An online estimation method and device for battery capacity

Abstract

The application provides an online estimation method and device for battery capacity, comprising: constructing an RC equivalent circuit model of a battery based on acquired online terminal voltage and current of the battery, and processing the RC equivalent circuit model to obtain a to-be-estimated parameter; setting a first set threshold, and processing the open circuit voltage based on the first set threshold to obtain an open circuit voltage first estimation value sequence; setting a second set threshold, and processing the open circuit voltage based on the second set threshold to obtain an open circuit voltage second estimation value sequence; determining a first SOC estimation value sequence corresponding to the open circuit voltage first estimation value sequence and a second SOC estimation value sequence corresponding to the open circuit voltage second estimation value sequence based on a pre-calibrated SOC-OCV relationship; and processing the first SOC estimation value sequence and the second SOC estimation value sequence according to a linear relationship between battery capacity change and SOC change to obtain the battery capacity. The application can more accurately estimate the battery capacity.

Description

Technical Field

[0001] This invention relates to battery technology, and in particular to an online method and apparatus for estimating battery capacity. Background Technology

[0002] With energy shortages and environmental pollution becoming a global concern, lithium batteries have become the primary choice for new energy vehicles, mobile communications, and smart grids due to their high energy density and long cycle life. However, with frequent use, batteries experience capacity decay and increased internal resistance, leading to a series of safety issues. Therefore, accurately estimating battery capacity is crucial for improving battery safety.

[0003] Currently, battery capacity estimation methods are mainly divided into three categories: direct detection method, model identification method, and data-driven method.

[0004] Direct testing methods mainly involve offline capacity calibration, which is time-consuming and labor-intensive.

[0005] Model identification methods primarily estimate capacity by identifying the parameters of the model, but parameter mismatch can occur during the estimation process.

[0006] Data-driven methods typically use voltage, current, and temperature during the charging process as inputs to a neural network to estimate battery capacity. However, this method requires a large amount of data, has severe time-series dependencies, and is difficult to model.

[0007] Therefore, it is urgent to solve the problem of inaccurate battery capacity estimation in existing technologies under offline conditions. Summary of the Invention

[0008] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide an online battery capacity estimation method and apparatus to solve the problem of inaccurate battery capacity estimation in the offline state in the prior art.

[0009] To achieve the above and other related objectives, the present invention provides an online method for estimating battery capacity, comprising at least the following steps:

[0010] An RC equivalent circuit model of the battery is constructed based on the obtained online terminal voltage and current of the battery, and the RC equivalent circuit model is processed to obtain the parameters to be estimated; the parameters to be estimated include the open circuit voltage.

[0011] A first preset threshold is set, and the open-circuit voltage is processed based on the first preset threshold to obtain a first estimated value sequence of open-circuit voltage;

[0012] A second preset threshold is set, and the open-circuit voltage is processed based on the second preset threshold to obtain a second estimated value sequence of open-circuit voltage;

[0013] Based on the pre-calibrated SOC-OCV relationship, a first SOC estimation sequence corresponding to the first open-circuit voltage estimation sequence and a second SOC estimation sequence corresponding to the second open-circuit voltage estimation sequence are determined.

[0014] Based on the linear relationship between changes in battery charge and changes in State of Charge (SOC), the battery capacity is obtained by processing the first SOC estimation sequence and the second SOC estimation sequence.

[0015] Preferably, the process of processing the RC equivalent circuit model to obtain the parameters to be estimated includes:

[0016] The transfer function is obtained by processing the RC equivalent circuit model.

[0017] The transfer function is identified online using the least squares method with a forgetting factor to estimate the battery parameters.

[0018] Preferably, the RC equivalent circuit model is as follows:

[0019]

[0020] Where U1 is the polarization voltage, U t R is the terminal voltage, I is the current, OCV is the open-circuit voltage, R1 is the polarization resistance, R0 is the ohmic resistance, and τ is the time constant.

[0021] Preferably, if a first set condition is met within a set sampling length, then a first estimated value sequence of open-circuit voltages satisfying the first set condition is obtained from the open-circuit voltages.

[0022] The first condition is:

[0023]

[0024] Where ΔT is the sampling time interval, R1 is the polarization internal resistance, C1 is the polarization capacitance, and n is the first set threshold.

[0025] Preferably, the step of processing the open-circuit voltage based on a second set threshold to obtain a second estimated sequence of open-circuit voltage values ​​includes:

[0026] A second set condition is determined based on a second set threshold; and at least two sets of data segments satisfying the second set condition are identified from the open-circuit voltage; the second set condition is:

[0027]

[0028] Where m is the second preset threshold;

[0029] Each data segment is processed to obtain the second estimated value of the open-circuit voltage corresponding to each data segment;

[0030] A sequence of second estimates of open-circuit voltage is obtained based on multiple sets of second estimates of open-circuit voltage.

[0031] Preferably, processing each data segment to obtain the second estimated value group of open-circuit voltage corresponding to each data segment includes:

[0032] The two data points of each group are respectively used as the second initial estimate of the open circuit voltage corresponding to the starting point of each group and the second final estimate of the open circuit voltage corresponding to the ending point of each group;

[0033] Linear interpolation is performed based on the second initial estimate of the open-circuit voltage and the second final estimate of the open-circuit voltage for each data segment to obtain the second estimate of the open-circuit voltage for each data segment.

[0034] Preferably, the step of processing the first SOC estimation sequence and the second SOC estimation sequence to obtain the battery capacity based on the linear relationship between battery charge change and SOC change includes:

[0035] Based on the linear relationship between battery charge change and SOC change, the first SOC estimation value sequence is processed to obtain the first capacity value, and the second SOC estimation value sequence is processed to obtain the second capacity value.

[0036] The battery capacity is obtained by weighting the first capacity value and the second capacity value.

[0037] Preferably, the linear relationship between battery charge change and SOC change is the SOC-charge gain method; the SOC-charge gain method calculates battery capacity by the ratio of charge / discharge charge change to the corresponding SOC change.

[0038] To achieve the above and other related objectives, the present invention also provides an online battery capacity estimation device, including a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the online battery capacity estimation method described above.

[0039] As described above, the online battery capacity estimation method and apparatus of the present invention have the following beneficial effects:

[0040] The online battery capacity estimation method and apparatus proposed in this invention achieve online battery capacity estimation by constructing an RC equivalent circuit model based on the acquired online battery electrical parameters and by weighting the processing results of the parameters to be estimated under different conditions. This significantly improves the accuracy of online battery capacity estimation and is more adaptable to the actual operating conditions of battery use. Attached Figure Description

[0041] Figure 1 The diagram shows a flowchart of the online battery capacity estimation method of the present invention.

[0042] Figure 2 The diagram shown is a structural schematic of the online battery capacity estimation device of the present invention. Detailed Implementation

[0043] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0044] Please see Figure 1-2 It should be noted that the illustrations provided in this embodiment are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0045] This invention proposes an online battery capacity estimation method that combines model identification and data-driven methods to estimate battery capacity online. This method can significantly improve the accuracy of online capacity estimation, is highly adaptable to actual working conditions, and has certain practical applications.

[0046] Based on the above technical concept, the present invention proposes an online estimation method and apparatus for battery capacity.

[0047] Method Implementation Examples:

[0048] like Figure 1 The diagram shown is a schematic flowchart of the online battery capacity estimation method of the present invention. Figure 1 The online battery capacity estimation method of the present invention is described in detail below:

[0049] S1. Based on the obtained online terminal voltage and current of the battery, construct the RC equivalent circuit model of the battery, and process the RC equivalent circuit model to obtain the parameters to be estimated.

[0050] In this embodiment of the invention, the constructed RC equivalent circuit model is a first-order RC model.

[0051]

[0052] Where U1 is the polarization voltage, U tR is the terminal voltage, I is the current, OCV is the open-circuit voltage, R1 is the polarization resistance, R0 is the ohmic resistance, and τ is the time constant.

