Method and device for quickly estimating battery capacity

Abstract

The present invention relates to a method and device for rapidly estimating battery capacity. The method comprises obtaining first current-time data from a battery under test during a constant-voltage charging phase and second current-time data from a rest phase after constant-voltage charging; the first and second current-time data are input into a pre-trained battery capacity estimation model to obtain the battery capacity. Advantageously, the method significantly shortens battery capacity estimation time and improves battery production efficiency. The method shortens the traditional full charge and discharge capacity testing process from four hours to less than 30 minutes, greatly improving detection efficiency and significantly reducing the electrical energy consumed by full charge and discharge, significantly reducing energy consumption. The method also features high detection accuracy and low detection error, with a maximum detection error of less than 1.5% and an average detection error of only 0.42%.

Description

Technical Field

[0001] The present invention relates to the field of battery detection and analysis, and in particular to a method and device for quickly estimating battery capacity, a computer device, and a computer-readable storage medium. Background Art

[0002] Battery capacity is an internal state of the battery and cannot be directly measured using sensors. Capacity is a key indicator for evaluating battery status and performance, directly impacting the quality of battery packs and the accuracy of battery recycling. However, existing capacity testing methods require fully charging and discharging the battery—that is, fully charging and then discharging it to calculate the battery's discharge capacity. This full-charge and discharge testing method is time-consuming, energy-intensive, and costly. The industry urgently needs a rapid battery capacity testing method to improve battery testing efficiency at both the production and decommissioning stages.

[0003] Chinese invention patent CN109828220A discloses a linear assessment method for lithium-ion battery health status. This method requires feature extraction using complete constant-current charging data and constant-voltage charging data from the lithium-ion battery. However, this method is time-consuming and extracts a large number of complex features, making it difficult to complete battery health assessment in a short period of time.

[0004] Chinese invention patent CN115598557A discloses a method for estimating the state of health (SOH) of lithium batteries based on constant-voltage charging current. This method extracts statistical features of current changes during constant-voltage charging, such as minimum, maximum, average, and standard deviation, to estimate the battery's state of health. However, this method extracts a large number of features, has a complex calculation process, and is inefficient.

[0005] Chinese invention patent CN117054914A discloses a method for estimating the state of health (SOH) of lithium batteries in energy storage power stations based on charging current curves. This method extracts features from the battery's constant-voltage charging current curve, such as the skewness and Shannon entropy of the battery data, the skewness and Shannon entropy of the first-order difference sequence of the current data, and the charging market, to assess the battery's state of health. However, this method extracts a large number of features, has a complex calculation process, and is inefficient.

[0006] Chinese invention patent CN115097344A discloses a method for collaborative end-to-end estimation of battery health status based on constant voltage charging segments. This method requires the establishment of a battery equivalent circuit model to design features. However, online identification of the parameters of the battery equivalent circuit model is difficult and easily affected by real-world noise, resulting in model inaccuracies and, in turn, inaccurate features extracted from the model, affecting the accuracy of battery health status estimation. Furthermore, the feature extraction process is difficult and complex, making it unsuitable for practical online applications.

[0007] Chinese invention patent CN111303379A discloses a battery health state estimation method based on local constant voltage charging data. This method predicts the current variation characteristics of the remaining CV charging period by establishing a second-order RC equivalent circuit model of the battery. However, this method is susceptible to interference from real-world environmental noise, resulting in model inaccuracies. Furthermore, the feature extraction process is difficult and complex, making it unsuitable for practical online applications.

[0008] At present, there is no effective solution to the problems in related technologies such as long time consumption, large number of feature extraction, complex calculation process, low accuracy, and susceptibility to interference from real environmental noise. Summary of the Invention

[0009] The purpose of the present invention is to address the deficiencies in the prior art and provide a method, device, computer equipment and computer-readable storage medium for rapid battery capacity estimation, so as to solve the problems existing in the related art such as long time consumption, large number of feature extraction quantities, complex calculation process, low accuracy and susceptibility to interference from real environmental noise.

[0010] To achieve the above object, the technical solution adopted by the present invention is:

[0011] In a first aspect, the present invention provides a method for quickly estimating battery capacity, comprising:

[0012] Acquire first current-time data of the battery under test during a constant voltage charging phase and second current-time data of the battery under test during a rest phase after constant voltage charging;

[0013] The first current-time data and the second current-time data are input into a pre-trained battery capacity estimation model to obtain the battery capacity.

[0014] In some embodiments, obtaining first current-time data of the constant voltage charging phase of the battery under test includes:

[0015] Acquire a plurality of first time data and a plurality of corresponding first current data in a constant voltage charging phase of the battery to be tested;

[0016] Constructing a first current-time curve with time as the abscissa and current as the ordinate;

[0017] Fitting the first current-time curve using a first polynomial function to obtain a first current-time fitting curve and a plurality of first polynomial coefficients;

[0018] The first current-time fitting curve and the first polynomial coefficients are used to form first current-time data.

[0019] In some embodiments, the first polynomial function is at least a third-order polynomial.

[0020] In some embodiments, the first polynomial function is a 5th-order polynomial, and several of the first polynomial coefficients are respectively -3.94403143e-06, 1.57437867e-04, -2.63657145e-03, 2.53981014e-02, -1.62843435e-01, and 6.33374973e-01.

[0021] In some embodiments, the first polynomial function is a 5th-order polynomial, and several of the first polynomial coefficients are -4.36179037e-07, 2.85444822e-05, -8.29644917e-04, 1.54047465e-02, -2.06722228e-01, and 1.77794943e+00, respectively.

