Method and system for determining soc of energy storage battery based on electrochemical impedance spectrum

Abstract

The application discloses a method and system for determining the SOC of an energy storage battery based on electrochemical impedance spectroscopy, comprising the following steps: obtaining the electrochemical impedance spectroscopy in the frequency domain under forward excitation by using forward AC sinusoidal excitation sweep measurement; obtaining the electrochemical impedance spectroscopy in the frequency domain under reverse excitation by using reverse AC sinusoidal excitation sweep measurement; comparing the electrochemical impedance spectroscopy in the frequency domain under forward excitation with the electrochemical impedance spectroscopy in the frequency domain under reverse excitation, and screening out the influence factors by Pearson correlation degree calculation; establishing a mathematical model to express the relationship between the influence factors and the SOC of the energy storage battery; and inputting the real-time collected data of the energy storage battery into the mathematical model to obtain the corresponding SOC; the application has the advantages of short modeling cycle, small calculation amount and high SOC estimation efficiency.

Description

Technical Field

[0001] This invention relates to the field of energy storage battery technology, and specifically to a method and system for determining the state of charge (SOC) of energy storage batteries based on electrochemical impedance spectroscopy. Background Technology

[0002] Energy storage batteries have been widely used in many fields such as photovoltaic power generation, wind power generation, and grid peak shaving and frequency regulation. Rapid state of charge (SOC) estimation is one of the key issues in battery management systems, with a wide range of applications. SOC estimation can prevent irreversible damage caused by overcharging and over-discharging, optimize battery operation, and improve the safety of the battery system. Furthermore, the changing patterns of SOC can reflect signs of battery performance degradation or aging, allowing for early maintenance measures and helping to extend battery life. In addition, in renewable energy systems, such as solar and wind power systems, SOC estimation can provide guidance for optimizing energy storage and distribution processes to improve energy storage efficiency. Therefore, SOC estimation technology has always been one of the focal points of research and development in battery technology and related applications.

[0003] Battery SOC estimation is a complex problem influenced by various factors, including battery type, temperature, current, charging and discharging rates. SOC estimation typically involves inference using the battery's current-voltage response. Internal electrochemical reactions, electrode material properties, and battery construction all significantly impact SOC estimation. Currently, numerous methods and techniques exist for battery SOC estimation, including those based on voltage, current, temperature, and electrochemical models, as well as those based on statistical methods and machine learning. Chinese Patent Publication No. CN114563716A discloses a method and apparatus for estimating the SOC of retired batteries based on electrochemical impedance spectroscopy. The method includes: acquiring electrochemical impedance spectra of several retired batteries at different SOC values; obtaining characteristic parameters based on the electrochemical impedance spectra; filtering the characteristic parameters to obtain filtered characteristic parameters; constructing an estimation model; and training the estimation model based on the filtered characteristic parameters to obtain a fully trained estimation model, ensuring the accuracy of retired battery SOC estimation. However, this patented method relies on a large number of samples and detection data for correlation calculations to filter characteristic parameters during the modeling process, resulting in a long modeling cycle, high computational load, and low SOC estimation efficiency. Summary of the Invention

[0004] The technical problem to be solved by this invention is that existing SOC estimation methods rely on a large number of samples and detection data for correlation calculation, resulting in long modeling cycles, large computational loads, and low SOC estimation efficiency.

[0005] This invention solves the above-mentioned technical problems through the following technical means: a method for determining the state of charge (SOC) of energy storage batteries based on electrochemical impedance spectroscopy, comprising the following steps:

[0006] Step 1: Obtain the electrochemical impedance spectrum in the frequency domain under positive excitation using a positive AC sinusoidal excitation sweep frequency measurement.

[0007] Step 2: Obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation using reverse AC sinusoidal excitation frequency sweep measurement;

[0008] Step 3: Compare the electrochemical impedance spectroscopy results in the frequency domain under positive excitation and the electrochemical impedance spectroscopy results in the frequency domain under reverse excitation, and screen out the influencing factors by calculating the Pearson correlation.

[0009] Step 4: Establish a mathematical model to express the relationship between influencing factors and the SOC of the energy storage battery;

[0010] Step 5: Input the real-time collected energy storage battery data into the mathematical model to obtain the corresponding SOC.

[0011] Further, step one includes:

[0012] Based on the target frequency domain on a logarithmic scale, a forward AC sinusoidal current excitation is sequentially applied to the energy storage battery. The forward excitation current equation is expressed as follows:

[0013]

[0014] in, For positive alternating current, A + ω is the positive current amplitude, ω is the angular frequency, and t is the current time.

