Battery charging and discharging data self-adaptive sampling method and system
By setting the DOC interval threshold and sampling capacity in battery charging and discharging data sampling, data checksum and trimming, and using dynamic downsampling method, the timing characteristics and diversity problems in data processing of different sampling frequencies are solved, and the accuracy of battery state prediction is improved.
Patent Information
- Application Number
- CN202411699309.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-26
- Publication Date
- 2025-03-14
AI Technical Summary
When processing battery charging and discharging data of different sampling frequencies, the prior art ignores timing characteristics and fails to effectively process data diversity, resulting in the limitation of the ability of deep learning models to accurately identify and predict battery states.
An adaptive sampling method for battery charging and discharging data is provided. By setting the DOC interval threshold, the input length threshold, the sampling start power and the sampling cutoff power, data verification, cutting and dynamic downsampling are performed to ensure that the data contains sufficient DOC interval and timing characteristics within a given length.
It improves the diversity of battery charging and discharging data sampling, retains timing characteristics, and improves the accurate prediction ability of deep learning models for battery state.
Smart Images
Figure CN119619854A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of battery detection, and in particular to a method and system for adaptively sampling battery charge and discharge data. Background Art
[0002] During the production and use of batteries, charging and discharging data including voltage and current need to be sampled to detect the battery status or used as a data set to train deep learning models.
[0003] Regarding the sampling of battery charge and discharge data, the sampling frequencies under different working conditions vary greatly. Some sampling frequencies are 1s, while others are 10s. This may lead to large differences in the effective feature quantities contained in the charge and discharge data of the same length. In order to adapt to different sampling frequencies and ensure the consistency of the input data length of the deep learning model, the traditional method is to directly truncate the charge and discharge data of a preset length, or to sample at equal distances. However, there are the following problems:
[0004] This may result in differences in the effective feature quantities contained in the charging and discharging data of the same length, and it is impossible to ensure that the charging and discharging data of different sampling frequencies contain sufficient DOC (depth of charge) interval data, that is, there is a lack of data diversity, and it is impossible to adapt well to the charging and discharging data of different sampling frequencies. It is impossible to ensure that there are sufficient effective features and data diversity in the input data to support the accurate learning and prediction of the deep learning model, which in turn limits the deep learning model's ability to accurately identify and predict the battery status.
[0005] Although using statistical features to process charging and discharging data can solve the problem of sampling frequency to a certain extent, this method ignores the timing characteristics of charging and discharging data, which may lead to information loss and limit the comprehensive understanding of battery behavior.
[0006] In summary, the existing technology not only ignores the timing characteristics when processing charging and discharging data with different sampling frequencies, but also fails to effectively handle data diversity. Therefore, how to provide a battery charging and discharging data adaptive sampling method and system to improve the diversity of battery charging and discharging data sampling and retain the timing characteristics has become a technical problem to be solved urgently. Summary of the invention
[0007] The technical problem to be solved by the present invention is to provide a method and system for adaptively sampling battery charge and discharge data, so as to improve the diversity of battery charge and discharge data sampling and retain the timing characteristics.
[0008] In a first aspect, the present invention provides a method for adaptively sampling battery charge and discharge data, comprising the following steps:
[0009] Step S1, setting a DOC interval threshold d, an input length threshold input_size, a sampling start power star_soc and a sampling end power end_soc;
[0010] Step S2, acquiring charge and discharge data of length L, and verifying the charge and discharge data based on the DOC interval threshold d and the input length threshold input_size;
[0011] Step S3, cutting the charge and discharge data based on the sampling start power star_soc and the sampling end power end_soc to obtain a plurality of first charge and discharge electronic data;
[0012] Step S4, verifying each first charge and discharge electronic data based on the input length threshold input_size;
[0013] Step S5: dynamically downsample each of the first charge and discharge electronic data and output the sampling result.
[0014] Furthermore, in step S1, the value of the sampling start power star_soc is a random number between [0, (Dd)], and the value of the sampling end power end_soc is a random number between [star_soc+d, maximum soc]; D is the difference between the maximum soc and the minimum soc in the charge and discharge data, that is, the DOC interval.
