Method, system and equipment for estimating health state of lithium ion battery of same kind and storage medium

Abstract

The invention relates to the technical field of battery management, in particular to a similar lithium ion battery health state estimation method, system and device and a storage medium. Comprising the following steps: acquiring voltage, current, capacity and corresponding sampling time data of the lithium ion battery in a charge-discharge cycle constant-current charging stage; obtaining a capacity increment curve and a constant-current charging curve based on the voltage, the current, the capacity and the corresponding sampling time data; performing dimension reduction processing on the capacity increment curve and the constant current charging curve based on a segmented aggregation approximation algorithm to obtain a low-dimensional capacity increment curve and a low-dimensional constant current charging curve; fusing the low-dimensional capacity increment curve and the low-dimensional constant-current charging curve to obtain fusion features; according to the method, the low-dimensional capacity increment curve and the low-dimensional constant-current charging curve are fused to serve as the characteristics for evaluating the health of the battery, the data dimension is reduced, information reflecting the health state of the battery is reserved, and therefore the efficiency and accuracy of characteristic extraction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of battery management, and in particular to a method, system, device and storage medium for estimating the health status of a similar lithium-ion battery. Background Art

[0002] Lithium-ion batteries play an important role in the fields of electric vehicles and energy storage due to their advantages such as long cycle life, high energy density and strong reliability. However, the electrochemical properties of lithium-ion batteries will change and decay continuously during the charge and discharge cycle. The continuous deterioration of battery performance will seriously affect its service life, resulting in a reduction in available capacity and even safety hazards. Therefore, in order to improve the performance of lithium batteries, it is necessary to accurately estimate the state of health (SOH). At present, there is no formal definition of the SOH of the battery, and any variable that can reflect the battery usage status can be used to define the SOH. In general, the SOH of a battery is usually defined as a decrease in battery capacity or an increase in internal resistance. This application regards the ratio of the current maximum available capacity to the nominal capacity as the SOH of the battery. The research on SOH estimation can be mainly divided into two categories: model-based methods and data-driven methods.

[0003] The model-based estimation method is to use electrochemical models or equivalent circuit models, etc., and combine them with other methods to describe the internal reaction mechanism of the battery during charging and discharging to achieve battery SOH estimation. The equivalent circuit model has limited accuracy and poor adaptability, and cannot describe the internal reaction mechanism of the battery; the electrochemical model has the problem of difficulty in parameter identification due to too many parameters. The model-based estimation method relies on the accuracy of the battery model, but the aging mechanism of lithium-ion batteries is very complex, and it is always challenging to establish an accurate estimation model.

[0004] In contrast, data-driven methods do not consider the internal reaction mechanism of the battery and accurately capture the complex nonlinear behavior of the battery through a large amount of battery data. Although data-driven methods have been widely used in lithium-ion battery SOH estimation, they still have two problems: (1) Since the aging of lithium-ion batteries is a very complex process, it is difficult to fully reflect the health status of lithium-ion batteries by only extracting some features during the charging and discharging process. It is necessary to extract and analyze various features from different perspectives. In the current research, battery health-related features are composed of local changes in the curve, and the extracted features may not fully reflect the aging process of lithium-ion batteries. (2) The current lithium-ion battery SOH estimation research has too much training data, too few SOH values ​​that can be accurately estimated, and a lack of SOH estimation methods under low training data. Summary of the invention

[0005] In response to the problems mentioned in the prior art, the present invention proposes a similar lithium-ion battery health status estimation method, system, device and storage medium to solve the technical problems in the existing related technology when estimating the battery health status, such as insufficient feature extraction, excessive use of training data, and difficulty in achieving accurate battery health status estimation.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: A similar lithium-ion battery health status estimation method of the present invention comprises the following steps: S1, collecting the voltage, current, capacity and corresponding sampling time data of the lithium-ion battery during the constant current charging stage of the charge and discharge cycle; S2, obtaining a capacity increment curve and a constant current charging curve based on the voltage, current, capacity and corresponding sampling time data; S3, performing dimensionality reduction processing on the capacity increment curve and the constant current charging curve based on a segmented aggregation approximation algorithm to obtain a low-dimensional capacity increment curve and a low-dimensional constant current charging curve; S4, fusing the low-dimensional capacity increment curve and the low-dimensional constant current charging curve to obtain a fusion feature; S5. Establish a health status estimation model based on CNN-Residual-BiGRU-Attention according to the fusion features; S6. Input the fusion features of the unknown lithium-ion battery into a health state estimation model to obtain a health state estimation result of the unknown lithium-ion battery.