[0022] In some embodiments, obtaining second current-time data of the battery under test during a rest phase after constant voltage charging includes:

[0023] Acquire a plurality of second time data and corresponding second current data of the battery under test during a rest phase after constant voltage charging;

[0024] Constructing a second current-time curve with time as the abscissa and current as the ordinate;

[0025] Fitting the second current-time curve using a second polynomial function to obtain a second current-time fitting curve and a plurality of second polynomial coefficients;

[0026] The second current-time fitting curve and a plurality of the second polynomial coefficients are used to form second current-time data.

[0027] In some embodiments, the second polynomial function is at least a third-order polynomial.

[0028] In some embodiments, the second polynomial function is a 5th-order polynomial, and several of the second polynomial coefficients are respectively -1.67312169e-05, 2.42717006e-04, -1.33710141e-03, 3.59136982e-03, -6.03967437e-03, and 2.55575012e+00.

[0029] In some embodiments, the second polynomial function is a fifth-order polynomial, and several second polynomial coefficients are -9.22933393e-06, 1.39798220e-04, -8.28311961e-04, 2.48732516e-03, -4.36348653e-03, and 3.64583619e+00, respectively.

[0030] In some embodiments, the training method of the battery capacity estimation model includes:

[0031] Obtain historical capacity data of several batteries;

[0032] Obtaining battery capacity data, first current-time data of a constant-voltage charging phase, and second current-time data of a rest phase after constant-voltage charging based on the historical capacity-dividing data;

[0033] The first current-time data and the second current-time data are used as input values ​​and the battery capacity data is used as output values, and are input into a battery capacity estimation model to train the battery capacity estimation model, so as to obtain a trained battery capacity estimation model.

[0034] In some embodiments, the battery capacity estimation model training method further includes:

[0035] Obtain historical capacity data of several batteries;

[0036] Obtaining, based on the historical capacity-dividing data, first current-time data during a constant-voltage charging phase and second current-time data during a rest phase after constant-voltage charging;

[0037] The first current-time data and the second current-time data are used as input values ​​and input into the trained battery capacity estimation model to test the battery capacity estimation model.

[0038] In some embodiments, the battery capacity estimation model training method further includes:

[0039] Get cross-validation results;

[0040] According to the cross-validation result, the parameters of the battery capacity estimation model are adjusted.

[0041] In a second aspect, a device for quickly estimating battery capacity is provided, for executing the method for quickly estimating battery capacity as described in the first aspect, comprising:

[0042] A current-time data acquisition module is used to acquire first current-time data of the battery under test during the constant voltage charging phase and second current-time data of the battery under test during the rest phase after the constant voltage charging phase;

[0043] The capacity estimation module is configured to input the first current-time data and the second current-time data into a pre-trained battery capacity estimation model to obtain the battery capacity.

[0044] In some embodiments, further comprising:

[0045] A training data acquisition module is used to acquire historical capacity data of a plurality of batteries, and obtain battery capacity data, first current-time data of a constant voltage charging phase, and second current-time data of a rest phase after constant voltage charging based on the historical capacity data;

[0046] A model training module is used to input the first current-time data and the second current-time data as input values ​​and the battery capacity data as output values ​​into a battery capacity estimation model to train the battery capacity estimation model to obtain a trained battery capacity estimation model.

[0047] In some embodiments, further comprising:

[0048] A test data acquisition module is used to obtain historical capacity data of a plurality of batteries, and obtain first current-time data of a constant voltage charging stage and second current-time data of a rest stage after constant voltage charging based on the historical capacity data;

[0049] The model testing module is used to input the first current-time data and the second current-time data as input values ​​into the trained battery capacity estimation model to test the battery capacity estimation model.

[0050] In some embodiments, further comprising:

[0051] The parameter optimization module is used to obtain cross-validation results and adjust the parameters of the battery capacity estimation model according to the cross-validation results.

[0052] In a third aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for rapidly estimating battery capacity as described in the first aspect when executing the computer program.

[0053] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for quickly estimating battery capacity as described in the first aspect.

[0054] The present invention adopts the above technical solution, which has the following technical effects compared with the prior art:

[0055] 1) Significantly shortened battery capacity separation time and improved battery production efficiency;

[0056] 2) The traditional full charge and discharge capacity test process is shortened from 4 hours to less than 30 minutes, greatly improving detection efficiency, significantly reducing the power consumed by full charge and discharge, and significantly reducing energy consumption;

[0057] 3) High detection accuracy and small detection error. The maximum detection error is less than 1.5% and the average detection error is only 0.42%. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a flowchart (1) of a method for rapid battery capacity estimation according to an embodiment of the present invention;

[0059] Figure 2 is a flowchart (II) of a method for rapid battery capacity estimation according to an embodiment of the present invention;

[0060] Figure 3 is a flowchart (III) of a method for rapid battery capacity estimation according to an embodiment of the present invention;

[0061] Figure 4 is a flowchart (four) of a method for quickly estimating battery capacity according to an embodiment of the present invention;

[0062] Figure 5 is a flowchart (V) of a method for quickly estimating battery capacity according to an embodiment of the present invention;

[0063] Figure 6 is a flowchart (six) of a method for rapid battery capacity estimation according to an embodiment of the present invention;

[0064] Figure 7 is a framework diagram of a device for quickly estimating battery capacity according to an embodiment of the present invention;