[0015] Furthermore, the step S2 is specifically as follows:
[0016] Obtain charging and discharging data of length L, and determine whether the DOC interval of the charging and discharging data is greater than the DOC interval threshold d. If not, end the process; if yes, then:
[0017] Determine whether L is greater than an input length threshold input_size. If not, fill the charge and discharge data to a length equal to the input length threshold input_size based on a preset filling character and output it as a sampling result. If yes, proceed to step S3.
[0018] Furthermore, the step S4 is specifically as follows:
[0019] It is determined whether the length of each of the first charge and discharge electronic data is greater than an input length threshold input_size. If not, the first charge and discharge electronic data is padded to a length equal to the input length threshold input_size based on a preset padding character and output as a sampling result. If yes, proceed to step S5.
[0020] Furthermore, the step S5 is specifically as follows:
[0021] A sampling frequency is set. When the charge and discharge current is a constant current, the sampling frequency is a1. When the charge and discharge current is switched, the sampling frequency is a2, where a1<a2.
[0022] Determine whether the length of the first charge-discharge electronic data is greater than 2*input_size; if so, dynamically downsample each of the first charge-discharge electronic data based on the sampling frequency to obtain corresponding second charge-discharge electronic data; if not, perform a random elimination operation on the first charge-discharge electronic data;
[0023] Determine whether the length of the second charge-discharge electronic data is greater than 2*input_size, if so, continue dynamic downsampling; if not, perform a random elimination operation on the second charge-discharge electronic data;
[0024] The random elimination operation on the first charge-discharge electronic data or the second charge-discharge electronic data is specifically performed as follows:
[0025] Randomly remove data points in the first charge-discharge electronic data or the second charge-discharge electronic data until the length of the first charge-discharge electronic data or the second charge-discharge electronic data is equal to input_size, and output the first charge-discharge electronic data or the second charge-discharge electronic data as a sampling result.
[0026] In a second aspect, the present invention provides a battery charging and discharging data adaptive sampling system, comprising the following modules:
[0027] An initialization module is used to set a DOC interval threshold d, an input length threshold input_size, a sampling start power star_soc and a sampling end power end_soc;
[0028] A charge and discharge data verification module, used to obtain charge and discharge data of length L, and verify the charge and discharge data based on the DOC interval threshold d and the input length threshold input_size;
[0029] A charge and discharge data trimming module, used for trimming the charge and discharge data based on the sampling start power star_soc and the sampling end power end_soc to obtain a plurality of first charge and discharge electronic data;
[0030] A first charge and discharge electronic data verification module, used to verify each first charge and discharge electronic data based on the input length threshold input_size;
[0031] The dynamic down-sampling module is used to dynamically down-sample each of the first charge and discharge electronic data and output the sampling result.
[0032] Furthermore, in the initialization module, the value of the sampling starting power star_soc is a random number between [0, (Dd)], and the value of the sampling end power end_soc is a random number between [star_soc+d, maximum soc]; D is the difference between the maximum soc and the minimum soc in the charge and discharge data, that is, the DOC interval.
[0033] Furthermore, the charge and discharge data verification module is specifically used for:
[0034] Obtain charging and discharging data of length L, and determine whether the DOC interval of the charging and discharging data is greater than the DOC interval threshold d. If not, end the process; if yes, then:
[0035] Determine whether L is greater than an input length threshold input_size. If not, fill the charge and discharge data to a length equal to the input length threshold input_size based on a preset fill character and output it as a sampling result. If yes, enter the charge and discharge data clipping module.
[0036] Furthermore, the first charge and discharge electronic data verification module is specifically used for:
[0037] It is determined whether the length of each of the first charge and discharge electronic data is greater than an input length threshold input_size. If not, the first charge and discharge electronic data is padded to a length equal to the input length threshold input_size based on a preset padding character, and output as a sampling result; if so, a dynamic downsampling module is entered.
[0038] Furthermore, the dynamic downsampling module is specifically used for:
[0039] A sampling frequency is set. When the charge and discharge current is a constant current, the sampling frequency is a1. When the charge and discharge current is switched, the sampling frequency is a2, where a1<a2.