[0007] As a further improvement of the present invention, the specific process of step S1 is: The constant current charging voltage data of the kth cycle is the sequence ,in and They are the initial voltage and the final voltage of the constant current charging stage respectively; The constant current charging current data of the kth cycle is the sequence ,in and They are the initial current and the termination current in the constant current charging stage respectively; The constant current charging sampling time data of the kth cycle is the sequence ,in and They are the initial sampling time and the terminal sampling time of the constant current charging stage respectively; The charging capacity data in the constant current charging stage is a sequence , ,in is the constant current charging capacity of the battery in the kth cycle, and are the current and sampling time of the kth cycle respectively.

[0008] As a further improvement of the present invention, the specific process of step S2 is: The capacity increment curve is expressed as follows:

[0009] Where: Indicates the change in battery capacity from time m to time m+1; Represents the voltage change from time m to time m+1; represents the voltage of the battery at time m+1; represents the voltage at time m; represents the battery capacity at time m; Indicates the capacity of the battery at time m+1; Indicates the battery current; represents the sampling time at time m+1, represents the sampling time at time m; The constant current charging curve is shown as follows:

[0010] Where: Represents the voltage at time m.

[0011] As a further improvement of the present invention, the specific process of step S3 is: Given an IC sequence of length m and constant current charging sequence , becomes a data sequence of length w and , where w <m; The resulting compression ratio is The dimensionality reduction process of time series data, and and (i=1,2,…w) satisfies ; Where: represents the ith IC data after dimension reduction, represents the jth original IC data, represents the i-th constant current charging data after dimension reduction, represents the jth original constant current charging data, is the compression ratio.

[0012] As a further improvement of the present invention, the specific process of step S4 is:

[0013] Where: represents the ith IC data after dimension reduction, represents the i-th constant current charging data after dimension reduction, represents vector concatenation, Represents the i-th fused feature.

[0014] As a further improvement of the present invention, the specific process of step S5 is: The CNN network is:

[0015]

[0016] Where: Represents the weight matrix of 1DCNN; Represents the feature vector after 1DCNN; represents the activation function; The feature vector representing the input; The Residual network is:

[0017] Where: Indicates The input vector of the layer network; Represents the feature vector after the activation function; Indicates The input vector of the layer network; The BiGRU network is:

[0018]

[0019]

[0020]

[0021]

[0022]

[0023] Where: represents the update gate; Represents the sigmoid function; represents the weight matrix of the update gate; represents the bias of the update gate; represents the input vector; Indicates the hidden state at the previous moment; Represents the input at the current moment; Reset gate. represents the weight matrix of the reset gate; represents the bias of the reset gate; Represents the new candidate state of the forward GRU; represents the updated hidden state of the forward GRU; Represents the forward GRU weight matrix; Represents the forward GRU bias; Represents the reset gate of the forward GRU; Represents the update gate of the forward GRU; represents the hyperbolic tangent function; Represents the new candidate state of the reverse GRU; represents the hidden state updated by the reverse GRU; Represents the inverse GRU weight matrix; Represents the reverse GRU bias; Represents the reset gate of the reverse GRU; Represents the update gate of the reverse GRU; The Attention is:

[0024]

[0025]

[0026]

[0027] Where: represents the randomly initialized attention weights; Represents the input Attention mechanism weight; represents the weight of the fully connected layer; Indicates bias; Represents the output after weighting by the attention mechanism; represents the hidden layer output; Indicates the weight of the current input; Represents the final output; represents the activation function; The fused features are input into the CNN-Residual-BiGRU-Attention model for training and learning to obtain a trained health status estimation model.

[0028] As a further improvement of the present invention, during the model training process, mean absolute error, mean absolute percentage error and root mean square error are used as evaluation indicators for training.