[0065] Figure 8 is a voltage variation curve diagram of a sodium ion battery fully charged and discharged according to an embodiment of the present invention;

[0066] Figure 9 2 is a schematic diagram of polynomial fitting of current-time data during a short-time constant-voltage charging phase of a sodium ion battery according to an embodiment of the present invention;

[0067] Figure 10 3 is a schematic diagram of polynomial fitting of current-time data during a rest period after short-time constant-voltage charging of a sodium ion battery according to an embodiment of the present invention;

[0068] Figure 11 is a comparison chart of the capacity detection value of the sodium ion battery according to an embodiment of the present invention and the reference value;

[0069] Figure 12 is an error diagram for rapid estimation of sodium ion battery capacity according to an embodiment of the present invention;

[0070] Figure 13 is a voltage variation curve diagram of a full charge and discharge test of a lithium-ion battery according to an embodiment of the present invention;

[0071] Figure 142 is a schematic diagram of polynomial fitting of current-time data during a short-time constant-voltage charging phase of a lithium-ion battery according to an embodiment of the present invention;

[0072] Figure 15 2 is a schematic diagram of polynomial fitting of current-time data during a rest period after short-time constant voltage charging of a lithium-ion battery according to an embodiment of the present invention;

[0073] Figure 16 is a comparison chart of the capacity detection value of a lithium-ion battery according to an embodiment of the present invention and a reference value;

[0074] Figure 17 4 is an error diagram of rapid estimation of lithium-ion battery capacity according to an embodiment of the present invention.

[0075] The accompanying drawings are marked as follows: 700, battery capacity rapid estimation device; 710, current-time data acquisition module; 720, capacity estimation module; 730, training data acquisition module; 740, model training module; 750, test data acquisition module; 760, model testing module; 770, parameter optimization module. DETAILED DESCRIPTION

[0076] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is described and illustrated below in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application. Based on the embodiments provided in this application, all other embodiments obtained by ordinary technicians in this field without making any creative work are within the scope of protection of this application.

[0077] Obviously, the drawings described below are merely examples or embodiments of the present application. Those skilled in the art can, without inventive effort, apply the present application to other similar scenarios based on these drawings. Furthermore, it is also understood that, although the effort involved in such a development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, changes in design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as an insufficiency of the content disclosed in this application.

[0078] References to "embodiments" in this application mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it refer to independent or alternative embodiments that are mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described in this application may be combined with other embodiments unless there is a conflict.

[0079] Unless otherwise defined, technical or scientific terms used herein shall have the ordinary meaning as understood by persons of ordinary skill in the art to which this application belongs. The terms "a," "an," "an," "the," and similar expressions used herein do not denote quantitative limitations and may refer to either the singular or the plural. The terms "comprise," "include," "have," and any variations thereof, used herein, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements (units) is not limited to the listed steps or elements but may also include steps or elements not listed, or may include other steps or elements inherent to the process, method, product, or apparatus. The terms "connected," "connected," "coupled," and similar expressions used herein are not limited to physical or mechanical connections but may include electrical connections, whether direct or indirect. As used herein, "plurality" means two or more. "And / or" describes an association between associated objects, indicating that three possible relationships exist. For example, "A and / or B" may mean: A exists alone; A and B exist simultaneously; or B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0080] Example 1

[0081] This embodiment relates to a method for quickly estimating battery capacity according to the present invention.

[0082] Figure 1 FIG1 is a flow chart of a method for quickly estimating battery capacity according to an embodiment of the present invention (I). Figure 1 As shown, a quick estimation of battery capacity includes:

[0083] Step S102, obtaining first current-time data of the battery under test during a constant voltage charging phase and second current-time data of the battery under test during a rest phase after constant voltage charging;

[0084] Step S104: input the first current-time data and the second current-time data into a pre-trained battery capacity estimation model to obtain the battery capacity.

[0085] In the present invention, batteries include but are not limited to sodium ion batteries, lithium ion batteries, lead-acid batteries, lithium-sulfur batteries, etc.

[0086] In the present invention, the application scenarios of the present invention include but are not limited to battery capacity division before leaving the factory, capacity estimation during operation, and capacity division during retirement.

[0087] In step S102 , the constant voltage charging stage is a short constant voltage charging stage after the constant voltage discharging stage of the battery.

[0088] In step S102 , the first current-time data includes a plurality of first current data and a plurality of first time data, wherein the plurality of first current data corresponds one-to-one to the plurality of first time data.

[0089] Furthermore, the first current-time data also includes a plurality of first fitting current data and a plurality of first fitting time data of a fitting curve formed based on the plurality of first current data and the plurality of first time data.

[0090] Furthermore, the first current-time data also includes a plurality of first coefficients of a fitting curve formed based on the plurality of first current data and the plurality of first time data.

[0091] In some embodiments, the plurality of first time data may be at equal time intervals or at unequal time intervals.

[0092] In some embodiments, the plurality of first fitting time data are spaced at equal intervals.

[0093] As can be seen from the above, the first current-time data can be a number of point values ​​(ie, a number of first current data, a number of first time data), or a curve (ie, a fitting curve obtained by fitting a number of first current data, a number of first time data).

[0094] In step S102 , the second current-time data includes a plurality of second current data and a plurality of second time data, wherein the plurality of second current data corresponds to the plurality of second time data in a one-to-one manner.

[0095] Furthermore, the second current-time data also includes a plurality of second fitting current data and a plurality of second fitting time data of a fitting curve formed based on the plurality of second current data and the plurality of second time data.

[0096] Furthermore, the second current-time data also includes a plurality of second coefficients of a fitting curve formed based on the plurality of second current data and the plurality of second time data.

[0097] In some embodiments, the plurality of second time data may be at equal time intervals or at unequal time intervals.

[0098] In some embodiments, the plurality of second fitting time data are spaced at equal intervals.

[0099] As can be seen from the above, the second current-time data can be a number of point values ​​(ie, a number of second current data, a number of second time data), or a curve (ie, a fitting curve obtained by fitting a number of second current data, a number of second time data).

[0100] In step S104 , the first current-time data and the second current-time data are input into a battery capacity estimation model to obtain the battery capacity.

[0101] In step S104 , the battery capacity estimation model includes but is not limited to a random forest regression model, a Gaussian process regression model, a support vector machine model, a neural network model, and the like.

[0102] In step S104, the time for obtaining the battery capacity is generally within 30 minutes. Preferably, the time for obtaining the battery capacity is within 20 minutes.

[0103] Through steps S102 to S104, the battery capacity division time is significantly shortened and the battery production efficiency is improved; the traditional full charge and discharge capacity test process is shortened from 4 hours to less than 30 minutes, which greatly improves the detection efficiency, significantly reduces the power consumed by full charge and discharge, and greatly reduces energy consumption; the detection accuracy is high and the detection error is small, with the maximum detection error less than 1.5% and the average detection error only 0.42%.

[0104] Figure 2 FIG2 is a flowchart of a method for quickly estimating battery capacity according to an embodiment of the present invention (II). Figure 2 As shown, obtaining the first current-time data of the constant voltage charging stage of the battery to be tested includes:

[0105] Step S202: obtaining a plurality of first time data and corresponding first current data during a constant voltage charging phase of the battery under test;

[0106] Step S204: construct a first current-time curve with time as the horizontal axis and current as the vertical axis;

[0107] Step S206: fitting the first current-time curve using a first polynomial function to obtain a first current-time fitting curve and a plurality of first polynomial coefficients;

[0108] Step S208 : forming first current-time data using the first current-time fitting curve and a plurality of first polynomial coefficients.

[0109] In step S202 , the plurality of first time data may be at equal time intervals or at unequal time intervals.

[0110] In step S204 , the first current-time curve is a point distribution graph.

[0111] In step S206 , the first polynomial function is at least a third-order polynomial.

[0112] In some embodiments, the first polynomial function is a fifth-order polynomial, and the first polynomial coefficients are -3.94403143e-06, 1.57437867e-04, -2.63657145e-03, 2.53981014e-02, -1.62843435e-01, and 6.33374973e-01, respectively.

[0113] In some embodiments, the first polynomial function is a fifth-order polynomial, and the first polynomial coefficients are -4.36179037e-07, 2.85444822e-05, -8.29644917e-04, 1.54047465e-02, -2.06722228e-01, and 1.77794943e+00, respectively.

[0114] In step S208 , only the first polynomial coefficients may be used to form the first current-time data. Specifically, only the first polynomial coefficients may be used as input values ​​to input into the battery capacity estimation model.

[0115] Through steps S202 to S208 , more accurate characteristic values ​​can be obtained as input values, thereby improving the output accuracy of the battery capacity estimation model.

[0116] Figure 3 FIG3 is a flowchart of a method for quickly estimating battery capacity according to an embodiment of the present invention. Figure 3 As shown, obtaining the second current-time data of the battery under test in the static stage after constant voltage charging includes:

[0117] Step S302: obtaining a plurality of second time data and corresponding second current data of the battery under test during a rest phase after constant voltage charging;

[0118] Step S304: construct a second current-time curve with time as the horizontal axis and current as the vertical axis;

[0119] Step S306: fitting the second current-time curve using a second polynomial function to obtain a second current-time fitting curve and a plurality of second polynomial coefficients;

[0120] Step S308 : forming second current-time data using the second current-time fitting curve and a plurality of second polynomial coefficients.

[0121] In step S302 , the plurality of second time data may be at equal time intervals or at unequal time intervals.

[0122] In step S304 , the second current-time curve is a point distribution graph.

[0123] In step S306 , the second polynomial function is at least a third-order polynomial.

[0124] In some embodiments, the second polynomial function is a fifth-order polynomial, and the second polynomial coefficients are -1.67312169e-05, 2.42717006e-04, -1.33710141e-03, 3.59136982e-03, -6.03967437e-03, and 2.55575012e+00, respectively.

[0125] In some embodiments, the second polynomial function is a fifth-order polynomial, and the second polynomial coefficients are -9.22933393e-06, 1.39798220e-04, -8.28311961e-04, 2.48732516e-03, -4.36348653e-03, and 3.64583619e+00, respectively.

[0126] In step S208 , only the second polynomial coefficients may be used to form the second current-time data. Specifically, only the second polynomial coefficients may be used as input values ​​to input into the battery capacity estimation model.

[0127] Through steps S302 to S308 , more accurate characteristic values ​​can be obtained as input values, thereby improving the output accuracy of the battery capacity estimation model.

[0128] Figure 4 FIG4 is a flowchart of a method for quickly estimating battery capacity according to an embodiment of the present invention. Figure 4 As shown, the training method of the battery capacity estimation model includes:

[0129] Step S402: Obtain historical capacity data of several batteries;

[0130] Step S404: Obtain battery capacity data, first current-time data during the constant-voltage charging phase, and second current-time data during the rest phase after constant-voltage charging based on the historical capacity-dividing data;

[0131] Step S406 : Input the first current-time data and the second current-time data as input values ​​and the battery capacity data as output values ​​into a battery capacity estimation model to train the battery capacity estimation model, thereby obtaining a trained battery capacity estimation model.

[0132] In step S402 , the historical capacity division data is generally battery capacity division data obtained based on full charge and discharge tests.

[0133] In step S402, historical capacity classification data of a number of batteries is used as a training set.

[0134] In step S404 , the battery capacity data is calculated using the ampere-hour integration method.

[0135] In step S404, the first current-time data includes a plurality of first current data and a plurality of first time data, wherein the plurality of first current data corresponds to the plurality of first time data in a one-to-one manner.

[0136] Furthermore, the first current-time data also includes a plurality of first fitting current data and a plurality of first fitting time data of a fitting curve formed based on the plurality of first current data and the plurality of first time data.

[0137] Furthermore, the first current-time data also includes a plurality of first coefficients of a fitting curve formed based on the plurality of first current data and the plurality of first time data.

[0138] In some embodiments, the plurality of first time data may be at equal time intervals or at unequal time intervals.

[0139] In some embodiments, the plurality of first fitting time data are spaced at equal intervals.

[0140] As can be seen from the above, the first current-time data can be a number of point values ​​(ie, a number of first current data, a number of first time data), or a curve (ie, a fitting curve obtained by fitting a number of first current data, a number of first time data).

[0141] In step S404, the second current-time data includes a plurality of second current data and a plurality of second time data, wherein the plurality of second current data corresponds to the plurality of second time data in a one-to-one manner.

[0142] Furthermore, the second current-time data also includes a plurality of second fitting current data and a plurality of second fitting time data of a fitting curve formed based on the plurality of second current data and the plurality of second time data.

[0143] Furthermore, the second current-time data also includes a plurality of second coefficients of a fitting curve formed based on the plurality of second current data and the plurality of second time data.

[0144] In some embodiments, the plurality of second time data may be at equal time intervals or at unequal time intervals.

[0145] In some embodiments, the plurality of second fitting time data are spaced at equal intervals.

[0146] As can be seen from the above, the second current-time data can be a number of point values ​​(ie, a number of second current data, a number of second time data), or a curve (ie, a fitting curve obtained by fitting a number of second current data, a number of second time data).

[0147] In step S406 , the battery capacity estimation model includes but is not limited to a random forest regression model, a Gaussian process regression model, a support vector machine model, a neural network model, and the like.

[0148] Through steps S402 to S406, a battery capacity estimation model can be trained to perform rapid and accurate detection of battery capacity, greatly improving detection efficiency, significantly reducing the power consumed by full charging and discharging, and significantly reducing energy consumption.

[0149] Figure 5 FIG5 is a flowchart of a method for quickly estimating battery capacity according to an embodiment of the present invention (V). Figure 5 As shown, the training method of the battery capacity estimation model also includes:

[0150] Step S502: Obtain historical capacity data of several batteries;

[0151] Step S504: obtaining first current-time data of the constant voltage charging phase and second current-time data of the rest phase after the constant voltage charging based on the historical capacity distribution data;

[0152] Step S506 : Using the first current-time data and the second current-time data as input values, and inputting them into the trained battery capacity estimation model to test the battery capacity estimation model.

[0153] Among them, steps S502 to S504 are basically the same as steps S402 to S404 and are not repeated here.

[0154] In step S502, historical capacity data of a number of batteries are used as a test set.

[0155] In step S506 , the first current-time data and the second current-time data are input as input values ​​to the trained battery capacity estimation model to obtain the battery capacity.

[0156] Furthermore, after step S506, the method further includes:

[0157] Step S508: Compare the battery capacity with the battery capacity obtained based on historical capacity data;

[0158] Step S510: Determine whether the battery capacity estimation model has been trained based on the comparison result.

[0159] Through steps S502 to S510, the battery capacity estimation model is tested to determine the training result.

[0160] Figure 6 FIG6 is a flowchart of a method for quickly estimating battery capacity according to an embodiment of the present invention. Figure 6 As shown, the training method of the battery capacity estimation model also includes:

[0161] Step S602: Obtain cross-validation results;

[0162] Step S604: Adjust the parameters of the battery capacity estimation model according to the cross-validation result.

[0163] In step S602, the cross-validation result is obtained from the training set of steps S402 to S406.

[0164] In step S604, hyperparameters such as “n_estimators”, “max_depth”, and “min_samples_split” are adjusted.

[0165] Through steps S602 to S604, the model parameters are adjusted using the cross-validation results, thereby improving the detection accuracy of the battery capacity estimation model and reducing the detection error.

[0166] In addition, the battery capacity rapid estimation method of the embodiment of the present application can be implemented by a computer device. The components of the computer device may include but are not limited to a processor and a memory storing computer program instructions.

[0167] In some embodiments, the processor may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0168] In some embodiments, the memory may include a large-capacity storage for data or instructions. By way of example, and not limitation, the memory may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory may include removable or non-removable (or fixed) media. Where appropriate, the memory may be internal or external to the data processing device. In certain embodiments, the memory is non-volatile memory. In certain embodiments, the memory includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically alterable ROM (EAROM) or a flash memory (FLASH), or a combination of two or more of these. Under appropriate circumstances, the RAM can be a static random access memory (SRAM) or a dynamic random access memory (DRAM), where the DRAM can be a fast page mode dynamic random access memory (FPMDRAM), an extended data out dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0169] The memory may be used to store or cache various data files required for processing and / or communication, as well as possible computer program instructions executed by the processor.