[0015] The excitation signal is converted into amplitude and angle form, and the equation is expressed as follows:

[0016]

[0017] Obtain the response voltage signal of the energy storage battery at the corresponding frequency point f, expressed by the equation as follows:

[0018]

[0019] in, For positive AC response voltage, B + The positive voltage amplitude, φ + It is a positive voltage phase;

[0020] The response signal is converted into amplitude and angle form, and the equation is as follows:

[0021]

[0022] Where θ1° is the phase angle of the positive AC response voltage;

[0023] The complex impedance at the target frequency f is obtained from equations (2) and (4), and the equation is expressed as follows:

[0024]

[0025] in, The positive complex impedance at the target frequency point f;

[0026] Convert to complex form

[0027]

[0028] Where j is the symbol for the imaginary part;

[0029] Adjust the excitation frequency within the target frequency domain and repeat the process of formulas (1)-(6) to obtain the electrochemical impedance spectrum in the frequency domain under positive excitation.

[0030] Furthermore, step two includes:

[0031] Based on the target frequency domain on a logarithmic scale, reverse AC sinusoidal current excitation is sequentially applied to the energy storage battery. The reverse excitation current equation is expressed as follows:

[0032]

[0033] in, To incentivize communication, A - This represents the reverse current amplitude.

[0034] The excitation signal is converted into amplitude and angle form, and the equation is expressed as follows:

[0035]

[0036] Obtain the response voltage signal of the energy storage battery at the corresponding frequency point f, expressed by the equation as follows:

[0037]

[0038] in, For the reverse AC response voltage, B - The reverse voltage amplitude, φ - It is the reverse voltage phase;

[0039] The response signal is converted into amplitude and angle form, and the equation is as follows:

[0040]

[0041] Where θ2° is the phase angle of the reverse AC response voltage;

[0042] The complex impedance at the target frequency f is obtained from equations (8) and (10), and the equation is expressed as follows:

[0043]

[0044] in, The reverse complex impedance at the target frequency point f;

[0045] Convert to complex form

[0046]

[0047] Adjust the excitation frequency in the target frequency domain and repeat the process of formulas (7)-(12) to obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation.

[0048] Furthermore, step three includes:

[0049] Observation of test data from randomly selected battery samples with health within a preset range; analysis was conducted on electrochemical impedance spectroscopy (EIS) in the 0.1-0.01 Hz frequency domain under forward excitation and EIS in the reverse excitation frequency domain, which showed good linear correlation with SOC. The real and imaginary parts of the EIS obtained by bidirectional frequency sweep detection were subtracted to extract the differences in the real and imaginary parts, impedance point distance, and impedance point slope of the forward and reverse EIS, which were named IF1-IF4 respectively. The feature with the best linear correlation with SOC was selected by Pearson correlation calculation as the influencing factor.

[0050] Furthermore, the influencing factor is the difference between the imaginary part of the positive electrochemical impedance spectrum and the negative electrochemical impedance spectrum.

[0051] Furthermore, the mathematical model in step four includes a polynomial fitting model, a piecewise function fitting model, and a trigonometric function fitting model.

[0052] Furthermore, in step four, the fitting effects and the number of parameters of the three mathematical models are compared, and the trigonometric function fitting model with good fitting effect, few parameters, and good parameter stability is selected as the final mathematical model.

[0053] Furthermore, the formula for the trigonometric function fitting model is y = a1tan(a2x + a3), where a1, a2, and a3 are model coefficients, x is an influence factor, and y is the SOC estimate.

[0054] This invention also provides a system for determining the state of charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy, comprising:

[0055] The positive excitation module is used to obtain the electrochemical impedance spectrum in the frequency domain under positive excitation by frequency sweep measurement using positive AC sinusoidal excitation.

[0056] The reverse excitation module is used to obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation by frequency sweep measurement using reverse AC sinusoidal excitation.

[0057] The impact factor acquisition module is used to compare the results of electrochemical impedance spectroscopy in the frequency domain under positive excitation and electrochemical impedance spectroscopy in the frequency domain under reverse excitation, and to screen the impact factors by calculating the Pearson correlation.

[0058] The model building module is used to establish mathematical models that express the relationship between influencing factors and the SOC of energy storage batteries.

[0059] The SOC estimation module is used to input real-time collected energy storage battery data into the mathematical model to obtain the corresponding SOC.

[0060] Furthermore, the positive excitation module is also used for:

[0061] Based on the target frequency domain on a logarithmic scale, a forward AC sinusoidal current excitation is sequentially applied to the energy storage battery. The forward excitation current equation is expressed as follows:

[0062]

[0063] in, For positive alternating current, A + ω is the positive current amplitude, ω is the angular frequency, and t is the current time.