[0029] A first-of-its-kind lithium-ion battery health status estimation system, comprising: The acquisition module is used to collect the voltage, current, capacity and corresponding sampling time data of the lithium-ion battery during the constant current charging stage of the charge and discharge cycle; A curve generation module, used to obtain a capacity increment curve and a constant current charging curve based on voltage, current, capacity and corresponding sampling time data; A curve processing module is used to perform dimensionality reduction processing on the capacity increment curve and the constant current charging curve based on a segmented aggregation approximation algorithm to obtain a low-dimensional capacity increment curve and a low-dimensional constant current charging curve; A feature fusion module is used to fuse the low-dimensional capacity increment curve and the low-dimensional constant current charging curve to obtain a fusion feature; A building block for building a CNN-Residual-BiGRU-Attention-based health state estimation model based on fused features.

[0030] A device for estimating the health status of a similar lithium-ion battery comprises a processor and a memory, wherein the processor implements the above-mentioned method for estimating the health status of a similar lithium-ion battery when executing a computer program stored in the memory.

[0031] A computer-readable storage medium is used to store a computer program, wherein when the computer program is executed by a processor, the above-mentioned lithium-ion battery health state estimation method is implemented.

[0032] Compared with the prior art, the present invention has achieved the following technical effects: The present invention combines the low-dimensional capacity increment curve and the low-dimensional constant current charging curve as the characteristics for evaluating battery health. Compared with the prior art which only extracts local features from high-dimensional raw data, it reduces the data dimension while retaining the information reflecting the battery health status, thereby improving the efficiency and accuracy of feature extraction.

[0033] The present invention adopts the CNN-Residual-BiGRU-Attention hybrid model architecture, which can deeply mine and effectively learn the deep-level information in the low-dimensional fusion features, thereby improving the accuracy and robustness of battery health state estimation. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 It is a schematic diagram of the process of the present invention; Figure 2 It is a schematic diagram of the original IC curve and the original constant current charging curve and the low-dimensional IC curve and the low-dimensional constant current charging curve processed by the segmented aggregation approximation algorithm of the present invention; Figure 3 It is a structural diagram of the CNN-Residual-BiGRU-Attention model of the present invention; Figure 4This is a graph showing the prediction results of the present invention based on NASA lithium-ion battery aging data; Figure 5 This is a diagram showing the prediction results of the lithium-ion battery aging data of the present invention based on the Oxford University. DETAILED DESCRIPTION

[0034] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0035] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments: like Figure 1 As shown, a similar lithium-ion battery health status estimation method of the present invention comprises the following steps: Step S1, collecting the voltage, current, capacity and corresponding sampling time data of the lithium-ion battery in the constant current charging stage of the charge and discharge cycle. The collection process in the embodiment is: The constant current charging voltage data of the kth cycle is the sequence ,in and They are the initial voltage and the final voltage of the constant current charging stage respectively; The constant current charging current data of the kth cycle is the sequence ,in and They are the initial current and the termination current in the constant current charging stage respectively; The constant current charging sampling time data of the kth cycle is the sequence ,in and They are the initial sampling time and the terminal sampling time of the constant current charging stage respectively; The charging capacity data in the constant current charging stage is a sequence , ,in is the constant current charging capacity of the battery in the kth cycle, and are the current and sampling time of the kth cycle respectively.

[0036] In some embodiments, this step also collects the health status data of the lithium battery, and the health status data of the lithium battery is used to evaluate the reliability of the final output result of this application. Specifically, the health status data sequence of the lithium battery , in,

[0037] In the formula, is the battery health status of the i-th (i=1,2,…,n) charge-discharge cycle, n is the number of battery charge-discharge cycles, is the maximum discharge capacity of the battery in the ith cycle, is the rated capacity of the battery.

[0038] Step S2, obtaining a capacity increment curve and a constant current charging curve based on voltage, current, capacity and corresponding sampling time data; like Figure 2 As shown, the incremental capacity (IC) curve in this step of the embodiment is expressed as follows:

[0039] Where: Indicates the change in battery capacity from time m to time m+1; Represents the voltage change from time m to time m+1; represents the voltage of the battery at time m+1; represents the voltage at time m; represents the battery capacity at time m; Indicates the capacity of the battery at time m+1; Indicates the battery current; represents the sampling time at time m+1, represents the sampling time at time m; The constant current charging curve in this step of the embodiment is shown as follows:

[0040] Where: Represents the voltage at time m.