[0170] The processor implements any one of the battery capacity rapid estimation methods in the above embodiments by reading and executing computer program instructions stored in the memory.

[0171] In some embodiments, the computer device may further include a communication interface and a bus, wherein the processor, the memory, and the communication interface are connected via the bus and communicate with each other.

[0172] The communication interface is used to enable communication between the various units, devices, units, and / or devices in the embodiments of the present application. The communication interface can also enable data communication with other components such as external devices, image / data acquisition equipment, databases, external storage, and image / data processing workstations.

[0173] A bus, which includes hardware, software, or both, connects components of a computer device. It includes, but is not limited to, at least one of the following: a data bus, an address bus, a control bus, an expansion bus, and a local bus. By way of example and not limitation, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of the above. Where appropriate, a bus may include one or more buses. Although embodiments herein describe and illustrate a particular bus, this application contemplates any suitable bus or interconnect.

[0174] The computer device can execute the battery capacity rapid estimation method in the embodiment of the present application.

[0175] In addition, in conjunction with the battery capacity rapid estimation method in the above embodiments, embodiments of the present application may provide a computer-readable storage medium for implementation. The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any of the battery capacity rapid estimation methods in the above embodiments is implemented.

[0176] Example 2

[0177] This embodiment relates to the device for quickly estimating battery capacity of the present invention.

[0178] Figure 7 FIG is a framework diagram of a device for quickly estimating battery capacity according to an embodiment of the present invention. Figure 7 As shown, a battery capacity rapid estimation device 700 is used to perform the battery capacity rapid estimation method as described in Example 1, and includes a current-time data acquisition module 710 and a capacity estimation module 720. The current-time data acquisition module 710 is used to acquire first current-time data during a constant voltage charging phase and second current-time data during a rest phase after constant voltage charging of a battery under test; the capacity estimation module 720 is used to input the first current-time data and the second current-time data into a pre-trained battery capacity estimation model to obtain the battery capacity.

[0179] The current-time data acquisition module 710 and the capacity estimation module 720 are used to execute steps S102 to S104 of the first embodiment.

[0180] The current-time data acquisition module 710 is further configured to execute steps S202 to S208 and steps S302 to S308 of the first embodiment.

[0181] Furthermore, the battery capacity rapid estimation device 700 also includes a training data acquisition module 730 and a model training module 740. The training data acquisition module 730 is configured to acquire historical capacity data of a plurality of batteries, and obtain battery capacity data, first current-time data during a constant voltage charging phase, and second current-time data during a rest phase after constant voltage charging based on the historical capacity data. The model training module 740 is configured to input the first current-time data and the second current-time data as input values ​​and the battery capacity data as output values ​​into a battery capacity estimation model to train the battery capacity estimation model and thereby obtain a trained battery capacity estimation model.

[0182] The training data acquisition module 730 and the model training module 740 are used to execute steps S402 to S406 of Example 1.

[0183] Furthermore, the battery capacity rapid estimation device 700 also includes a test data acquisition module 750 and a model testing module 760. The test data acquisition module 750 is configured to acquire historical capacity data of a plurality of batteries and, based on the historical capacity data, obtain first current-time data during a constant voltage charging phase and second current-time data during a rest phase after constant voltage charging. The model testing module 760 is configured to input the first current-time data and the second current-time data as input values ​​into a trained battery capacity estimation model to test the battery capacity estimation model.

[0184] The test data acquisition module 750 and the model testing module 760 are used to execute steps S502 to S506 of the first embodiment.

[0185] Furthermore, the battery capacity rapid estimation device 700 further includes a parameter optimization module 770. The parameter optimization module 770 is used to obtain cross-validation results and adjust the parameters of the battery capacity estimation model according to the cross-validation results.

[0186] The parameter optimization module 770 is used to execute steps S602 to S604 of the first embodiment.

[0187] The technical effects of this embodiment are basically the same as those of embodiment 1 and will not be described again here.

[0188] Example 3

[0189] This embodiment is a specific implementation of the present invention.

[0190] A method for quickly estimating battery capacity, comprising:

[0191] Step 1: Collect historical data of battery charging and discharging and complete data preprocessing

[0192] Step 1.1: Collect historical data from the battery capacity grading production line, generally based on full charge and discharge battery capacity grading data, and record the voltage, current, and time data at each moment of the battery's charge and discharge process;

[0193] Step 1.2: Based on the voltage and current data of the complete discharge process of the battery, the capacity of each battery is calculated using the ampere-hour integration method as the actual label value of the capacity;

[0194] Step 1.3: Remove battery data with abnormal voltage, current, and capacity, complete data preprocessing, and save the data.

[0195] Step 2: Extract battery health features based on the battery constant voltage charging and static phase to form a training data set

[0196] Step 2.1. Extract the current and time variation data of the battery during the constant voltage charging phase, fit the current-time curve using a polynomial function, select the appropriate polynomial order based on experience, and record all coefficients of the fitted polynomial equation;

[0197] Step 2.2: Extract the voltage-time variation data during the rest phase after constant-voltage charging of the battery, fit the voltage-time curve using a polynomial function, select an appropriate polynomial order based on experience, and record all coefficients of the fitted polynomial equation;

[0198] In step 2.3, the polynomial fitting coefficients of steps 2.1 and 2.2 are used as battery health features and integrated with the capacity labels of step 1.3 to form a training data set.

