Lithium battery charge state estimation method, battery management system and storage medium

Abstract

The invention provides a lithium battery charge state estimation method, a battery management system and a storage medium, and relates to the field of lithium battery data processing. The method comprises the following steps: firstly, acquiring state data and open-circuit voltage data of each charge-discharge cycle of the lithium battery, and constructing an equivalent RC model of the lithium battery; predicting the state of charge of the lithium battery by using the constructed lithium battery RC model, and correcting the prediction result of the state of charge by using an extended Kalman filtering algorithm according to the relationship between the open-circuit voltage and the state of charge; the method comprises the following steps: acquiring a relationship between an initial charge state and a charge state at the current moment through a form of extending along with time of a state-space equation, and then calculating an unbiased expectation of the initial charge state based on a charge state estimation value at the current moment by means of an inverse function; and finally, based on the unbiased expectation of the initial state of charge, executing an extended Kalman filtering algorithm to smooth the state of charge. According to the method, the estimation precision of the state of charge is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of lithium battery data processing, and in particular to a lithium battery state of charge estimation method, a battery management system and a storage medium. Background Art

[0002] Lithium-ion batteries have received widespread attention in the energy field due to their performance advantages such as high energy density, many available cycles and less environmental pollution. In the recycling process of lithium batteries, the state of charge is a key indicator for evaluating the battery usage effect. In the actual energy storage power station or household new energy vehicles, in the process of lithium battery charging and discharging, in order to improve efficiency, most of them choose the fast charging mode, that is, the charging is immediately terminated when the specified voltage is reached through constant current charging, rather than continuing to use a small current charging method to charge to 100% state of charge after the charging reaches the constant voltage stage. Therefore, the initial value of the state of charge of this type of charging and discharging process is not 100%, so it is very important to accurately obtain the initial value of the lithium battery state of charge.

[0003] In setting the initial state of charge value for the lithium battery charging and discharging process, the equivalent RC model is generally used to estimate the current state of charge of the lithium battery based on the initial state of charge value set in advance. This method regards the inside of the lithium battery as a combination of multiple groups of resistors and capacitors, and models the characteristic variables such as current and voltage outside the battery, which has great advantages in computing efficiency and cost. However, common modeling methods cannot solve the problem of inaccurate initial state of charge caused by constant current charging.

[0004] Although the method based on the equivalent RC model can efficiently establish the state of charge estimation model, a special algorithm is required to smooth the initial value of the state of charge. Using the traditional nonlinear Kalman iterative filtering strategy, the setting of the initial state of charge can only be obtained based on prior information, and its accuracy can only be judged based on the initial variance. The variance is a statistic that targets the average error of multiple initial value statistical results and cannot accurately reflect the error of a single initial value. Summary of the invention

[0005] Purpose of the invention: To propose a method for estimating the state of charge (SOC) of a lithium battery, and further propose a battery management system (BMS) for executing the SOC estimation method, so as to solve the above-mentioned problems existing in the prior art.

[0006] The present invention provides a method for estimating the state of charge of a lithium battery, comprising the following steps:

[0007] S1. Construct a state space equation according to the parameters of the lithium battery equivalent RC model, and estimate the current state of charge based on the state space model;

[0008] S2. Execute an extended Kalman filter algorithm according to the estimated state of charge at the current moment and the state space equation to calculate an unbiased expectation of the initial value of the state of charge;

[0009] S3, using the unbiased expected value of the initial value of the state of charge, continuing to execute the extended Kalman filter algorithm until the current moment; smoothing the estimated value of the state of charge from the initial moment to the current moment;

[0010] S4. Taking the state of charge estimation result based on the unbiased estimation of the initial value at the current moment as the most reasonable estimation value at the current moment, and executing the state of charge estimation again at the next moment.

[0011] In a further embodiment, the lithium battery equivalent RC model equates the internal structure of the lithium battery to the power source U OC 、Internal resistance R 0 And two RC combinations R 1 C 1 , R 2 C 2 The form of series combination satisfies the following equation:

[0012]

[0013] U L (t) = U OC (t)-U 1 (t)-U 2 (t)-R 0 I(t)

[0014] Where U L (t) is the open circuit voltage; I(t) is the main circuit current; Represents the divided voltage of the first group of capacitors and resistors in parallel; Indicates the partial voltage of the second group of capacitors and resistors in parallel; U OC (t) represents the discharge voltage; see Figure 1 .