[0041] Step S3, performing dimensionality reduction processing on the capacity increment curve and the constant current charging curve based on a segmented aggregation approximation algorithm to obtain a low-dimensional capacity increment curve and a low-dimensional constant current charging curve; In step S3 of the embodiment, the capacity increment curve and the constant current charging curve are reduced in dimension based on the segmented aggregation approximation algorithm. It should be noted that in the embodiment, the segmented aggregation approximation algorithm simplifies the data volume while retaining the main features and details of the original data.

[0042] Given an IC sequence of length m and constant current charging sequence , becomes a data sequence of length w and , where w <m; The resulting compression ratio is The dimensionality reduction process of time series data, and and (i=1,2,…w) satisfies ; Where: represents the ith IC data after dimension reduction, represents the jth original IC data, represents the i-th constant current charging data after dimension reduction, represents the jth original constant current charging data, is the compression ratio.

[0043] Step S4: Fusing the low-dimensional capacity increment curve and the low-dimensional constant current charging curve to obtain a fusion feature. The fusion process is:

[0044] Where: represents the ith IC data after dimension reduction, represents the i-th constant current charging data after dimension reduction, represents vector concatenation, Represents the i-th fused feature.

[0045] Step S5: Establish a health status estimation model based on CNN-Residual-BiGRU-Attention according to the fusion features; Figure 3 As shown, in the embodiment, the present invention combines the deep learning model of convolutional neural network (CNN), residual network (Residual), bidirectional gated recurrent unit (BiGRU) and attention mechanism (Attention), inputs the fusion features into the CNN-Residual-BiGRU-Attention model for learning and training, and obtains the health status estimation model after the training is completed.

[0046] During the training process, mean absolute error, mean absolute percentage error and root mean square error are used as evaluation indicators for training.

[0047] The low-dimensional capacity increment curve and the low-dimensional constant current charging curve of a battery of the same type are fused and input into the CNN-Residual-BiGRU-Attention model for pre-training. After the training is completed, all BiGRU layers of the health state estimation model are frozen; 40% of the low-dimensional fusion features composed of low-dimensional capacity increment curve and low-dimensional constant current charging curve of other similar batteries are taken as the training set to fine-tune the health status estimation model. 60% of the low-dimensional fusion features are taken as the test set and input into the health status estimation model to output the estimated results of the health status.

[0048] The CNN network is:

[0049]

[0050] Where: Represents the weight matrix of 1DCNN; Represents the feature vector after 1DCNN; represents the activation function; The feature vector representing the input; The residual network is:

[0051] Where: Indicates The input vector of the layer network; Represents the feature vector after the activation function; Indicates The input vector of the layer network; The BiGRU network is:

[0052]

[0053]

[0054]

[0055]

[0056]

[0057] Where: represents the update gate; Represents the sigmoid function; represents the weight matrix of the update gate; represents the bias of the update gate; represents the input vector; Indicates the hidden state at the previous moment; Represents the input at the current moment; Reset gate. represents the weight matrix of the reset gate; represents the bias of the reset gate; Represents the new candidate state of the forward GRU; represents the updated hidden state of the forward GRU; Represents the forward GRU weight matrix; Represents the forward GRU bias; Represents the reset gate of the forward GRU; Represents the update gate of the forward GRU; represents the hyperbolic tangent function; Represents the new candidate state of the reverse GRU; represents the hidden state updated by the reverse GRU; Represents the inverse GRU weight matrix; Represents the reverse GRU bias; Represents the reset gate of the reverse GRU; Represents the update gate of the reverse GRU; Attention is:

[0058]

[0059]

[0060]

[0061] Where: represents the randomly initialized attention weights; Represents the input Attention mechanism weight; represents the weight of the fully connected layer; Indicates bias; Represents the output after weighting by the attention mechanism; represents the hidden layer output; Indicates the weight of the current input; Represents the final output; Represents the activation function.

[0062] In order to demonstrate the feasibility and accuracy of the similar lithium-ion battery health status estimation method based on segmented approximate aggregation and CNN-Residual-BiGRU-Attention model in this application, this application uses the following examples to illustrate, and compares the MAE, MAPE and RMSE of multiple algorithms to illustrate the accuracy of this application.

[0063] This application uses the lithium-ion battery cycle life data set numbered B0005, B0006, and B0007 of the National Aeronautics and Space Administration (NASA) of a certain country and the lithium-ion battery cycle life data set numbered Cell1, Cell3, and Cell8 of the University of Oxford to verify the proposed hybrid model. The operating temperature of the NASA battery is 24°C, the charging cut-off voltage is 4.2V, and the rated capacity is 2Ah. The battery aging experiment is carried out using the constant current and constant voltage charging mode and the constant current discharge mode. The rated capacity of the Oxford University lithium-ion battery is 740mAh, the ambient temperature is 40°C, and the constant current and constant voltage charging mode and the ARTEMIS urban driving cycle discharge mode are used. Characterization tests are carried out after every 100 cycles. The characterization measurement is 1C constant current charging and discharging, and the data sampling frequency is 1hz.

[0064] Use the CNN-BiGRU model and the CNN-BiGRU-Attention model for comparison The hybrid model was pre-trained using the B0006 battery data from NASA. After pre-training, the model was fine-tuned using 40% of the data from B0005 and B0007 batteries. The SOH estimation results are shown in the figure. Figure 4 As shown, Figure 4 (a) is the health status estimation result of B0005 battery. Figure 4 (b) is the health status estimation result of the B0007 battery. It can be seen that compared with the CNN-BiGRU and CNN-BiGRU-Attention models, the CNN-Residual-BiGRU-Attention model proposed in the present invention has better estimation performance, smaller fluctuations in the estimation curve, and is closer to the actual SOH curve. The model also has a good estimation effect in the capacity regeneration area. In contrast, the estimation curves of the CNN-BiGRU and CNN-BiGRU-Attention models show obvious fluctuations.

[0065] The MAE, MAPE and RMSE of the estimation results of the three models are shown in Table 1. It can be seen that for B0005 batteries, the MAE, MAPE and RMSE of the CNN-BiGRU model are 0.0073, 0.0106, and 0.0089, respectively; the MAE, MAPE and RMSE of the CNN-BiGRU-Attention model are 0.0048, 0.0070, and 0.0060, respectively; the MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are 0.0032, 0.0045, and 0.0040, respectively; the MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are smaller than those of the CNN-BiGRU model and the CNN-BiGRU-Attention model.

[0066] Table 2 shows the average error indicators of B0005 and B0007 batteries. The MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are 0.0030, 0.0041 and 0.0041, respectively, all of which are less than 0.005. Compared with the CNN-BiGRU model, the MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are reduced by 60.53%, 60.95% and 55.43%, respectively. Compared with the CNN-BiGRU-Attention model, the average MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are reduced by 44.44%, 46% and 38.81%, respectively. In summary, the CNN-Residual-BiGRU-Attention model proposed in the present invention shows good performance in NASA's lithium-ion battery aging dataset, and its MAE, MAPE and RMSE are better than those of the CNN-BiGRU model and the CNN-BiGRU-Attention model, which shows that it has good stability and accuracy.

[0067] Table 1 Error indicators of model comparison experiments

[0068] Table 2 Average error indicators of B0005 and B0007 batteries

[0069] The hybrid model was pre-trained using the Cell1 battery data from Oxford University Battery. After pre-training, the model was fine-tuned using 40% of the data from Cell3 and Cell8 batteries. The SOH estimation results are shown in Figure 2. Figure 5 As shown, Figure 5 (a) is the health status estimation result of Cell3 battery. Figure 5 (b) is the health status estimation result of Cell8 battery. It can be seen that compared with the CNN-BiGRU and CNN-BiGRU-Attention models, the CNN-Residual-BiGRU-Attention model proposed in the present invention has better estimation performance, smaller fluctuation of the estimation curve, and is closer to the actual SOH curve.

[0070] The MAE, MAPE and RMSE of the estimation results of the three models are shown in Table 1. It can be seen that for the Cell3 battery, the MAE, MAPE and RMSE of the CNN-BiGRU model are 0.0041, 0.0052, and 0.0046 respectively; the MAE, MAPE and RMSE of the CNN-BiGRU-Attention model are 0.0027, 0.0033, and 0.0031 respectively; the MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are 0.0012, 0.0015, and 0.0017 respectively; the MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are smaller than those of the CNN-BiGRU model and the CNN-BiGRU-Attention model.