[0199] Step 3: Select and train a machine learning model based on the training dataset

[0200] Step 3.1. Based on the size of the training dataset, select several appropriate machine learning models, such as random forest regression, Gaussian process regression, support vector machine, neural network, etc.

[0201] Step 3.2: Based on the training data set in step 2.3, use the polynomial coefficient feature as input and the capacity label as output, and use the model in (1) to train on the training data set. The model with the smallest cross-validation error is selected as the final model.

[0202] Step 3.3: After the final model training is completed, save the model for online estimation.

[0203] Step 4: Perform a short constant voltage charge test on the battery to be tested, extract features, and input the model to complete the rapid capacity estimation.

[0204] Step 4.1, perform a short constant current and constant voltage charging test on the device to be tested, and record the test data during the constant voltage stage and the subsequent static stage;

[0205] Step 4.2: Based on the test data, use the method in step 2 to extract the corresponding polynomial fitting coefficients as features;

[0206] Step 4.3: The features are directly input into the model in step 3.3, and the model outputs a quick estimation result of the battery capacity.

[0207] The technical effects of this embodiment are as follows:

[0208] 1) Significantly shorten battery capacity testing time: The traditional full charge and discharge capacity testing process is shortened from 4 hours to approximately 17 minutes, greatly improving battery production efficiency;

[0209] 2) Significantly reduce energy consumption: significantly reduce the electrical energy consumed by full charging and discharging, and greatly reduce energy consumption;

[0210] 3) High capacity estimation accuracy: the maximum detection error is less than 1.5%, and the average detection error is only 0.42%

[0211] Example 4

[0212] This embodiment is a specific implementation of the present invention.

[0213] In this embodiment, a cylindrical 18650 sodium ion battery is taken as an example for description.

[0214] A method for quickly estimating the capacity of a sodium ion battery, comprising:

[0215] Step 1: Select a batch of historical capacity data of sodium ion batteries. The voltage change over time during the historical capacity change process of a typical battery is as follows: Figure 8 As shown. In industry, the ampere-hour integration method is used to calculate the true capacity of a battery based on the "constant current discharge phase." The voltage, current, and time data for each battery at each moment of the charge and discharge process, along with the corresponding true capacity data, are recorded. Data from batteries with abnormal voltage, current, and capacity are removed, and the data is preprocessed and saved.

[0216] Step 2: Extract the current-time test data of the short constant voltage charging phase of each battery, such as Figure 9 As shown. The current-time curve is fitted with a polynomial function. Here, a fifth-order polynomial is selected, and the obtained coefficients are [-3.94403143e-06, 1.57437867e-04, -2.63657145e-03, 2.53981014e-02, -1.62843435e-01, 6.33374973e-01], which are recorded. Similarly, the voltage and time variation data of each battery in the static stage after the constant voltage charging is completed are extracted, as shown in the figure below. Figure 10 As shown in the figure, a polynomial function is used to fit the voltage-time curve. Here, a fifth-order polynomial is selected. The resulting coefficients are [-1.67312169e-05, 2.42717006e-04, -1.33710141e-03, 3.59136982e-03, -6.03967437e-03, 2.55575012e+00] and recorded. Repeat the above steps for all batteries and record the resulting polynomial coefficients as features. Combine the features of each battery and its corresponding battery capacity to form a training dataset, which is then recorded and saved.

[0217] Step 3: Based on the training dataset, select several appropriate machine learning models, such as linear regression, random forest regression, Gaussian process regression, support vector regression, and neural network models for training. Here, the random forest model is selected as the final model. Model hyperparameters and other parameters are adjusted based on the training set. Once the hyperparameters are determined, model training is complete. Save the model locally for easy reference during online capacity estimation.

[0218] Step 4: Select another batch of sodium-ion batteries as test subjects and use the feature extraction method in Step 2 to obtain polynomial coefficients as input features. These features are then fed into the random forest model trained in Step 3 to obtain rapid capacity estimation results for each battery.

[0219] The capacity quick estimation result is as follows Figure 11 As shown, it can be seen that the estimated capacity value and the reference value (battery capacity obtained by fully charging and discharging the battery) are very close.

[0220] The error of the rapid capacity estimation is as follows Figure 12 As shown, the average absolute percentage error is 0.42% and the maximum percentage error is 1.33%.

[0221] From the above results, we can see that:

[0222] 1) The technology of the present invention uses the fragmented data of short-term constant voltage charging to achieve high-precision battery capacity estimation.

[0223] 2) If Figure 9 、 Figure 10 As shown, the traditional capacity testing method based on full charge and discharge takes an average of more than 4 hours for a single battery cell, while the technology of the present invention only requires about 17 minutes of detection time. This shows that while maintaining high-precision capacity estimation, this technology also greatly shortens the test time, has great industrial application value, and can significantly reduce the time, energy consumption and related costs of the battery cell capacity separation process.

[0224] Example 5

[0225] This embodiment is a specific implementation of the present invention.

[0226] In this embodiment, a cylindrical 18650 ternary lithium-ion battery is taken as an example for description.

