capacity fading curve C1 with the smallest distance in the set C, wherein the predicted capacity fading curve has data with the largest similarity, as shown in FIG. 7;
S603, outputting the smallest predicted capacity fading curve c1 and the corresponding distance index value. To measure the magnitude of the difference between the generated predicted capacity fading curve and the true capacity fading curve, the present embodiment uses the Mean Square Error (MSE) as the evaluation index.
Where k is the total number of test sets, y k andA curve capacity value and a true curve capacity value are generated, respectively.
The generator performance trained by this embodiment can be further illustrated using the following table.
TABLE 2 error results of real voltage sequences and generated Curve voltage sequences
As can be seen from the data in Table 2, the real capacity fade curve approximates the predicted capacity fade curve generated by the CGAN network model with less loss, so the generator performs better.
The battery capacity attenuation curve generation method based on the condition generation countermeasure network solves the problems that the battery capacity attenuation curve data set in the existing battery capacity attenuation curve prediction field is small, the existing data enhancement method in the field is large in difficulty, poor in generalization performance and the like. The research expands the existing battery degradation data set through CGAN network model, can improve the sample number and diversity of the data set, improves the accuracy and robustness of the deep learning network model prediction analysis, and promotes the research and development of lithium battery capacity degradation track prediction to a certain extent.
It should be noted that any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and that preferred embodiments of the present application include additional implementations in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present application.
It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of techniques known in the art, discrete logic circuits with logic gates for implementing logic functions on data signals, application specific integrated circuits with appropriate combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, where the program may be stored in a computer readable storage medium, and where the program, when executed, includes one or a combination of the steps of the method embodiments.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing module, or each unit may exist alone physically, or two or more units may be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules may also be stored in a computer readable storage medium if implemented in the form of software functional modules and sold or used as a stand-alone product.
The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like.
While embodiments of the present application have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the application, and that variations, modifications, alternatives and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the application.
Claims (8)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310772148.6A CN116660756B (en) | 2023-06-28 | 2023-06-28 | Battery capacity decay curve generation method based on conditional generative adversarial network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310772148.6A CN116660756B (en) | 2023-06-28 | 2023-06-28 | Battery capacity decay curve generation method based on conditional generative adversarial network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116660756A CN116660756A (en) | 2023-08-29 |
| CN116660756B true CN116660756B (en) | 2025-01-24 |
Family
ID=87719041
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310772148.6A Active CN116660756B (en) | 2023-06-28 | 2023-06-28 | Battery capacity decay curve generation method based on conditional generative adversarial network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116660756B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118330471B (en) * | 2024-06-14 | 2024-08-23 | 南通乐创新能源有限公司 | Battery EIS data enhancement method based on data enhancement model and related equipment |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113176518B (en) * | 2021-03-24 | 2023-08-15 | 西安理工大学 | Lithium battery health state estimation method |
| CN115526368B (en) * | 2021-06-25 | 2024-12-10 | 比亚迪股份有限公司 | Power battery capacity prediction method, device and equipment |
| US20230182616A1 (en) * | 2021-12-06 | 2023-06-15 | Sustainable Energy Technologies, Inc. | System and method for determining range and capacity of supercapacitor battery storage for electric vehicle |
| CN114282654B (en) * | 2021-12-24 | 2024-09-27 | 杭州电子科技大学 | EV charging load calculation method based on condition generation countermeasure network |
| CN114662653B (en) * | 2022-02-24 | 2024-06-04 | 苏州恒美电子科技股份有限公司 | Double-LSTM battery capacity estimation method based on generation type countermeasure network |
| CN115456020B (en) * | 2022-09-08 | 2024-12-17 | 哈尔滨工业大学 | Lithium battery life detection method, device and medium based on SDP aging spectrum and CGAN |
| CN115356639B (en) * | 2022-09-29 | 2023-04-18 | 深圳先进技术研究院 | A bidirectional lithium-ion battery intelligent health monitoring method and system |
| CN115980586A (en) * | 2022-11-28 | 2023-04-18 | 浙江凌骁能源科技有限公司 | Power battery health state prediction method and device and computer equipment |
-
2023
- 2023-06-28 CN CN202310772148.6A patent/CN116660756B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| A synthetic data generation method and evolutionary transformer model for degradation trajectory prediction in lithium-ion batteries;Haiyan Jin 等;《Applied Energy》;20241022(第377期);124629 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116660756A (en) | 2023-08-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jiang et al. | Efficient network architecture search via multiobjective particle swarm optimization based on decomposition | |
| CN111352965B (en) | Training method of sequence mining model, and processing method and equipment of sequence data | |
| KR20210040248A (en) | Generative structure-property inverse computational co-design of materials | |
| CN111507521A (en) | Power load forecasting method and forecasting device in Taiwan area | |
| CN113632106A (en) | Hybrid precision training of artificial neural networks | |
| CN110473592A (en) | The multi-angle of view mankind for having supervision based on figure convolutional network cooperate with lethal gene prediction technique | |
| CN112834927A (en) | Method, system, device and medium for predicting remaining life of lithium battery | |
| JP7542793B2 (en) | Method and system for lightweight artificial intelligence inference models | |
| US20220284261A1 (en) | Training-support-based machine learning classification and regression augmentation | |
| CN116660756B (en) | Battery capacity decay curve generation method based on conditional generative adversarial network | |
| CN115203631A (en) | A method and system for multimodal data analysis based on improved genetic algorithm | |
| WO2023044927A1 (en) | Rna-protein interaction prediction method and apparatus, and medium and electronic device | |
| CN116468095A (en) | Neural network architecture searching method and device, equipment, chip and storage medium | |
| Jiang et al. | PDAS: Improving network pruning based on progressive differentiable architecture search for DNNs | |
| CN112365033A (en) | Wind power interval prediction method, system and storage medium | |
| EP3965015A1 (en) | Partial-activation of neural network based on heat-map of neural network activity | |
| CN111831955B (en) | Lithium ion battery remaining life prediction method and system | |
| CN118641969A (en) | A method for predicting the remaining life of lithium batteries | |
| CN115222012A (en) | Method and apparatus for compressing neural networks | |
| CN118261223A (en) | Method for training generating capacity prediction model of micro-grid photovoltaic power generation system | |
| CN110457814A (en) | Multi-modal cross entropy optimization algorithm based on mixed Gauss model | |
| Shuai et al. | A Self-adaptive neuroevolution approach to constructing Deep Neural Network architectures across different types | |
| WO2023044931A1 (en) | Rna-protein interaction prediction method and apparatus, and medium and electronic device | |
| CN112990255A (en) | Method and device for predicting equipment failure, electronic equipment and storage medium | |
| Zhang et al. | Adaptive-Precision Framework for SGD Using Deep Q-Learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |