A method for accurately calculating SOC based on battery model and least squares

Abstract

The present invention relates to a method for accurately calculating SOC based on a battery model and least squares, the method comprising the following steps: S1-determining the relationship between SOC and current and voltage, and obtaining a mathematical data correspondence table of SOC and current and voltage; S2-determining the actual capacity Q by using the least squares method; c and the coulomb counter deviation Q0; S3-using Q c The SOC value at any time t is calculated based on #imgabs0# and the coulomb counter deviation Q0. This method for accurately calculating SOC based on a battery model and least squares overcomes the shortcomings of existing voltage lookup table methods and current integration methods. By determining that SOC depends on the relationship between current and voltage, it applies the least squares principle to combine the voltage lookup table method and current integration method to estimate the actual capacity and deviation, thereby obtaining a stable and accurate SOC value.

Description

Technical Field

[0001] The present invention relates to a method for accurately calculating SOC based on a battery model and least squares. Background Art

[0002] Battery state-of-charge (SOC) detection has evolved over the past two decades: from 2000 to 2005, it was primarily based on voltage-SOC comparison tables; from 2005 to 2010, filtering, coulomb counters, and voltage correction integration were applied; from 2010 to 2015, a combination of voltage lookup tables and coulomb counters was employed; and since 2016, progress has been muted. The application of big data concepts, data-based learning, the selection of feature data for learning, and the use of neural network methods to train nonlinear relationships for SOC value acquisition have not been ideal. The industry generally believes that the difficulty lies in determining the conditions of a fully charged or discharged battery. A common solution is to account for battery aging and employ a patching approach based on voltage lookup tables and coulomb counters to make the SOC curve display as linear as possible. Attempts have been made to construct universal algorithm models based on voltage, capacity, and coulomb counter principles, and to develop concise, universal formulas. However, these results have proven inadequate for generalization. Therefore, the most viable avenue for improvement is to identify the SOC dependency relationship (SOC = f(v, temperature, k, ...)) related to factors such as voltage, temperature, number of charge cycles, and discharge rate. Commonly used SOC estimation methods in the industry include voltage lookup tables and current integration. The voltage lookup table method suffers from unstable SOC values ​​and large errors due to voltage fluctuations. While the current integration method provides stable SOC, the error between the predicted and actual SOC values ​​increases with age due to the decay of rated capacity and the accumulation of coulomb counter measurement errors. Summary of the Invention

[0003] The technical problem to be solved by the present invention is: to overcome the defects of the existing voltage lookup table method and current integration method, and to provide a method for accurately calculating SOC based on a battery model and least squares. By determining that SOC depends on the relationship between current and voltage, applying the least squares principle, the voltage lookup table method and the current integration method are combined to estimate the actual capacity and deviation, thereby obtaining a stable and accurate SOC value.

[0004] The method for accurately calculating SOC based on battery model and least squares includes the following steps:

[0005] S1-determine the relationship between SOC and current and voltage, and obtain a mathematical data correspondence table of SOC and current and voltage;

[0006] S2- Determine the actual capacity Q using the least squares method c and the coulomb counter deviation Q0;

[0007] S3-Utilize Q c And the coulomb counter deviation Q0, according to Calculate the SOC value at any time t.

[0008] Furthermore, the SOC in step S1 is determined to depend on the relationship between current and voltage and includes the following steps:

[0009] S11: Establish a voltage-current-SOC correspondence table, discharge the battery from full to empty at a specified current, and select a specific discharge current sequence {I k}, k=1,…,n, for each specified current I k , observe that it is in a certain voltage sequence {V j}, j = 1, ..., m. Under the SOC value, the SOC value {SOC kj}, a mathematical two-dimensional m×n table is generated, where each k-th row represents the SOC value of the k-th current value in a series of voltage sequences, and the voltage SOC corresponding curve under the specified current is obtained;

[0010] S12: For each experimental current I k The SOC value at any voltage V is calculated by the voltage sequence (V j ,SOC kj ), (j=1,…,m) and perform piecewise linear interpolation to calculate, so as to obtain the current I k The SOC depends on the piecewise linear function SOC(I k ,V), where SOC(I k ,V j )=SOC kj .;

[0011] S13: Solve the variation law of SOC dependence on voltage under any current I and any voltage V, and calculate the SOC dependence on voltage under current I by the piecewise linear function in S12. k The value of SOC(I k ,V)=SOC k (V), over (I k ,SOC k (V)), k=1,2,…,m, and perform cubic spline interpolation to obtain the SOC value SOC(I,V) at any current I and any voltage V;

[0012] S14: Set the voltage interval [V min ,V max ]N equal parts, current interval [I min ,I max ]M is divided equally, N, M are both relatively large, especially N>>n; the points on the grid (Ik ,V j ),k=0,1,…,N;j=0,1,…,M. According to the above method, we can get the value of (I k ,V j ) SOC value SOC k,j , k=0,1,…,N;j=0,1,…,M., and obtain a precise mathematical data correspondence table of SOC dependence on current and voltage.