[0064] The excitation signal is converted into amplitude and angle form, and the equation is expressed as follows:

[0065]

[0066] Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows:

[0067]

[0068] in, For positive AC response voltage, B + The positive voltage amplitude, φ + It is a positive voltage phase;

[0069] The response signal is converted into amplitude and angle form, and the equation is as follows:

[0070]

[0071] Where θ1° is the phase angle of the positive AC response voltage;

[0072] The complex impedance at the target frequency f is obtained from equations (2) and (4), and the equation is expressed as follows:

[0073]

[0074] in, The positive complex impedance at the target frequency point f;

[0075] Convert to complex form

[0076]

[0077] Where j is the symbol for the imaginary part;

[0078] Adjust the excitation frequency within the target frequency domain and repeat the process of formulas (1)-(6) to obtain the electrochemical impedance spectrum in the frequency domain under positive excitation.

[0079] Furthermore, the reverse excitation module is also used for:

[0080] Based on the target frequency domain on a logarithmic scale, reverse AC sinusoidal current excitation is sequentially applied to the energy storage battery. The reverse excitation current equation is expressed as follows:

[0081]

[0082] in, To incentivize communication, A - This represents the reverse current amplitude.

[0083] The excitation signal is converted into amplitude and angle form, and the equation is expressed as follows:

[0084]

[0085] Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows:

[0086]

[0087] in, For the reverse AC response voltage, B - The reverse voltage amplitude, φ - It is the reverse voltage phase;

[0088] The response signal is converted into amplitude and angle form, and the equation is as follows:

[0089]

[0090] Where θ2° is the phase angle of the reverse AC response voltage;

[0091] The complex impedance at the target frequency f is obtained from equations (8) and (10), and the equation is expressed as follows:

[0092]

[0093] in, The reverse complex impedance at the target frequency point f;

[0094] Convert to complex form

[0095]

[0096] Adjust the excitation frequency in the target frequency domain and repeat the process of formulas (7)-(12) to obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation.

[0097] Furthermore, the impact factor acquisition module is also used for:

[0098] Observation of test data from randomly selected multi-cell battery samples with health within a preset range; analysis was conducted on electrochemical impedance spectra in the 0.1-0.01 Hz frequency domain under forward excitation and electrochemical impedance spectra in the frequency domain under reverse excitation, which showed good linear correlation with SOC; the real and imaginary parts of the electrochemical impedance spectra obtained by bidirectional frequency sweep detection were subtracted, and the differences in the real, imaginary, impedance point distance, and impedance point slope of the forward and reverse electrochemical impedance spectra were extracted and named IF1-IF4 respectively; the features with the best linear correlation with SOC were selected by Pearson correlation calculation as influencing factors.

[0099] Furthermore, the influencing factor is the difference between the imaginary part of the positive electrochemical impedance spectrum and the negative electrochemical impedance spectrum.

[0100] Furthermore, the mathematical models in the model building module include polynomial fitting models, piecewise function fitting models, and trigonometric function fitting models.

[0101] Furthermore, the model building module compares the fitting effects and the number of parameters of the three mathematical models, and selects the trigonometric function fitting model with good fitting effect, few parameters, and good parameter stability as the final mathematical model.

[0102] Furthermore, the formula for the trigonometric function fitting model is y = a1tan(a2x + a3), where a1, a2, and a3 are model coefficients, x is an influence factor, and y is the SOC estimate.

[0103] The advantages of this invention are:

[0104] (1) This invention obtains the bidirectional electrochemical impedance spectroscopy of the battery by positive and reverse excitation, compares the results of the bidirectional electrochemical impedance spectroscopy, and screens out the influencing factors by Pearson correlation calculation. Compared with the prior art, some possible influencing parameters have been obtained in advance by comparing the bidirectional electrochemical impedance spectroscopy, and then the influencing factors are screened out by Pearson correlation calculation. It is not necessary to perform correlation calculation on all parameters, the amount of calculation is small, the modeling cycle is short, thereby improving the efficiency of SOC estimation.

[0105] (2) The present invention utilizes the difference in electrochemical impedance spectroscopy of bidirectional AC detection to estimate the SOC of a battery. This method can directly extract the SOC influencing factors with good regularity through electrochemical impedance spectroscopy, while avoiding the dependence of traditional SOC estimation methods on a large number of samples and detection data in the modeling process, reducing the amount of data processing and shortening the modeling cycle.

[0106] (3) The literature mentioned in the background directly uses the real part, imaginary part, and open-circuit voltage of the EIS (Electronic Information System) of the battery under various SOC (State of Charge) states as feature parameters and inputs them into a support vector machine regression prediction model to construct the correlation with SOC. Extensive testing data has confirmed that the real and imaginary part features extracted directly from the single EIS detection under various SOC states have poor regularity. This requires a large number of experimental samples and testing data as a model training set, resulting in high modeling complexity and low modeling efficiency. Furthermore, the SOC prediction model built on this basis does not consider the application scenarios of battery samples under different SOH (State of Charge) states. This invention uses bidirectional EIS detection, i.e., comparing the results of forward and reverse EIS detection. The extracted features, such as the difference in the imaginary part and the difference in impedance point distance between the forward and reverse EIS, have a high correlation with the battery SOC. Multiple samples have verified that these features also show a good correlation with SOC in the testing data of batteries under different SOH states. In addition, since the extracted features have a high correlation with the battery SOC, the complexity of modeling can be greatly reduced. Attached Figure Description

[0107] Figure 1 This is a flowchart of the method for determining the state of charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy, as disclosed in the embodiments of the present invention.