[0053] The process of processing the RC equivalent circuit model to obtain the parameters to be estimated includes:

[0054] S11, The transfer function is obtained by processing the RC equivalent circuit model;

[0055] The transfer function is obtained by processing the RC equivalent circuit model based on the bilinear transform in frequency domain analysis.

[0056] First, the transfer function of the RC series circuit in the frequency domain is transformed by a bilinear transformation to obtain the transfer function model;

[0057]

[0058] Where ΔT is the sampling time interval and C1 is the polarization capacitor.

[0059] Then, by combining the first-order RC equivalent circuit model, the transfer function model is transformed to obtain the transfer function as follows:

[0060]

[0061] make:

[0062] The simplified transfer function is:

[0063]

[0064] S12, the transfer function is identified online using the least squares method with a forgetting factor to estimate the battery parameters.

[0065] The parameters to be estimated include open-circuit voltage, ohmic internal resistance, polarization internal resistance, and time constant;

[0066] Discretizing the transfer function yields:

[0067] U t,k =bU t,k-1 +(1-b)OCV+cI k +dI k-1

[0068] Among them, U t,k Let U be the terminal voltage at time k in the sampling time t. t,k-1 Let I be the terminal voltage at time k-1 in the sampling time t. k Let I be the current at time k. k-1 Let be the current at time k-1.

[0069] In a least squares system with a forgetting factor (FFRLS), the vector expressions for the observation vector, output vector, and vector to be estimated are as follows:

[0070] y k =U t,k

[0071] φ k =[1,U t,k-1 ,I k ,I k-1 ] T

[0072] θ k =[(1-b)OCV k [b,c,d]

[0073] Among them, y k For the output vector, φ k Let θ be the observation vector. k Let be the vector to be estimated.

[0074] The open-circuit voltage, ohmic internal resistance R0, polarization internal resistance R1 and time constant τ1 are identified online by calculation using least squares with forgetting factor (FFRLS).

[0075] Wherein, the ohmic resistance R0, the polarization resistance R1, and the time constant τ1 are respectively:

[0076]

[0077] S2, set a first preset threshold, and process the open circuit voltage based on the first preset threshold to obtain a first estimated value sequence of open circuit voltage;

[0078] The first set threshold of the present invention is n, and the set sampling length (sampling time is greater than or equal to t seconds) is set. If the first set condition is met within the set sampling length, the first estimated value sequence of open circuit voltage that meets the first set condition is solved in the open circuit voltage. The first estimated value sequence is the first estimated value of open circuit voltage corresponding to each sampling point within the set sampling length.

[0079] The first condition is: b ≤ n, that is,

[0080]

[0081] In this embodiment of the invention, the first set threshold is n. The first set threshold is set according to the battery cell and the actual operating conditions. That is to say, the first set threshold can be different for different models of battery cells and operating conditions.

[0082] This invention takes a 174Ah ternary lithium battery as an example. Under NEDC conditions, n = 0.96, and the time length is set to 3000s. When the value of n is less than or equal to 0.96 within the time duration, and the time duration is greater than or equal to 3000s, the first estimated value of the open-circuit voltage at each sampling point within that time duration is calculated, which is OCV. 1,1 OCV 1,2 OCV 1,3 OCV 1,4 ...

[0083] S3, set a second preset threshold, and process the open circuit voltage based on the second preset threshold to obtain a second estimated value sequence of open circuit voltage;

[0084] In this invention, processing the open-circuit voltage based on a second set threshold to obtain a second estimated sequence of open-circuit voltage values ​​includes:

[0085] S31, determine the second setting condition based on the second set threshold; and determine at least two sets of data segments in the open circuit voltage that satisfy the second setting condition;

[0086] The second condition is: |1-b|≤m, that is:

[0087]

[0088] The second set threshold of this invention is m. This second set threshold is set according to the battery cell and actual operating conditions; that is, the second set threshold can be different for different cell models and operating conditions. This step achieves the identification of multiple sets of data points that meet the second set condition within the open-circuit voltage.