[0040] Determine whether the length of the first charge-discharge electronic data is greater than 2*input_size; if so, dynamically downsample each of the first charge-discharge electronic data based on the sampling frequency to obtain corresponding second charge-discharge electronic data; if not, perform a random elimination operation on the first charge-discharge electronic data;
[0041] Determine whether the length of the second charge-discharge electronic data is greater than 2*input_size, if so, continue dynamic downsampling; if not, perform a random elimination operation on the second charge-discharge electronic data;
[0042] The random elimination operation on the first charge-discharge electronic data or the second charge-discharge electronic data is specifically performed as follows:
[0043] Randomly remove data points in the first charge-discharge electronic data or the second charge-discharge electronic data until the length of the first charge-discharge electronic data or the second charge-discharge electronic data is equal to input_size, and output the first charge-discharge electronic data or the second charge-discharge electronic data as a sampling result.
[0044] The advantages of the present invention are:
[0045] By setting the DOC interval threshold d, input length threshold input_size, sampling start power star_soc and sampling end power end_soc; then obtaining the charge and discharge data of length L, verifying the charge and discharge data based on the DOC interval threshold d and the input length threshold input_size, trimming the charge and discharge data based on the sampling start power star_soc and the sampling end power end_soc, obtaining a number of first charge and discharge electronic data, verifying each first charge and discharge electronic data based on the input length threshold input_size, and finally dynamically downsampling each first charge and discharge electronic data, and outputting the sampling result; since only the charge and discharge data of the DOC interval > d are sampled, and the basis The charge and discharge data are trimmed at star_soc and end_soc, the value of star_soc is a random number between [0, (Dd)], and the value of end_soc is a random number between [star_soc+d, maximum soc]. This not only ensures that the DOC interval of the first charge and discharge electronic data after trimming is greater than d, that is, different sampling frequencies can ensure sufficient DOC intervals within a given length, but also ensures the randomness of sampling; through dynamic downsampling, the DOC intervals sampled from different batches are different, ensuring the diversity of the data. Compared with filling methods such as interpolation, the detailed features of the current and voltage curves in the charge and discharge data are retained, which ultimately greatly improves the diversity of battery charge and discharge data sampling and retains the timing characteristics. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] The present invention will be further described below in conjunction with embodiments with reference to the accompanying drawings.
[0047] Figure 1 The present invention is a flow chart of a method for adaptively sampling battery charge and discharge data.
[0048] Figure 2 It is a structural schematic diagram of a battery charging and discharging data adaptive sampling system of the present invention.
[0049] Figure 3 It is a schematic diagram of the process of the present invention.
[0050] Figure 4 It is a schematic diagram of the sampling effect of the present invention. DETAILED DESCRIPTION
[0051] The technical solution in the embodiment of the present application has the following overall idea: only the charge and discharge data with a DOC interval of > d are sampled, and the charge and discharge data are trimmed based on star_soc and end_soc, the value of star_soc is a random number between [0, (Dd)], and the value of end_soc is a random number between [star_soc+d, maximum soc]. This not only ensures that the DOC interval of the first charge and discharge electronic data after trimming is > d, that is, different sampling frequencies can ensure sufficient DOC intervals within a given length, but also ensures the randomness of sampling; through dynamic downsampling, the DOC intervals sampled from different batches are different, ensuring the diversity of the data. Compared with filling methods such as interpolation, the detailed features of the current and voltage curves in the charge and discharge data are retained, thereby improving the diversity of the battery charge and discharge data sampling and retaining the timing characteristics.
[0052] Please refer to Figures 1 to 4 As shown, a preferred embodiment of a method for adaptively sampling battery charge and discharge data of the present invention comprises the following steps:
[0053] Step S1, setting a DOC interval threshold d, an input length threshold input_size, a sampling start power star_soc and a sampling end power end_soc; the input length threshold input_size is the data input length of the deep learning model;
[0054] Step S2, acquiring charge and discharge data of length L, and verifying the charge and discharge data based on the DOC interval threshold d and the input length threshold input_size;
[0055] Step S3, cutting the charge and discharge data based on the sampling start power star_soc and the sampling end power end_soc to obtain a plurality of first charge and discharge electronic data;
[0056] Step S4, verifying each first charge and discharge electronic data based on the input length threshold input_size;
[0057] Step S5: dynamically downsample each of the first charge and discharge electronic data and output the sampling result.