[0108] Figure 2 This is a graph showing the trend of IF as a function of SOC in the energy storage battery SOC determination method based on electrochemical impedance spectroscopy disclosed in the embodiments of the present invention.

[0109] Figure 3 This is a graph showing the fitting effect of the mathematical model in the energy storage battery SOC determination method based on electrochemical impedance spectroscopy disclosed in the embodiments of the present invention. Detailed Implementation

[0110] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0111] Example 1

[0112] like Figure 1 As shown, Embodiment 1 of the present invention provides a method for determining the State of Charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy. In this embodiment, the detection equipment is an electrochemical workstation (model: CSC350M). The test sample is a lithium iron phosphate battery (capacity: 15Ah). The method includes the following steps:

[0113] S1. Obtain the electrochemical impedance spectroscopy in the frequency domain under positive excitation using a forward AC sinusoidal excitation frequency sweep measurement; the specific process is as follows:

[0114] Establishing a target frequency domain on a logarithmic scale (0.01Hz-1000Hz), a forward AC sinusoidal current excitation is sequentially applied to the energy storage battery. The forward excitation current equation is expressed as follows:

[0115]

[0116] in, For positive alternating current, A + ω is the positive current amplitude, ω is the angular frequency, and t is the current time.

[0117] The excitation signal is converted into amplitude and angle form, and the equation is expressed as follows:

[0118]

[0119] Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows:

[0120]

[0121] in, For positive AC response voltage, B + The positive voltage amplitude, φ + It is a positive voltage phase;

[0122] The response signal is converted into amplitude and angle form, and the equation is as follows:

[0123]

[0124] Where θ1° is the phase angle of the positive AC response voltage;

[0125] The complex impedance at the target frequency f is obtained from equations (2) and (4), and the equation is expressed as follows:

[0126]

[0127] in, The positive complex impedance at the target frequency point f;

[0128] Convert to complex form

[0129]

[0130] Where j is the symbol for the imaginary part;

[0131] By adjusting the excitation frequency within the target frequency domain and repeating the process of formulas (1)-(6), the electrochemical impedance spectrum in the frequency domain (0.01Hz-1000Hz) under positive excitation is obtained.

[0132] S2. Obtain the electrochemical impedance spectroscopy in the frequency domain under reverse excitation using reverse AC sinusoidal excitation frequency sweep measurement; the specific process is as follows:

[0133] Establishing a target frequency domain on a logarithmic scale (0.01Hz-1000Hz), reverse AC sinusoidal current excitation is sequentially applied to the energy storage battery. The reverse excitation current equation is expressed as follows:

[0134]

[0135] in, To incentivize communication, A - This represents the reverse current amplitude.

[0136] The excitation signal is converted into amplitude and angle form, and the equation is expressed as follows:

[0137]

[0138] Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows:

[0139]

[0140] in, For the reverse AC response voltage, B - The reverse voltage amplitude, φ - It is the reverse voltage phase;

[0141] The response signal is converted into amplitude and angle form, and the equation is as follows:

[0142]

[0143] Where θ2° is the phase angle of the reverse AC response voltage;

[0144] The complex impedance at the target frequency f is obtained from equations (8) and (10), and the equation is expressed as follows:

[0145]

[0146] in, The reverse complex impedance at the target frequency point f;

[0147] Convert to complex form

[0148]

[0149] By adjusting the excitation frequency within the target frequency domain and repeating the process of formulas (7)-(12), the electrochemical impedance spectrum in the frequency domain (0.01Hz-1000Hz) under reverse excitation is obtained.