[0089] In this embodiment of the invention, taking a 174Ah ternary battery as an example, under NEDC conditions, setting m=0.002, we find sampling points that satisfy |1-b|≤m, thereby obtaining multiple sets of data segments.

[0090] S32, process each data segment to obtain the second estimated value group of open circuit voltage corresponding to each data segment;

[0091] In this invention, processing each data segment to obtain a second estimated value group of open-circuit voltage corresponding to each data segment includes:

[0092] S321, take the two data points of each group as the second initial estimate of the open circuit voltage corresponding to the starting point of each group and the second final estimate of the open circuit voltage corresponding to the ending point of each group, respectively.

[0093] In this invention, the earliest time in each data segment is taken as the starting point, and the corresponding data is the second initial estimate of the open circuit voltage for that group. The latest time in each data segment is taken as the ending point, and the corresponding data is the second final estimate of the open circuit voltage.

[0094] S322, perform linear interpolation based on the second initial estimate of the open-circuit voltage and the second final estimate of the open-circuit voltage for each data segment to obtain the second estimate of the open-circuit voltage for each data segment.

[0095] This step aims to perform linear interpolation on each data segment that meets the second set conditions to obtain multiple intermediate second estimates of open-circuit voltage, and to obtain a set of second estimates of open-circuit voltage for each data segment based on the second initial estimate of open-circuit voltage, the second final estimate of open-circuit voltage, and multiple intermediate estimates of open-circuit voltage.

[0096] The data points between the second initial estimate and the second final estimate of the open-circuit voltage in each data segment are deleted. Then, the second initial estimate and the second final estimate of the open-circuit voltage are used as the endpoint values ​​for each group, and the data within the endpoint values ​​are interpolated. This step is to prevent the open-circuit voltage OCV from having a large error when the b value is close to 1.

[0097] S33, obtain the sequence of second estimates of open-circuit voltage based on the second estimate group of open-circuit voltage corresponding to all data segments.

[0098] In this embodiment of the invention, multiple sets of second estimates of open-circuit voltage are aggregated according to the sampling time to obtain a sequence of second estimates of open-circuit voltage.

[0099] In this embodiment of the invention, the second estimated value of the open-circuit voltage at each sampling point included in the second estimated value sequence is OCV. 2,1 OCV 2,2 OCV 2,3 OCV 2,4 ...

[0100] S4, based on the pre-calibrated SOC-OCV relationship, determine the first SOC sequence corresponding to the first estimated open-circuit voltage sequence and the second SOC estimated sequence corresponding to the second estimated open-circuit voltage sequence;

[0101] The SOC-OCV relationship of this invention is based on pre-calibrated experiments. The SOC-OCV relationship can be represented by a table or a graph.

[0102] In this embodiment of the invention, the SOC-OCV relationship is represented by an SOC-OCV relationship table for cells calibrated at different temperatures. By looking up the SOC-OCV relationship table, the SOC (State of Charge) corresponding to different open-circuit voltages can be obtained.

[0103] Specifically, in the embodiments of the present invention, in the SOC-OCV relationship table of battery cells calibrated at different temperatures, under NEDC conditions:

[0104] Each open-circuit voltage first estimate in the open-circuit voltage first estimate sequence is compared with the SOC-OCV relationship table to obtain multiple first SOC estimates at the current temperature, thus forming a first SOC estimate sequence.

[0105] Each open-circuit voltage second estimate in the open-circuit voltage second estimate sequence is compared with the SOC-OCV relationship table to obtain multiple second SOC estimates at the current temperature, thus forming a second SOC estimate sequence.

[0106] S5. Based on the linear relationship between battery charge change and SOC change, the first SOC estimation sequence and the second SOC estimation sequence are processed to obtain the battery capacity.

[0107] In this embodiment of the invention, the linear relationship between battery charge change and SOC change is the SOC-charge gain method;

[0108] The State of Charge (SOC) gain method calculates capacity by the ratio of the change in charge / discharge capacity to the corresponding change in State of Charge (SOC). The capacity calculation formula is as follows:

[0109]

[0110] Where C represents the battery capacity.