[0071] Table 2 shows the average error indicators of Cell3 and Cell8. The MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are 0.0013, 0.0016 and 0.0017, respectively, all of which are less than 0.002. Compared with the CNN-BiGRU model, the MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are reduced by 63.24%, 63.64% and 59.76%, respectively. Compared with the CNN-BiGRU-Attention model, the average MAE, MAPE and RMSE of the CNN-Residual-BiGRU-Attention model are reduced by 47.92%, 45.76% and 43.10%, respectively. In summary, the CNN-Residual-BiGRU-Attention model proposed in the present invention shows good performance in NASA's lithium-ion battery aging dataset, and its MAE, MAPE and RMSE are better than those of the CNN-BiGRU model and the CNN-BiGRU-Attention model, which shows that it has good stability and accuracy.

[0072] Table 1 Error indicators of model comparison experiments

[0073] Table 2 Average error indicators of Cell3 and Cell8 batteries

[0074] Based on the same inventive concept, an embodiment of the present invention also provides a similar lithium-ion battery health status estimation system. Since the principle of solving the problem by the similar lithium-ion battery health status estimation system is similar to the aforementioned similar lithium-ion battery health status estimation method, the implementation of the similar lithium-ion battery health status estimation system can refer to the implementation of the similar lithium-ion battery health status estimation method, and the repeated parts will not be repeated.

[0075] In specific implementation, the similar lithium-ion battery health status estimation system provided by the embodiment of the present invention specifically includes: The acquisition module is used to collect the voltage, current, capacity and corresponding sampling time data of the lithium-ion battery during the constant current charging stage of the charge and discharge cycle; A curve generation module, used to obtain a capacity increment curve and a constant current charging curve based on voltage, current, capacity and corresponding sampling time data; A curve processing module is used to perform dimensionality reduction processing on the capacity increment curve and the constant current charging curve based on a segmented aggregation approximation algorithm to obtain a low-dimensional capacity increment curve and a low-dimensional constant current charging curve; A feature fusion module is used to fuse the low-dimensional capacity increment curve and the low-dimensional constant current charging curve to obtain a fusion feature; A building block for building a CNN-Residual-BiGRU-Attention-based health state estimation model based on fused features.

[0076] Correspondingly, an embodiment of the present invention further provides a similar lithium-ion battery health status estimation device, including a processor and a memory, wherein the processor implements the similar lithium-ion battery health status estimation method provided in the embodiment of the present invention when executing the computer program stored in the memory.

[0077] For more specific processes of the above method, please refer to the corresponding contents disclosed in the aforementioned embodiments, which will not be repeated here.

[0078] Accordingly, an embodiment of the present invention further provides a computer-readable storage medium for storing a computer program, wherein when the computer program is executed by a processor, the above-mentioned similar lithium-ion battery health status estimation method provided in the embodiment of the present invention is implemented.

[0079] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the systems, devices, and storage media disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method part.

[0080] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0081] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0082] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0083] The above is a detailed introduction to the similar lithium-ion battery health status estimation method, system, device and storage medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technicians in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A similar lithium-ion battery health status estimation method, characterized in that: The following steps are involved: S1, collecting the voltage, current, capacity and corresponding sampling time data of the lithium-ion battery during the constant current charging stage of the charge and discharge cycle; S2, obtaining a capacity increment curve and a constant current charging curve based on the voltage, current, capacity and corresponding sampling time data; S3, performing dimensionality reduction processing on the capacity increment curve and the constant current charging curve based on a segmented aggregation approximation algorithm to obtain a low-dimensional capacity increment curve and a low-dimensional constant current charging curve; S4, fusing the low-dimensional capacity increment curve and the low-dimensional constant current charging curve to obtain a fusion feature; S5. Establish a health status estimation model based on CNN-Residual-BiGRU-Attention according to the fusion features; S6. Input the fusion features of the unknown lithium-ion battery into a health state estimation model to obtain a health state estimation result of the unknown lithium-ion battery.

2. A method for estimating the health status of a similar lithium-ion battery according to claim 1, characterized in that: The specific process of step S1 is: The constant current charging voltage data of the kth cycle is the sequence ,in and They are the initial voltage and the final voltage of the constant current charging stage respectively; The constant current charging current data of the kth cycle is the sequence ,in and They are the initial current and the termination current in the constant current charging stage respectively; The constant current charging sampling time data of the kth cycle is the sequence ,in and They are the initial sampling time and the terminal sampling time of the constant current charging stage respectively; The charging capacity data in the constant current charging stage is a sequence , ,in is the constant current charging capacity of the battery in the kth cycle, and are the current and sampling time of the kth cycle respectively.