[0150] Based on constant current charging and discharging, the voltage variation law shows that within the 50%-100% SOC range, the absolute value of the slope of the voltage curve increases slowly with increasing SOC. Therefore, the response voltage amplitude of the positive half-cycle of the forward AC sinusoidal current excitation gradually increases with increasing SOC, and the voltage reset effect also increases the corresponding negative half-cycle amplitude. However, since the reverse AC sinusoidal current excitation injects the negative half-cycle first, the negative half-cycle amplitude of the response voltage is slightly smaller than that of the positive half-cycle of the forward excitation. Furthermore, the voltage reset effect has limited effect on the positive half-cycle of the response in this SOC region. Therefore, within the 50%-100% SOC range, the impedance amplitude of the forward-detected EIS (electrochemical impedance spectroscopy) in the low-frequency domain is larger than that of the reverse-detected EIS, and this difference tends to increase with increasing SOC. Similarly, within the 50%-0% SOC range, the absolute value of the slope of the voltage curve increases slowly as the SOC decreases. Therefore, the response voltage amplitude of the reverse AC sinusoidal current excitation in the negative half-cycle gradually increases with the decrease of SOC, and the voltage reset effect will increase the corresponding negative half-cycle amplitude to a certain extent. However, since the positive half-cycle is injected first for the forward AC sinusoidal current excitation, the positive half-cycle amplitude of the response voltage is slightly smaller than that of the reverse excitation in the negative half-cycle. Furthermore, the voltage reset effect has limited impact on the negative half-cycle response within this SOC region. Therefore, within the 50%-0% SOC range, the impedance amplitude of the reverse-detected EIS in the low-frequency domain is larger than that of the forward-detected EIS, and this difference tends to increase as the SOC decreases. This bidirectional EIS detection method can directly extract influencing factors reflecting the SOC, avoiding the dependence on a large number of detection samples and data in the modeling process of traditional SOC estimation methods.

[0151] S3. Compare the electrochemical impedance spectroscopy (EIS) results in the frequency domain under forward excitation and reverse excitation, and screen out influencing factors using Pearson correlation calculation; the specific process is as follows:

[0152] By observing the test data of three randomly selected battery samples with similar health levels, such as... Figure 2 As shown, impedance spectra in the frequency domain (0.1-0.01Hz) with good linear correlation to SOC were selected for analysis. The real and imaginary parts of the electrochemical impedance spectra obtained by bidirectional frequency sweep detection were subtracted, and the differences in the real and imaginary parts, impedance point distance, and impedance point slope of the forward and reverse electrochemical impedance spectra were named IF1-IF4 respectively. The IFs with good linear correlation to SOC were selected by Pearson correlation calculation, and the results are shown in Tables 1 to 3.

[0153] Table 1. Mean Correlation Coefficients of Three Samples in the 10%-90% SOC Range

[0154] Pearson correlation coefficient -0.3607 -0.9770 0.6353 0.1450 Significance level (P-value) 0.3400 6.67E-06 0.1360 0.7283

[0155] Table 2. Mean Correlation Coefficients of Three Samples in the 0%-90% SOC Range

[0156] Pearson correlation coefficient -0.3550 -0.7743 0.6630 0.2010 Significance level (P-value) 0.3403 0.0087 0.0407 0.5800

[0157] Table 3. Mean Correlation Coefficients of Three Samples in the 0%-100% SOC Range

[0158]

[0159] Comparing Tables 1-3, it can be seen that IF2 and IF3 have a good correlation with SOC. Furthermore, the significance levels of IF in Tables 2 and 3 are lower than those in Table 1. This is because the change in IF is relatively gradual in the 20%-80% SOC range, while the slope of IF change increases sharply below 20% SOC and above 80% SOC, and both show a negative correlation under full charge. Since the Pearson correlation coefficient of IF2 in the 10%-90% SOC range can be as high as -0.977, it can be expressed through a mathematical model. The lower significance levels of IF1 and IF4 in Tables 1-3 are because, in actual detection, the real part of EIS (electrochemical impedance spectroscopy) is easily affected by differences in line impedance and contact impedance, while the imaginary part of EIS is less affected. The imaginary part information is mainly related to the electron and ion transport processes in the battery and is used to assess the electrochemical reaction rate, charge transfer, and mass transfer processes at the electrode surface. A higher imaginary part value usually indicates a slower charge transfer process inside the battery, which is a limiting factor for the electrochemical reaction rate. In summary, IF2, with the optimal Pearson correlation coefficient, was selected for modeling. Therefore, in this embodiment, Pearson correlation coefficients greater than or equal to 0.9 are used as influencing factors, while others are discarded. The influencing factor in this embodiment is the difference between the imaginary parts of the forward and reverse electrochemical impedance spectra.

[0160] S4. Establish a mathematical model to express the relationship between influencing factors and the SOC of energy storage batteries; the specific process is as follows:

[0161] By observing the changing trends of the test sample data, a mathematical model is selected, the model parameters are determined, and the best-performing mathematical model is selected through comparison. The mathematical models include polynomial fitting models, piecewise function fitting models, and trigonometric function fitting models. In this embodiment, the fitting effects and number of parameters of the three mathematical models are compared, and the trigonometric function fitting model, which has a good fitting effect, fewer parameters, and better parameter stability, is selected as the final mathematical model. The fitting formulas for each model are described below.