[0111] In this invention, the step of processing the first SOC estimation sequence and the second SOC estimation sequence to obtain the battery capacity based on the linear relationship between battery charge change and SOC change includes:

[0112] S51, based on the linear relationship between battery charge change and SOC change, process the first SOC estimation value sequence to obtain a first capacity value, and process the second SOC estimation value sequence to obtain a second capacity value;

[0113] In this embodiment of the invention, the step of processing the first SOC estimation value sequence to obtain the first capacity value includes:

[0114] First, convert the capacity calculation formula into a matrix format:

[0115]

[0116] make y i =SOC(t2)-SOC(t1)

[0117] Then, Y = X·H + V

[0118] Where: H = [k1, k2] T Let C = 1 / k1, and V represent random noise, where the coefficients are to be determined.

[0119] Then, the coefficients to be determined in the matrix pattern are obtained by fitting the matrix pattern using the least squares method.

[0120] Specifically, the coefficient H to be determined is obtained through the least squares method, with the aim of further determining the battery capacity C.

[0121] The principle of the least squares algorithm is to determine the parameter vector value H by minimizing the squared residuals S of all observations, where the sum of squared residuals is:

[0122] S = (Y - XH) T (Y-XH)

[0123] Based on the first SOC estimate sequence, the least squares estimate of the coefficient H to be determined is obtained by differentiating the above equation and setting the partial derivative to zero:

[0124] H = (X) T X) -1 X T Y

[0125] The first estimate of the coefficient H to be determined is obtained by the least squares method;

[0126] Finally, the first battery capacity value is obtained based on the relationship between the coefficient to be determined and the battery capacity.

[0127] The relationship between the first estimated value of the coefficient H and the battery capacity is as follows:

[0128] H = [k1, k2] T And C1 = 1 / k1;

[0129] The first battery capacity value is calculated to be C1 = 1 / k1.

[0130] In this embodiment of the invention, the step of processing the second SOC estimation sequence to obtain the second capacity value is the same as the step of processing the first SOC estimation sequence to obtain the first capacity value.

[0131] Assume the coefficients to be determined are H = [k3, k4] TThe capacity value of the second battery, C2 = 1 / k3, is obtained for the coefficient to be determined.

[0132] S52, the first capacity value and the second capacity value are weighted to obtain the battery capacity.

[0133] Since the ranges and precisions of the two OCV sequence values ​​obtained in steps S2 and S3 are different, the weights of the first capacity value and the second capacity value are different when calculating the battery capacity.

[0134] In this embodiment of the invention, the first battery capacity value C1 and the second battery capacity value C2 are obtained based on the least squares method. Since the OCV estimate of the first battery capacity value C1 has a reasonable range of parameter b and high accuracy, C1 is given a weight of 0.6 and C2 is given a weight of 0.4, resulting in the final capacity estimate C = C1 × 0.6 + C2 × 0.4.

[0135] Device Example:

[0136] The present invention also provides an online battery capacity estimation device, such as... Figure 2 As shown, it includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the online battery capacity estimation method described above.

[0137] The detailed steps of the online battery capacity estimation method have been described in detail in the method embodiments and will not be repeated here.

[0138] Beneficial effects:

[0139] Using a first-order RC equivalent circuit as a model, the terminal voltage and current are used as observations to identify the parameters to be estimated (OCV, R0, R1, tao1) online. The model is simple and the computation is small. The online identification of OCV avoids the waste of time caused by long-term shelving and has practical application significance.

[0140] By setting a first threshold and a second threshold for filtering, the estimation accuracy of effective OCV is improved, which indirectly improves the accuracy of obtaining SOC by looking up the table through OCV in the data preprocessing stage, thereby improving the accuracy of capacity estimation.

[0141] By assigning weights to the first capacity value C1 and the second capacity value C2, the accuracy of battery capacity can be further improved on the original basis. Furthermore, it ensures the robustness of capacity estimation in the event that the data for a certain capacity value is too small.