3. A method for estimating the health status of similar lithium-ion batteries according to claim 1, characterized in that: The specific process of step S2 is: The capacity increment curve is expressed as follows: Where: Indicates the change in battery capacity from time m to time m+1; Represents the voltage change from time m to time m+1; represents the voltage of the battery at time m+1; represents the voltage at time m; represents the battery capacity at time m; Indicates the capacity of the battery at time m+1; Indicates the battery current; represents the sampling time at time m+1, represents the sampling time at time m; The constant current charging curve is shown as follows: Where: Represents the voltage at time m.

4. A method for estimating the health status of similar lithium-ion batteries according to claim 1, characterized in that: The specific process of step S3 is: Given an IC sequence of length m and constant current charging sequence , becomes a data sequence of length w and , where w <m; The resulting compression ratio is The dimensionality reduction process of time series data, and and (i=1,2,…w) satisfies ; Where: represents the ith IC data after dimension reduction, represents the jth original IC data, represents the i-th constant current charging data after dimension reduction, represents the jth original constant current charging data, is the compression ratio.

5. The method for estimating the health status of a similar lithium-ion battery according to claim 1, characterized in that: The specific process of step S4 is: Where: represents the ith IC data after dimension reduction, represents the i-th constant current charging data after dimension reduction, represents vector concatenation, Represents the i-th fused feature.

6. A method for estimating the health status of similar lithium-ion batteries according to claim 1, characterized in that: The specific process of step S5 is: The CNN network is: Where: Represents the weight matrix of 1DCNN; Represents the feature vector after 1DCNN; represents the activation function; The feature vector representing the input; The Residual network is: Where: Indicates The input vector of the layer network; Represents the feature vector after the activation function; Indicates The input vector of the layer network; The BiGRU network is: Where: represents the update gate; Represents the sigmoid function; represents the weight matrix of the update gate; represents the bias of the update gate; represents the input vector; Indicates the hidden state at the previous moment; Represents the input at the current moment; Reset gate. represents the weight matrix of the reset gate; represents the bias of the reset gate; Represents the new candidate state of the forward GRU; represents the updated hidden state of the forward GRU; Represents the forward GRU weight matrix; Represents the forward GRU bias; Represents the reset gate of the forward GRU; Represents the update gate of the forward GRU; represents the hyperbolic tangent function; Represents the new candidate state of the reverse GRU; represents the hidden state updated by the reverse GRU; Represents the inverse GRU weight matrix; Represents the reverse GRU bias; Represents the reset gate of the reverse GRU; Represents the update gate of the reverse GRU; The Attention is: Where: represents the randomly initialized attention weights; Represents the input Attention mechanism weight; represents the weight of the fully connected layer; Indicates bias; Represents the output after weighting by the attention mechanism; represents the hidden layer output; Indicates the weight of the current input; Represents the final output; represents the activation function; The fused features are input into the CNN-Residual-BiGRU-Attention model for training and learning to obtain a trained health status estimation model.

7. A method for estimating the health status of similar lithium-ion batteries according to claim 6, characterized in that: During the model training process, mean absolute error, mean absolute percentage error and root mean square error are used as evaluation indicators for training.

8. A similar lithium-ion battery health status estimation system, characterized in that: include: The acquisition module is used to collect the voltage, current, capacity and corresponding sampling time data of the lithium-ion battery during the constant current charging stage of the charge and discharge cycle; A curve generation module, used to obtain a capacity increment curve and a constant current charging curve based on voltage, current, capacity and corresponding sampling time data; A curve processing module is used to perform dimensionality reduction processing on the capacity increment curve and the constant current charging curve based on a segmented aggregation approximation algorithm to obtain a low-dimensional capacity increment curve and a low-dimensional constant current charging curve; A feature fusion module is used to fuse the low-dimensional capacity increment curve and the low-dimensional constant current charging curve to obtain a fusion feature; A building block for building a CNN-Residual-BiGRU-Attention-based health state estimation model based on fused features.

9. A similar lithium-ion battery health status estimation device, characterized in that: The method comprises a processor and a memory, wherein when the processor executes the computer program stored in the memory, the method for estimating the health status of a similar lithium-ion battery as claimed in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, wherein when the computer program is executed by a processor, the method for estimating the health status of a similar lithium-ion battery as described in any one of claims 1 to 7 is implemented.

Images