[0162] (1) Polynomial fitting model

[0163] Based on the variation patterns of the IF data selected in step S3, a polynomial mathematical model is established. A better fitting effect is achieved when the polynomial reaches the ninth degree or higher; therefore, a tenth-degree polynomial is chosen, and the equation is expressed as follows:

[0164]

[0165] The model parameters a0-a10 were determined using the impact factor and SOC, and the results are shown in Table 4.

[0166] Table 4. Parameters of the polynomial fitting model

[0167]

[0168] (2) Piecewise function fitting model

[0169] A piecewise mathematical model is established. Based on the variation pattern of the IF data selected in step S3, it can be seen that the slope of the function increases sharply in the 0%-20% SOC and 80%-100% SOC intervals. Therefore, three segments of the function are defined to correspond to the SOC variation patterns in different intervals. The piecewise function equation is expressed as follows:

[0170]

[0171] The model parameters a1-a10 were determined using the impact factor and SOC, and the results are shown in Table 5.

[0172] Table 5. Parameters of the piecewise function fitting model

[0173]

[0174] (3) Trigonometric function fitting model

[0175] Based on the changing patterns of the IF data selected in step S3, a trigonometric function model is established, and the functional equation is expressed as follows:

[0176] y = a1tan (a2x + a3) (15)

[0177] The model parameters a1-a3 were determined using the impact factor and SOC, and the results are shown in Table 6.

[0178] Table 6. Parameters of the trigonometric function fitting model

[0179]

[0180]

[0181] The fitting effect of the trigonometric function fitting model is as follows: Figure 3 .

[0182] The results of comparing the three mathematical models are shown in Table 7.

[0183] Table 7. Parameter Comparison

[0184] Parameter fluctuation mean % 192.4019 63.8 0.26 Parameters 11 10 3

[0185] In summary, all three mathematical models achieve good fitting results. Model 3 contains only three parameters, and each parameter exhibits the best stability. Therefore, estimating the SOC of the energy storage battery using the trigonometric function model is the optimal solution.

[0186] S5. Input the real-time collected energy storage battery data into the mathematical model to obtain the corresponding SOC.

[0187] Through the above technical solutions, this invention provides a battery SOC estimation method based on bidirectional detection of electrochemical impedance spectroscopy (EIS) of energy storage batteries. This method has a relatively simple detection and modeling approach. It uses the differences in EIS of bidirectional AC detection of energy storage batteries to estimate battery SOC. It can directly extract SOC influencing factors with good regularity through EIS, while avoiding the dependence of traditional SOC estimation methods on a large number of samples and detection data in the modeling process, reducing the amount of data processing and shortening the modeling cycle.

[0188] Example 2

[0189] Based on Example 1, Example 2 of the present invention also provides a SOC determination system for energy storage batteries based on electrochemical impedance spectroscopy, including:

[0190] The positive excitation module is used to obtain the electrochemical impedance spectrum in the frequency domain under positive excitation by frequency sweep measurement using positive AC sinusoidal excitation.

[0191] The reverse excitation module is used to obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation by frequency sweep measurement using reverse AC sinusoidal excitation.

[0192] The impact factor acquisition module is used to compare the results of electrochemical impedance spectroscopy in the frequency domain under positive excitation and electrochemical impedance spectroscopy in the frequency domain under reverse excitation, and to screen the impact factors by calculating the Pearson correlation.

[0193] The model building module is used to establish mathematical models that express the relationship between influencing factors and the SOC of energy storage batteries.

[0194] The SOC estimation module is used to input real-time collected energy storage battery data into the mathematical model to obtain the corresponding SOC.

[0195] Specifically, the positive incentive module is also used for:

[0196] Based on the target frequency domain on a logarithmic scale, a forward AC sinusoidal current excitation is sequentially applied to the energy storage battery. The forward excitation current equation is expressed as follows:

[0197]

[0198] in, For positive alternating current, A + ω is the positive current amplitude, ω is the angular frequency, and t is the current time.

[0199] The excitation signal is converted into amplitude and angle form, and the equation is expressed as follows:

[0200]

[0201] Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows:

[0202]

[0203] in, For positive AC response voltage, B + The positive voltage amplitude, φ + It is a positive voltage phase;

[0204] The response signal is converted into amplitude and angle form, and the equation is as follows:

[0205]

[0206] Where θ1° is the phase angle of the positive AC response voltage;

[0207] The complex impedance at the target frequency f is obtained from equations (2) and (4), and the equation is expressed as follows:

[0208]

[0209] in, The positive complex impedance at the target frequency point f;

[0210] Convert to complex form

[0211]

[0212] Where j is the symbol for the imaginary part;

[0213] Adjust the excitation frequency within the target frequency domain and repeat the process of formulas (1)-(6) to obtain the electrochemical impedance spectrum in the frequency domain under positive excitation.