[0058] The present invention focuses on the information imbalance of different sampling frequencies and proposes adaptive dynamic downsampling to ensure that within a given length, the charging and discharging data of different sampling frequencies can guarantee a sufficient DOC interval. At the same time, the DOC intervals (DOC lengths) of data sampled from different batches are different to ensure data diversity; and with a view to retaining the timing characteristics of the charging and discharging data, the dynamic downsampling does not use filling methods such as interpolation while reducing the sampling frequency, but designs different downsampling frequencies for different working conditions to ensure that the detailed features of the current, voltage and other curves in the input charging and discharging data are retained, thereby improving the deep learning model's ability to accurately predict the battery status.
[0059] In step S1, the sampling start power star_soc is a random number between [0, (Dd)], and the sampling end power end_soc is a random number between [star_soc+d, maximum soc]; D is the difference between the maximum soc and the minimum soc in the charge and discharge data, that is, the DOC interval.
[0060] The values of star_soc and end_soc are designed in this way to ensure that the DOC interval obtained by the final sampling start power and sampling cut-off power can be greater than d, while ensuring the randomness and diversity of data after sampling.
[0061] The step S2 is specifically as follows:
[0062] Obtain charging and discharging data of length L, and determine whether the DOC interval of the charging and discharging data is greater than the DOC interval threshold d. If not, end the process; if yes, then:
[0063] Determine whether L is greater than an input length threshold input_size. If not, fill the charge and discharge data to a length equal to the input length threshold input_size based on a preset filling character and output it as a sampling result. If yes, proceed to step S3.
[0064] The step S4 is specifically as follows:
[0065] It is determined whether the length of each of the first charge and discharge electronic data is greater than an input length threshold input_size. If not, the first charge and discharge electronic data is padded to a length equal to the input length threshold input_size based on a preset padding character and output as a sampling result. If yes, proceed to step S5.
[0066] The step S5 is specifically as follows:
[0067] A sampling frequency is set. When the charge and discharge current is a constant current, the sampling frequency is a1. When the charge and discharge current is switched, the sampling frequency is a2, where a1<a2.
[0068] Dynamic downsampling refers to dynamically changing the sampling frequency according to the change of current. When the current is constant, a relatively low sampling frequency is used, and when the current switches, a relatively high sampling frequency is used. This is because when the current switches, there will be richer detailed information about characteristics such as voltage. Using a relatively high sampling frequency can retain the feature details as much as possible.
[0069] Determine whether the length of the first charge-discharge electronic data is greater than 2*input_size; if so, dynamically downsample each of the first charge-discharge electronic data based on the sampling frequency to obtain corresponding second charge-discharge electronic data; if not, perform a random elimination operation on the first charge-discharge electronic data;
[0070] Determine whether the length of the second charge-discharge electronic data is greater than 2*input_size, if so, continue dynamic downsampling; if not, perform a random elimination operation on the second charge-discharge electronic data;
[0071] The random elimination operation on the first charge-discharge electronic data or the second charge-discharge electronic data is specifically performed as follows:
[0072] Randomly remove data points in the first charge-discharge electronic data or the second charge-discharge electronic data until the length of the first charge-discharge electronic data or the second charge-discharge electronic data is equal to input_size, and output the first charge-discharge electronic data or the second charge-discharge electronic data as a sampling result.
[0073] A preferred embodiment of a battery charging and discharging data adaptive sampling system of the present invention includes the following modules:
[0074] An initialization module is used to set a DOC interval threshold d, an input length threshold input_size, a sampling start power star_soc, and a sampling end power end_soc; the input length threshold input_size is the data input length of the deep learning model;
[0075] A charge and discharge data verification module, used to obtain charge and discharge data of length L, and verify the charge and discharge data based on the DOC interval threshold d and the input length threshold input_size;
[0076] A charge and discharge data trimming module, used for trimming the charge and discharge data based on the sampling start power star_soc and the sampling end power end_soc to obtain a plurality of first charge and discharge electronic data;
[0077] A first charge and discharge electronic data verification module, used to verify each first charge and discharge electronic data based on the input length threshold input_size;
[0078] The dynamic down-sampling module is used to dynamically down-sample each of the first charge and discharge electronic data and output the sampling result.