[0227] A method for quickly estimating battery capacity, comprising:

[0228] Step 1: Select a batch of historical capacity data for lithium-ion batteries. The voltage change over time during the historical capacity change process of a typical battery is as follows: Figure 13 As shown. In industry, the ampere-hour integration method is used to calculate the true capacity of a battery based on the "constant current discharge phase." The voltage, current, and time data for each battery at each moment of the charge and discharge process, along with the corresponding true capacity data, are recorded. Data from batteries with abnormal voltage, current, and capacity are removed, and the data is preprocessed and saved.

[0229] Step 2: Extract the current-time test data of the short constant voltage charging phase of each battery, such as Figure 14As shown. The current-time curve is fitted using a polynomial function. Here, a fifth-order polynomial is selected, and the obtained coefficients are [-4.36179037e-07, 2.85444822e-05, -8.29644917e-04, 1.54047465e-02, -2.06722228e-01, 1.77794943e+00], which are recorded. Similarly, the voltage and time variation data of each battery in the static stage after the constant voltage charging is completed are extracted, as shown in the figure. Figure 15 As shown in the figure, a polynomial function is used to fit the voltage-time curve. Here, a fifth-order polynomial is selected. The resulting coefficients are [-9.22933393e-06, 1.39798220e-04, -8.28311961e-04, 2.48732516e-03, -4.36348653e-03, 3.64583619e+00] and recorded. Repeat the above steps for all batteries and record the resulting polynomial coefficients as features. Combine the features of each battery and its corresponding battery capacity to form a training dataset, which is then recorded and saved.

[0230] Step 3: Based on the training dataset, select several appropriate machine learning models, such as linear regression, random forest regression, Gaussian process regression, support vector regression, and neural network models for training. Here, the random forest model is selected as the final model. Model hyperparameters and other parameters are adjusted based on the training set. Once the hyperparameters are determined, model training is complete. Save the model locally for easy reference during online capacity estimation.

[0231] Step 4: Select another batch of lithium-ion batteries as test subjects and use the feature extraction method in Step 2 to obtain polynomial coefficients as input features. These features are then fed into the random forest model trained in Step 3 to obtain a rapid capacity estimation result for each battery.

[0232] The capacity quick estimation result is as follows Figure 16 As shown, it can be seen that the estimated capacity value and the reference value (battery capacity obtained by fully charging and discharging the battery) are very close.

[0233] The error of the rapid capacity estimation is as follows Figure 17 As shown, the average absolute percentage error is 0.56% and the maximum percentage error is 1.70%.

[0234] From the above results, we can see that:

[0235] 1) The technology of the present invention uses the fragmented data of short-term constant voltage charging to achieve high-precision battery capacity estimation.

[0236] 2) If Figure 14 、 Figure 15As shown in the figure, the traditional capacity testing method based on full charge and discharge takes an average of more than 4 hours for a single battery cell, while the technology of the present invention only requires about 25 minutes of detection time. This shows that while maintaining high-precision capacity estimation, this technology also greatly shortens the test time, has great industrial application value, and can significantly reduce the time, energy consumption and related costs of the battery cell capacity separation process.

[0237] The above description is only a preferred embodiment of the present invention and does not limit the implementation mode and protection scope of the present invention. For those skilled in the art, it should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for quickly estimating battery capacity, characterized in that: include: Step 1: Collect historical data of battery charging and discharging and complete data preprocessing Step 1.1: Collect historical data from the battery capacity grading production line. This includes recording the voltage, current, and time data at each moment of the battery's charge and discharge process based on the battery capacity grading data of full charge and discharge cycles. Step 1.2: Based on the voltage and current data of the complete discharge process of the battery, the capacity of each battery is calculated using the ampere-hour integration method as the actual label value of the capacity; Step 1.3: Remove battery data with abnormal voltage, current or capacity, complete data preprocessing, and save the data; Step 2: Extract battery health features based on the battery constant voltage charging and static phase to form a training data set Step 2.

1. Extract the current and time variation data of the battery during the constant voltage charging phase, fit the current-time curve using a polynomial function, select the appropriate polynomial order based on experience, and record all coefficients of the fitted polynomial equation; Step 2.2, extract the voltage and time change data of the static stage after the constant voltage charging of the battery, use a polynomial function to fit the voltage-time curve, select an appropriate polynomial order, and record all coefficients of the fitted polynomial equation; Step 2.3: The polynomial fitting coefficients from steps 2.1 and 2.2 are used as battery health features and integrated with the capacity labels from step 1.3 to form a training dataset. Step 3: Select and train a machine learning model based on the training dataset Step 3.

1. Select an appropriate machine learning model based on the size of the training dataset. Machine learning models include random forest regression model, Gaussian process regression model, support vector machine model, and neural network model. Step 3.2: Based on the training dataset in step 2.3, use the polynomial coefficient features as input and the capacity label as output. Use the model in step 3.1 to train on the training dataset and select the model with the smallest cross-validation error as the final model. Step 3.3: After the final model training is completed, save the model for online estimation. Step 4: Perform a short constant voltage charge test on the battery to be tested, extract features, and input the model to complete the rapid capacity estimation. Step 4.1, perform a short constant current and constant voltage charging test on the battery under test, and record the test data during the constant voltage stage and the subsequent static stage; Step 4.2: Based on the test data, use the method in step 2 to extract the corresponding polynomial fitting coefficients as features; Step 4.3: The features are directly input into the model in step 3.3, and the model outputs a quick estimation result of the battery capacity.

2. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for quickly estimating battery capacity according to claim 1 is implemented.

3. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for quickly estimating battery capacity as claimed in claim 1 is implemented.

Images