[0015] According to the experimental measurement results, the open circuit voltage U OC The fitting curve with the state of charge SOC can be described as the following high-order polynomial:

[0016] U OC =a 6 SOC 6 +a 5 SOC 5 +a 4 SOC 4 +a 3 SOC 3 +a 2 SOC 2 +a 1SOC+a 0

[0017] In the formula, parameter a 0 ~a 6 Obtained by fitting the experimental measurement data.

[0018] According to the above relationship, the state space equation can be listed as follows:

[0019]

[0020] C = [0 -1-1]

[0021] D=R 0

[0022] Where △t represents the interval time, w(k) represents the system noise signal at time k, v(k) represents the measurement noise signal at time k, and u(k) represents the input current at time k.

[0023] In a further embodiment, before calculating the unbiased expectation of the initial value of the state of charge, the state space equation is converted into the following extended state space equation form:

[0024]

[0025] Where x(0) is the initial value of the state; F(k-1), G(k-1), G w2 (k-1) is the state space model parameter at time k-1 after expansion.

[0026] In a further embodiment, the unbiased expectation of the extended state at time k is The unbiased expectation of the initial state Written as the following equation:

[0027]

[0028] At this point, the initial value based on the current state estimation result is calculated

[0029] In a further embodiment, step S4 specifically includes:

[0030] After the lithium battery state of charge completes the smoothing of the variational initial value at each moment, the extended Kalman filter is used again to obtain the current state estimation result with smaller error, and then the state of charge at the next moment is estimated based on the state estimation result with smaller error, and then steps S2 to S3 are repeated to calculate the unbiased expectation of the initial value based on the state of charge estimation value at the next moment, and the extended Kalman filter is repeatedly executed to obtain the state estimation value at the next moment.

[0031] In addition, the present invention also provides a battery management system, which includes:

[0032] At least one monitoring module, used to monitor the current charging and discharging information of the lithium battery in real time;

[0033] At least one calculation module, used to read the current lithium battery charge and discharge information and calculate the lithium battery state of charge;

[0034] At least one monitoring module and at least one computing module are packaged on a mainboard; a memory for storing computer program instructions is also provided on the mainboard; and the above-mentioned lithium battery state of charge estimation method is implemented when the computing module executes the computer program instructions.

[0035] In addition, the present invention also proposes a computer-readable storage medium, in which at least one executable instruction is stored. When the executable instruction is executed on an electronic device, the electronic device executes the above-mentioned lithium battery state of charge estimation method.

[0036] Beneficial effects:

[0037] Compared with the traditional extended Kalman filter algorithm, the variational initialization technology is introduced. After the current state of charge estimation is completed, the functional relationship between the initial state of charge and the current state of charge is obtained by expanding the state space equation to time, and then the unbiased expectation of the initial state of charge based on the current state of charge estimation value is solved with the help of the inverse function matrix. Then, the nonlinear Kalman filter algorithm is used to smooth the state of charge estimation value from the initial moment to the current moment based on the unbiased expectation of the initial state of charge. Before the iterative nonlinear Kalman filter algorithm starts, the setting of the initial state of charge only considers the prior information, while during the execution of the iterative filtering algorithm, the information from the observed voltage is introduced, so as to improve the accuracy of the initial value setting through iterative smoothing. The applicability of this method has been verified in the experiment.

[0038] The variational initialization technology is introduced to reduce the impact of inaccurate model parameters and poor state of charge estimation caused by inaccurate initial value setting. The extended Kalman filter algorithm that introduces variational initialization technology can effectively overcome the problem of inaccurate initial value setting caused by fast charging, improve the accuracy and robustness of the state of charge estimation at the current moment, and solve the problem of inaccurate state of charge estimation of lithium batteries caused by inaccurate initial values. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is an execution flow chart of the lithium battery charge state estimation method of the present invention.

[0040] Figure 2 It is a schematic diagram of a model for predicting the state of charge based on an equivalent n-th order RC model of the present invention.

[0041] Figure 3 4 is a graph showing the relationship between the open circuit voltage and the state of charge of a lithium battery provided by an embodiment of the present invention.