[0079] The present invention focuses on the information imbalance of different sampling frequencies and proposes adaptive dynamic downsampling to ensure that within a given length, the charging and discharging data of different sampling frequencies can guarantee a sufficient DOC interval. At the same time, the DOC intervals (DOC lengths) of data sampled from different batches are different to ensure data diversity; and with a view to retaining the timing characteristics of the charging and discharging data, the dynamic downsampling does not use filling methods such as interpolation while reducing the sampling frequency, but designs different downsampling frequencies for different working conditions to ensure that the detailed features of the current, voltage and other curves in the input charging and discharging data are retained, thereby improving the deep learning model's ability to accurately predict the battery status.
[0080] In the initialization module, the value of the sampling start power star_soc is a random number between [0, (Dd)], and the value of the sampling end power end_soc is a random number between [star_soc+d, maximum soc]; D is the difference between the maximum soc and the minimum soc in the charge and discharge data, that is, the DOC interval.
[0081] The values of star_soc and end_soc are designed in this way to ensure that the DOC interval obtained by the final sampling start power and sampling cut-off power can be greater than d, while ensuring the randomness and diversity of data after sampling.
[0082] The charge and discharge data verification module is specifically used for:
[0083] Obtain charging and discharging data of length L, and determine whether the DOC interval of the charging and discharging data is greater than the DOC interval threshold d. If not, end the process; if yes, then:
[0084] Determine whether L is greater than an input length threshold input_size. If not, fill the charge and discharge data to a length equal to the input length threshold input_size based on a preset fill character and output it as a sampling result. If yes, enter the charge and discharge data clipping module.
[0085] The first charge and discharge electronic data verification module is specifically used for:
[0086] It is determined whether the length of each of the first charge and discharge electronic data is greater than an input length threshold input_size. If not, the first charge and discharge electronic data is padded to a length equal to the input length threshold input_size based on a preset padding character, and output as a sampling result; if so, a dynamic downsampling module is entered.
[0087] The dynamic downsampling module is specifically used for:
[0088] A sampling frequency is set. When the charge and discharge current is a constant current, the sampling frequency is a1. When the charge and discharge current is switched, the sampling frequency is a2, where a1<a2.
[0089] Dynamic downsampling refers to dynamically changing the sampling frequency according to the change of current. When the current is constant, a relatively low sampling frequency is used, and when the current switches, a relatively high sampling frequency is used. This is because when the current switches, there will be richer detailed information about characteristics such as voltage. Using a relatively high sampling frequency can retain the feature details as much as possible.
[0090] Determine whether the length of the first charge-discharge electronic data is greater than 2*input_size; if so, dynamically downsample each of the first charge-discharge electronic data based on the sampling frequency to obtain corresponding second charge-discharge electronic data; if not, perform a random elimination operation on the first charge-discharge electronic data;
[0091] Determine whether the length of the second charge-discharge electronic data is greater than 2*input_size, if so, continue dynamic downsampling; if not, perform a random elimination operation on the second charge-discharge electronic data;
[0092] The random elimination operation on the first charge-discharge electronic data or the second charge-discharge electronic data is specifically performed as follows:
[0093] Randomly remove data points in the first charge-discharge electronic data or the second charge-discharge electronic data until the length of the first charge-discharge electronic data or the second charge-discharge electronic data is equal to input_size, and output the first charge-discharge electronic data or the second charge-discharge electronic data as a sampling result.