[0142] In summary, this invention proposes an online battery capacity estimation method. This method combines model-driven and data-driven approaches to estimate battery capacity online, significantly improving the accuracy of online capacity estimation and adapting well to real-world operating conditions, thus possessing considerable practicality. Therefore, this invention effectively overcomes the various shortcomings of existing technologies and has high industrial application value.

[0143] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A method for online estimation of battery capacity, characterized in that, It should include at least the following steps: An RC equivalent circuit model of the battery is constructed based on the obtained online terminal voltage and current of the battery, and the RC equivalent circuit model is processed to obtain the parameters to be estimated; the parameters to be estimated include the open circuit voltage. A first preset threshold is set, and the open-circuit voltage is processed based on the first preset threshold to obtain a first estimated value sequence of open-circuit voltage. If a first preset condition is met within a set sampling length, then the first estimated value sequence of open-circuit voltage that meets the first preset condition is solved from the open-circuit voltage. The first preset condition is: in, R1 is the sampling time interval, C1 is the polarization internal resistance, and n is the polarization capacitance. Setting a second preset threshold, and processing the open-circuit voltage based on the second preset threshold to obtain a second estimated sequence of open-circuit voltage values, wherein processing the open-circuit voltage based on the second preset threshold to obtain the second estimated sequence of open-circuit voltage values ​​includes: A second set condition is determined based on a second set threshold; and at least two sets of data segments satisfying the second set condition are identified from the open-circuit voltage; the second set condition is: Where m is the second preset threshold; Each data segment is processed to obtain the second estimated value of the open-circuit voltage corresponding to each data segment; A sequence of second estimates of open-circuit voltage is obtained based on multiple sets of second estimates of open-circuit voltage; Based on the pre-calibrated SOC-OCV relationship, a first SOC estimation sequence corresponding to the first open-circuit voltage estimation sequence and a second SOC estimation sequence corresponding to the second open-circuit voltage estimation sequence are determined. Based on the linear relationship between changes in battery charge and changes in State of Charge (SOC), the battery capacity is obtained by processing the first SOC estimation sequence and the second SOC estimation sequence.

2. The online battery capacity estimation method according to claim 1, characterized in that, The process of processing the RC equivalent circuit model to obtain the parameters to be estimated includes: The transfer function is obtained by processing the RC equivalent circuit model. The transfer function is identified online using the least squares method with a forgetting factor to estimate the battery parameters.

3. The online battery capacity estimation method according to claim 2, characterized in that, The RC equivalent circuit model is as follows: Where U1 is the polarization voltage, U t R is the terminal voltage, I is the current, OCV is the open-circuit voltage, R1 is the polarization resistance, R0 is the ohmic resistance, and τ is the time constant.

4. The online battery capacity estimation method according to claim 3, characterized in that, The second estimated value group of open-circuit voltage corresponding to each data segment is obtained by processing each data segment. The two data points of each group are respectively used as the second initial estimate of the open circuit voltage corresponding to the starting point of each group and the second final estimate of the open circuit voltage corresponding to the ending point of each group; Linear interpolation is performed based on the second initial estimate of the open-circuit voltage and the second final estimate of the open-circuit voltage for each data segment to obtain the second estimate of the open-circuit voltage for each data segment.

5. The online battery capacity estimation method according to claim 1, characterized in that, The step of processing the first SOC estimation sequence and the second SOC estimation sequence to obtain the battery capacity based on the linear relationship between battery charge change and SOC change includes: Based on the linear relationship between battery charge change and SOC change, the first SOC estimation value sequence is processed to obtain the first capacity value, and the second SOC estimation value sequence is processed to obtain the second capacity value. The battery capacity is obtained by weighting the first capacity value and the second capacity value.

6. The online battery capacity estimation method according to claim 1, characterized in that, The linear relationship between battery capacity change and SOC change is known as the SOC-capacity gain method; the SOC-capacity gain method calculates battery capacity by the ratio of charge / discharge capacity change to the corresponding SOC change.

7. An online battery capacity estimation device, characterized in that, The method includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the steps of the online battery capacity estimation method according to any one of claims 1-6.

Images