[0013] Furthermore, let the value of the coulomb meter at time t be Q t , then the SOC value at time t During a normal discharge process, a series of t0, t1, ..., t K Voltage at the moment Current and capacity j=0,1,…,K.,S2 is the actual capacity Q c The estimation of the coulomb counter deviation Q0 includes the following steps:

[0014] S21: Look up or calculate the SOC value from the table according to the data corresponding to the SOC dependence on current and voltage. The SOC value obtained by the conventional current integration method is The difference between the two is The sum of squared errors is

[0015] S22: Determine Q0,Q using the least squares method c , so that ψ(Q0,Q c ) is minimum, and the rated capacity Q is obtained c And the deviation Q0.

[0016] As a practical optimization, specifically for the single chip algorithm selection, the least squares method is used to determine Q0, Q c In, order After the parameters q1 and q2 are calculated, the actual capacity Q is restored. c Furthermore, the single chip microcomputer uses the integer data budget calculation for the least squares operation algorithm of S22, specifically using rational numbers for the real data q1, q2 Approximation representation.

[0017] Since Q0,Q c The least squares problem of is a nonlinear least squares problem, so it is converted into a linear least squares problem. c As a fitting parameter, this is a nonlinear fitting. By cleverly replacing variables, let Then it becomes finding the best straight line SOC=q1Qt +q2 fitting. When the parameters q1 and q2 are obtained, the actual capacity Q can be restored. c And the coulomb counter deviation Q0 In addition, in order to adapt to the problem of small storage and computing power of the single-chip computer, the algorithm uses integer data budget to implement the calculation of the actual data. Specifically, for the real data q1, q2, rational numbers are used. To approximate the representation, save memory length and calculation time, and improve calculation efficiency.

[0018] The present invention provides a method for accurately calculating SOC based on a battery model and least squares, which overcomes the shortcomings of the existing voltage lookup table method and current integration method. By determining that SOC depends on the relationship between current and voltage, applying the least squares principle, the voltage lookup table method and the current integration method are combined to estimate the actual capacity and deviation, thereby obtaining a stable and accurate SOC value. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The following is a further description of a method for accurately calculating SOC based on a battery model and least squares according to the present invention with reference to the accompanying drawings:

[0020] Figure 1 It is a voltage-SOC curve under constant current discharge;

[0021] Figure 2 SOC depends on the voltage and current relationship surface. DETAILED DESCRIPTION

[0022] The technical solution of the present invention is further described below through specific implementation methods, but the protection scope of the present invention is not limited to the following implementation methods.

[0023] Implementation 1: This method for accurately calculating SOC based on battery model and least squares includes the following steps:

[0024] S1-determine the relationship between SOC and current and voltage, and obtain a mathematical data correspondence table of SOC and current and voltage;

[0025] S2- Determine the actual capacity Q using the least squares method c and the coulomb counter deviation Q0;

[0026] S3-Utilize Q c And the coulomb counter deviation Q0, according to Calculate the SOC value at any time t.

[0027] Furthermore, the SOC in S1 is determined to depend on the relationship between current and voltage and includes the following steps:

[0028] S11: Discharge the battery from full to empty at the specified current. Figure 1As shown, a specific discharge current sequence {I k}, k=1,…,n, for each specified current I k , observe that it is in a certain voltage sequence {V j}, j = 1, ..., m. Under the SOC value, the SOC value {SOC kj}, a mathematical two-dimensional m×n table is generated, where each k-th row represents the SOC value of the k-th current value in a series of voltage sequences, and the voltage SOC corresponding curve under the specified current is obtained;

[0029] S12: For each experimental current I k The SOC value at any voltage V is calculated by the voltage sequence (V j ,SOC kj ), (j=1,…,m) and perform piecewise linear interpolation to calculate, so as to obtain the current I k The SOC depends on the piecewise linear function SOC(I k ,V), where SOC(I k ,V j )=SOC kj .;

[0030] S13: Solve the variation law of SOC dependence on voltage under any current I and any voltage V, and calculate the SOC dependence on voltage under current I by the piecewise linear function in S12. k The value of SOC(I k ,V)=SOC k (V), over (I k ,SOC k (V)), k=1,2,…,m to do cubic spline interpolation, and get the SOC value SOC(I,V) under any current I and any voltage V. The function graph corresponds to a surface on the geometry. The SOC depends on the voltage and current relationship surface as follows: Figure 2 As shown;

[0031] S14: Set the voltage interval [V min ,V max ]N equal parts, current interval [I min ,I max ]M is divided equally, N, M are both relatively large, especially N>>n; the points on the grid (I k ,V j ),k=0,1,…,N;j=0,1,…,M. According to the above method, we can get the value of (I k ,V j ) SOC value SOC k,j, k=0,1,…,N;j=0,1,…,M., and obtain a precise mathematical data correspondence table of SOC dependence on current and voltage.

[0032] Furthermore, let the value of the coulomb meter at time t be Q t , then the SOC value at time t During a normal discharge process, a series of t0, t1, ..., t K Voltage at the moment Current and capacity j=0,1,…,K.,S2 is the actual capacity Q c The estimation of the coulomb counter deviation Q0 includes the following steps:

[0033] S21: Look up or calculate the SOC value from the table according to the data corresponding to the SOC dependence on current and voltage. The SOC value obtained by the conventional current integration method is The difference between the two is The sum of squared errors is

[0034] S22: Determine Q0,Q using the least squares method c , so that ψ(Q0,Q c ) is minimum, and the rated capacity Q is obtained c And the deviation Q0.

[0035] Implementation 2: Specifically for the MCU algorithm selection, the least square method is used to determine Q0, Q c In, order After the parameters q1 and q2 are calculated, the actual capacity Q is restored. c And the coulomb counter deviation Q0. Further, the single chip microcomputer adopts the integer data budget calculation for the least squares operation algorithm of S22, specifically, the rational number is used for the real data q1, q2 Approximation representation.

[0036] Since Q0,Q c The least squares problem of is a nonlinear least squares problem, so it is converted into a linear least squares problem. c As a fitting parameter, this is a nonlinear fitting. By cleverly replacing variables, let Then it becomes finding the best straight line SOC=q1Q t +q2 fitting. When the parameters q1 and q2 are obtained, the actual capacity Q can be restored. cAnd the coulomb counter deviation Q0 In addition, in order to adapt to the problem of small storage and computing power of the single-chip computer, the algorithm uses integer data budget to implement the calculation of the actual data. Specifically, for the real data q1, q2, rational numbers are used. To approximate the representation, save memory length and operation time, and improve calculation efficiency. The remaining steps are as described in embodiment 1 and will not be repeated.

[0037] This method for accurately calculating SOC based on a battery model and least squares overcomes the shortcomings of the existing voltage lookup table method and current integration method. It determines that SOC depends on the relationship between current and voltage, applies the least squares principle, combines the voltage lookup table method and the current integration method, estimates the actual capacity and deviation, and thus obtains a stable and accurate SOC value.

[0038] The above description shows the main features, basic principles, and advantages of the present invention. It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments or examples described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. The above embodiments or examples should therefore be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be encompassed within the present invention. Any reference numerals in the claims should not be construed as limiting the claims to which they relate.

[0039] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A method for accurately calculating SOC based on a battery model and least squares, characterized by: The method The following steps are included: S1-determine the relationship between SOC and current and voltage, and obtain a mathematical data correspondence table of SOC and current and voltage; this step specifically includes: S11-According to a set of specific discharge current sequences {I k }, k=1,…,n, discharge the battery throughout the entire process and record the current at each specified current I k Under a series of voltage sequences {V j }, j = 1, ..., m SOC value {SOC kj }; S12: For each experimental current I k , piecewise linear interpolation is used to calculate the SOC value at any voltage V, and the piecewise linear function SOC(I k ,V), where SOC(I k ,V j )=SOC kj ; S13: For any current I and any voltage V, by k ,SOC(I k ,V)), (k=1,…,n) perform cubic spline interpolation to obtain the SOC value SOC(I,V), thereby obtaining a data correspondence table in which the SOC depends on the current and voltage; S2- Determine the actual capacity Q using the least squares method c and the coulomb counter deviation Q0; S3-Utilize actual capacity Q c And the coulomb counter deviation Q0, according to Calculate the SOC value at any time t, where Q t is the measurement value of the coulomb meter at time t.

2. The method for accurately calculating SOC based on a battery model and least squares according to claim 1, wherein: Step S2 uses the least squares method to determine the actual capacity Q c The steps of estimating the coulomb counter deviation Q0 include: S21: During a normal discharge process, record a series of moments t j Voltage of (j=0,1,...,k) Current and coulombmeter measurements The SOC value is obtained by looking up or calculating the data corresponding table. The SOC value obtained by the current integration method is The sum of the squares of the two errors is S22: Determine Q0,Q using the least squares method c , so that ψ(Q0,Q c )minimum.

3. The method for accurately calculating SOC based on a battery model and least squares according to claim 2, wherein: In step S22, After determining the parameters q1, q2 by the linear least squares method, the actual capacity Q is restored. c And the coulomb counter deviation Q0.

4. The method for accurately calculating SOC based on a battery model and least squares according to claim 3, wherein: When the least squares method is performed using a single chip microcomputer, the real data q1, q2 are represented by rational numbers. Approximate representation for integer data operations.

Images