[0094] like Figure 4 As shown, input_size is set to 128, and the DOC interval after sampling is >40%. The first line is the charge and discharge data, and the second line is the sampling result after dynamic downsampling. It can be seen that after sampling using the method of the present invention, the original form of characteristics such as voltage and current in the sampled SOC interval is retained, and the DOC interval after sampling can meet the requirements. Figure 4 The original SOC range of the left figure is 0.2-1.0, the number of sampling points is greater than 500 points, and the SOC range after sampling is 0.2-0.8, and the number of sampling points is 128; Figure 4 The original SOC range of the right figure is 0-1, the number of sampling points is greater than 250 points, and the SOC range after sampling is 0.4-1.0, and the number of sampling points is 128.
[0095] In summary, the advantages of the present invention are:
[0096] By setting the DOC interval threshold d, input length threshold input_size, sampling start power star_soc and sampling end power end_soc; then obtaining the charge and discharge data of length L, verifying the charge and discharge data based on the DOC interval threshold d and the input length threshold input_size, trimming the charge and discharge data based on the sampling start power star_soc and the sampling end power end_soc, obtaining a number of first charge and discharge electronic data, verifying each first charge and discharge electronic data based on the input length threshold input_size, and finally dynamically downsampling each first charge and discharge electronic data, and outputting the sampling result; since only the charge and discharge data of the DOC interval > d are sampled, and the basis The charge and discharge data are trimmed at star_soc and end_soc, the value of star_soc is a random number between [0, (Dd)], and the value of end_soc is a random number between [star_soc+d, maximum soc]. This not only ensures that the DOC interval of the first charge and discharge electronic data after trimming is greater than d, that is, different sampling frequencies can ensure sufficient DOC intervals within a given length, but also ensures the randomness of sampling; through dynamic downsampling, the DOC intervals sampled from different batches are different, ensuring the diversity of the data. Compared with filling methods such as interpolation, the detailed features of the current and voltage curves in the charge and discharge data are retained, which ultimately greatly improves the diversity of battery charge and discharge data sampling and retains the timing characteristics.
[0097] Although the specific implementation modes of the present invention are described above, those skilled in the art should understand that the specific implementation modes described are only illustrative and are not intended to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included in the scope of protection of the claims of the present invention.
Claims
1. A method for adaptively sampling battery charge and discharge data, characterized in that: The steps include: Step S1, setting a DOC interval threshold d, an input length threshold input_size, a sampling start power star_soc and a sampling end power end_soc; Step S2, acquiring charge and discharge data of length L, and verifying the charge and discharge data based on the DOC interval threshold d and the input length threshold input_size; Step S3, cutting the charge and discharge data based on the sampling start power star_soc and the sampling end power end_soc to obtain a plurality of first charge and discharge electronic data; Step S4, verifying each first charge and discharge electronic data based on the input length threshold input_size; Step S5: dynamically downsample each of the first charge and discharge electronic data and output the sampling result.
2. A battery charge and discharge data adaptive sampling method as claimed in claim 1, characterized in that: In step S1, the sampling start power star_soc is a random number between [0, (Dd)], and the sampling end power end_soc is a random number between [star_soc+d, maximum soc]; D is the difference between the maximum soc and the minimum soc in the charge and discharge data, that is, the DOC interval.
3. A battery charge and discharge data adaptive sampling method as claimed in claim 1, characterized in that: The step S2 is specifically as follows: Obtain charging and discharging data of length L, and determine whether the DOC interval of the charging and discharging data is greater than the DOC interval threshold d. If not, end the process; if yes, then: Determine whether L is greater than an input length threshold input_size. If not, fill the charge and discharge data to a length equal to the input length threshold input_size based on a preset filling character and output it as a sampling result. If yes, proceed to step S3.
4. A battery charge and discharge data adaptive sampling method as claimed in claim 1, characterized in that: The step S4 is specifically as follows: It is determined whether the length of each of the first charge and discharge electronic data is greater than an input length threshold input_size. If not, the first charge and discharge electronic data is padded to a length equal to the input length threshold input_size based on a preset padding character and output as a sampling result. If yes, proceed to step S5.