[0214] More specifically, the reverse excitation module is also used for:

[0215] Based on the target frequency domain on a logarithmic scale, reverse AC sinusoidal current excitation is sequentially applied to the energy storage battery. The reverse excitation current equation is expressed as follows:

[0216]

[0217] in, To incentivize communication, A - This represents the reverse current amplitude.

[0218] The excitation signal is converted into amplitude and angle form, and the equation is expressed as follows:

[0219]

[0220] Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows:

[0221]

[0222] in, For the reverse AC response voltage, B - The reverse voltage amplitude, φ - It is the reverse voltage phase;

[0223] The response signal is converted into amplitude and angle form, and the equation is as follows:

[0224]

[0225] Where θ2° is the phase angle of the reverse AC response voltage;

[0226] The complex impedance at the target frequency f is obtained from equations (8) and (10), and the equation is expressed as follows:

[0227]

[0228] in, The reverse complex impedance at the target frequency point f;

[0229] Convert to complex form

[0230]

[0231] Adjust the excitation frequency in the target frequency domain and repeat the process of formulas (7)-(12) to obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation.

[0232] Specifically, the impact factor acquisition module is also used for:

[0233] Observation of test data from randomly selected multi-cell battery samples with health within a preset range; analysis was conducted on electrochemical impedance spectra in the 0.1-0.01 Hz frequency domain under forward excitation and electrochemical impedance spectra in the frequency domain under reverse excitation, which showed good linear correlation with SOC; the real and imaginary parts of the electrochemical impedance spectra obtained by bidirectional frequency sweep detection were subtracted, and the differences in the real, imaginary, impedance point distance, and impedance point slope of the forward and reverse electrochemical impedance spectra were extracted and named IF1-IF4 respectively; the features with the best linear correlation with SOC were selected by Pearson correlation calculation as influencing factors.

[0234] More specifically, the influencing factor is the difference between the imaginary part of the positive and negative electrochemical impedance spectra.

[0235] Specifically, the mathematical models in the model building module include polynomial fitting models, piecewise function fitting models, and trigonometric function fitting models.

[0236] More specifically, in the model building module, the fitting effect and number of parameters of the three mathematical models are compared, and the trigonometric function fitting model with good fitting effect, few parameters, and good parameter stability is selected as the final mathematical model.

[0237] More specifically, the formula for the trigonometric function fitting model is y = a1tan(a2x + a3), where a1, a2, and a3 are model coefficients, x is an influence factor, and y is the SOC estimate.

[0238] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for determining the state of charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy, characterized in that, Includes the following steps: Step 1: Obtain the electrochemical impedance spectrum in the frequency domain under positive excitation using a positive AC sinusoidal excitation sweep frequency measurement. Step 2: Obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation using reverse AC sinusoidal excitation frequency sweep measurement; Step 3: Compare the electrochemical impedance spectroscopy (EIS) results in the frequency domain under forward excitation and under reverse excitation, and screen out influencing factors using Pearson correlation calculation; Step 3 includes: Observe the test data of randomly selected multi-cell battery samples with health within the preset range, and select the electrochemical impedance spectra in the frequency domain of 0.1-0.01Hz under positive excitation and the electrochemical impedance spectra in the frequency domain under reverse excitation with good linear correlation with SOC for analysis. The real and imaginary parts of the electrochemical impedance spectroscopy obtained by bidirectional frequency sweep detection are subtracted to extract the real part difference, imaginary part difference, impedance point distance difference, and impedance point slope difference features of the forward and reverse electrochemical impedance spectra, which are named IF1-IF4 respectively. The features with the best linear correlation with SOC are selected by Pearson correlation calculation and used as influencing factors. Step 4: Establish a mathematical model to express the relationship between influencing factors and SOC of energy storage battery; the mathematical model in step 4 includes a polynomial fitting model, a piecewise function fitting model and a trigonometric function fitting model; in step 4, the fitting effect and number of parameters of the three mathematical models are compared, and the trigonometric function fitting model with good fitting effect, few parameters and good parameter stability is selected as the final mathematical model. Step 5: Input the real-time collected energy storage battery data into the mathematical model to obtain the corresponding SOC.

2. The method for determining the state of charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy according to claim 1, characterized in that, Step one includes: Based on the target frequency domain on a logarithmic scale, a forward AC sinusoidal current excitation is sequentially applied to the energy storage battery. The forward excitation current equation is expressed as follows: （1） in, For positive alternating current, The positive current amplitude, Angular frequency, The current moment; Converting the positive AC excitation current into amplitude and angle form, the equation is as follows: (2) Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows: (3) in, This is the positive AC response voltage. This represents the positive voltage amplitude. It is a positive voltage phase; The positive AC response voltage is converted into amplitude and angle form, expressed by the equation as follows: (4) in, The phase angle of the positive AC response voltage; The target frequency is obtained from equations (2) and (4). The complex impedance at point is expressed by the equation as follows: (5) in, Target frequency The positive complex impedance at the location; Convert to complex form (6) in, The symbol for the imaginary part; Adjust the excitation frequency within the target frequency domain and repeat the process of formulas (1)-(6) to obtain the electrochemical impedance spectrum in the frequency domain under positive excitation.