5. A battery charge and discharge data adaptive sampling method as claimed in claim 1, characterized in that: The step S5 is specifically as follows: A sampling frequency is set. When the charge and discharge current is a constant current, the sampling frequency is a1. When the charge and discharge current is switched, the sampling frequency is a2, where a1<a2. Determine whether the length of the first charge-discharge electronic data is greater than 2*input_size; if so, dynamically downsample each of the first charge-discharge electronic data based on the sampling frequency to obtain corresponding second charge-discharge electronic data; if not, perform a random elimination operation on the first charge-discharge electronic data; Determine whether the length of the second charge-discharge electronic data is greater than 2*input_size, if so, continue dynamic downsampling; if not, perform a random elimination operation on the second charge-discharge electronic data; The random elimination operation on the first charge-discharge electronic data or the second charge-discharge electronic data is specifically performed as follows: Randomly remove data points in the first charge-discharge electronic data or the second charge-discharge electronic data until the length of the first charge-discharge electronic data or the second charge-discharge electronic data is equal to input_size, and output the first charge-discharge electronic data or the second charge-discharge electronic data as a sampling result.
6. A battery charge and discharge data adaptive sampling system, characterized in that: Includes the following modules: An initialization module is used to set a DOC interval threshold d, an input length threshold input_size, a sampling start power star_soc and a sampling end power end_soc; A charge and discharge data verification module, used to obtain charge and discharge data of length L, and verify the charge and discharge data based on the DOC interval threshold d and the input length threshold input_size; A charge and discharge data trimming module, used for trimming the charge and discharge data based on the sampling start power star_soc and the sampling end power end_soc to obtain a plurality of first charge and discharge electronic data; A first charge and discharge electronic data verification module, used to verify each first charge and discharge electronic data based on the input length threshold input_size; The dynamic down-sampling module is used to dynamically down-sample each of the first charge and discharge electronic data and output the sampling result.
7. A battery charge and discharge data adaptive sampling system as claimed in claim 6, characterized in that: In the initialization module, the value of the sampling start power star_soc is a random number between [0, (Dd)], and the value of the sampling end power end_soc is a random number between [star_soc+d, maximum soc]; D is the difference between the maximum soc and the minimum soc in the charge and discharge data, that is, the DOC interval.
8. A battery charge and discharge data adaptive sampling system as claimed in claim 6, characterized in that: The charge and discharge data verification module is specifically used for: Obtain charging and discharging data of length L, and determine whether the DOC interval of the charging and discharging data is greater than the DOC interval threshold d. If not, end the process; if yes, then: Determine whether L is greater than an input length threshold input_size. If not, fill the charge and discharge data to a length equal to the input length threshold input_size based on a preset fill character and output it as a sampling result. If yes, enter the charge and discharge data clipping module.
9. A battery charge and discharge data adaptive sampling system as claimed in claim 6, characterized in that: The first charge and discharge electronic data verification module is specifically used for: It is determined whether the length of each of the first charge and discharge electronic data is greater than an input length threshold input_size. If not, the first charge and discharge electronic data is padded to a length equal to the input length threshold input_size based on a preset padding character, and output as a sampling result; if so, a dynamic downsampling module is entered.
10. A battery charge and discharge data adaptive sampling system as claimed in claim 6, characterized in that: The dynamic downsampling module is specifically used for: A sampling frequency is set. When the charge and discharge current is a constant current, the sampling frequency is a1. When the charge and discharge current is switched, the sampling frequency is a2, where a1<a2. Determine whether the length of the first charge-discharge electronic data is greater than 2*input_size; if so, dynamically downsample each of the first charge-discharge electronic data based on the sampling frequency to obtain corresponding second charge-discharge electronic data; if not, perform a random elimination operation on the first charge-discharge electronic data; Determine whether the length of the second charge-discharge electronic data is greater than 2*input_size, if so, continue dynamic downsampling; if not, perform a random elimination operation on the second charge-discharge electronic data; The random elimination operation on the first charge-discharge electronic data or the second charge-discharge electronic data is specifically performed as follows: Randomly remove data points in the first charge-discharge electronic data or the second charge-discharge electronic data until the length of the first charge-discharge electronic data or the second charge-discharge electronic data is equal to input_size, and output the first charge-discharge electronic data or the second charge-discharge electronic data as a sampling result.