3. The method for determining the state of charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy according to claim 2, characterized in that, Step two includes: Based on the target frequency domain on a logarithmic scale, reverse AC sinusoidal current excitation is sequentially applied to the energy storage battery. The reverse excitation current equation is expressed as follows: （7） in, To encourage reverse communication, This represents the reverse current amplitude. The reverse excitation AC is converted into amplitude and angle form, expressed by the equation as follows: (8) Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows: (9) in, This is the reverse AC response voltage. This is the reverse voltage amplitude. It is the reverse voltage phase; The reverse AC response voltage is converted into amplitude and angle form, expressed by the equation as follows: (10) in, The phase angle of the reverse AC response voltage; The target frequency is obtained from equations (8) and (10). The complex impedance at point is expressed by the equation as follows: (11) in, Target frequency The reverse complex impedance at the point; Convert to complex form (12) Adjust the excitation frequency in the target frequency domain and repeat the process of formulas (7)-(12) to obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation.

4. The method for determining the state of charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy according to claim 1, characterized in that, The influencing factor is the difference between the imaginary part of the positive and negative electrochemical impedance spectra.

5. The method for determining the state of charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy according to claim 1, characterized in that, The formula for the trigonometric function fitting model is as follows: ,in, , and All are model coefficients. As the impact factor, This is the estimated SOC value.

6. A system for determining the state of charge (SOC) of an energy storage battery based on electrochemical impedance spectroscopy, characterized in that, include: The positive excitation module is used to obtain the electrochemical impedance spectrum in the frequency domain under positive excitation by frequency sweep measurement using positive AC sinusoidal excitation. The reverse excitation module is used to obtain the electrochemical impedance spectrum in the frequency domain under reverse excitation by frequency sweep measurement using reverse AC sinusoidal excitation. The impact factor acquisition module is used to compare the results of electrochemical impedance spectroscopy (EIS) in the frequency domain under positive excitation and electrochemical impedance spectroscopy in the frequency domain under reverse excitation, and to screen the impact factors through Pearson correlation calculation; the impact factor acquisition module is also used for: Observe the test data of randomly selected multi-cell battery samples with health within the preset range, and select the electrochemical impedance spectra in the frequency domain of 0.1-0.01Hz under positive excitation and the electrochemical impedance spectra in the frequency domain under reverse excitation with good linear correlation with SOC for analysis. The real and imaginary parts of the electrochemical impedance spectroscopy obtained by bidirectional frequency sweep detection are subtracted to extract the real part difference, imaginary part difference, impedance point distance difference, and impedance point slope difference features of the forward and reverse electrochemical impedance spectra, which are named IF1-IF4 respectively. The features with the best linear correlation with SOC are selected by Pearson correlation calculation and used as influencing factors. The model building module is used to establish a mathematical model to express the relationship between influencing factors and the SOC of the energy storage battery. The mathematical models in the model building module include a polynomial fitting model, a piecewise function fitting model, and a trigonometric function fitting model. The model building module compares the fitting effect and the number of parameters of the three mathematical models, and selects the trigonometric function fitting model, which has a good fitting effect, fewer parameters, and better parameter stability, as the final mathematical model. The SOC estimation module is used to input real-time collected energy storage battery data into the mathematical model to obtain the corresponding SOC.

7. The energy storage battery SOC determination system based on electrochemical impedance spectroscopy according to claim 6, characterized in that, The positive excitation module is also used for: Based on the target frequency domain on a logarithmic scale, a forward AC sinusoidal current excitation is sequentially applied to the energy storage battery. The forward excitation current equation is expressed as follows: （1） in, For positive alternating current, The positive current amplitude, Angular frequency, The current moment; Converting the positive AC excitation current into amplitude and angle form, the equation is as follows: (2) Obtain the response voltage signal of the energy storage battery at the corresponding frequency point, expressed by the equation as follows: (3) in, This is the positive AC response voltage. This represents the positive voltage amplitude. It is a positive voltage phase; The positive AC response voltage is converted into amplitude and angle form, expressed by the equation as follows: (4) in, The phase angle of the positive AC response voltage; The target frequency is obtained from equations (2) and (4). The complex impedance at point is expressed by the equation as follows: (5) in, Target frequency The positive complex impedance at the location; Convert to complex form (6) in, The symbol for the imaginary part; Adjust the excitation frequency within the target frequency domain and repeat the process of formulas (1)-(6) to obtain the electrochemical impedance spectrum in the frequency domain under positive excitation.

Images