[0042] Figure 4 This is a comparison chart of the results of the state of charge estimation based on different initial state of charge setting values ​​under the condition of the lithium battery model of the present invention.

[0043] Figure 5 This is a comparison chart of the state of charge estimation error results under different initial state of charge setting values ​​based on the lithium battery model of the present invention.

[0044] Figure 6 It is a schematic diagram showing the application results of the present invention on an automobile example. DETAILED DESCRIPTION

[0045] In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is apparent to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features known in the art are not described.

[0046] This embodiment provides a lithium battery state of charge estimation method based on variational initialization and extended Kalman filtering. A high-order function is fitted based on the historical state of charge data and the relationship between the equivalent RC model parameters; the extended Kalman filter algorithm is applied to calculate the current state of charge estimate, and then the relationship between the initial state of charge and the current state of charge is obtained based on the extended form of the state space model for time, and the unbiased expectation of the initial state of charge based on the current state of charge estimate is calculated based on the inverse function matrix, and then the extended Kalman filter algorithm is repeated with the help of the unbiased expectation until the current moment to complete the smoothing of the state of charge time series. The present invention can effectively reduce the influence of unreasonable initial state of charge on the current state of charge estimation result through the iteration of the initial state of charge setting value of the lithium battery and the state of charge estimation smoothing method, and is a state of charge estimation method with the advantages of fastness and accuracy. Figure 2 It is a second-order equivalent RC model.

[0047] The lithium battery state of charge estimation method based on variational initialization and extended Kalman filtering proposed in this embodiment has a process as shown in Figure 1 , the specific steps are as follows:

[0048] S1. According to the system parameters of the lithium battery RC model, a state space model is constructed, and the state of charge is estimated based on this model; the equivalent RC model of the lithium battery is analyzed according to Kirchhoff's law to obtain the relationship between the open circuit voltage, current and various impedances, and then according to the functional relationship between the open circuit voltage and current in the RC circuit U L (t) = U OC (t)-∑ i＝1 U i (t)-R 0 I(t), and then describe the open circuit voltage as a nonlinear high-order function of the state of charge according to the fitting curve of the open circuit voltage and the state of charge, and then list the state space equation.

[0049] According to the equivalent principle of the battery Figure 2 It can be seen that the internal structure of the lithium battery is equivalent to the power supply U OC With internal resistance R 0 And two RC combinations R 1 C 1 , R 2 C 2 In the form of a series combination, according to Kirchhoff's law, the equivalent RC model satisfies the following equation:

[0050]

[0051]

[0052] U L (t) = U OC (t)-U 1 (t)-U 2 (t)-R 0 I(t) (3)

[0053] Among them U L (t) is the open circuit voltage, I(t) is the main circuit current, U 1 (t), U 2 (t) is the voltage of the two RC combinations, and according to experimental measurement results, the open circuit voltage U LC The fitting curve with the state of charge SOC can be described as the following high-order polynomial:

[0054] U OC =a 6 SOC 6 +a 5 SOC 5 +a 4 SOC 4 +a 3 SOC 3 +a 2 SOC 2 +a1 SOC+a 0

[0055] The corresponding parameter a 0 ~a 6 It can be obtained by fitting based on experimental measurement data.

[0056] According to the above relationship, the state space equation can be listed as follows:

[0057]

[0058] in C=[0 -1-1],D=R 0 .

[0059] S2, calculating the unbiased expectation of the initial value of the state of charge based on the estimated value of the state of charge at the current moment and the state space model parameters;

[0060] After estimating the state of charge of the lithium battery based on the lithium battery RC state space model using the extended Kalman filter algorithm, the initial state expansion form of the state space equation is used to show the impact of the initial value of the state of charge on the current moment. Then, according to the expansion form and the inverse function matrix of the state space model, the unbiased expected value of the initialized state of charge based on the estimated value at the current moment is calculated.

[0061] To calculate the initial value of the state of charge, it is necessary to use the extended Kalman filter algorithm; before using the extended Kalman filter algorithm, the above recursive state space equation needs to be converted into the following extended state space equation form:

[0062]

[0063] Where x(0) is the initial value of the state, and the above extended state vector is:

[0064] X(k)=[x(1) T ,x(2) T ,...,x(k) T ] T

[0065] U(k-1)=[u(0) T ,u(1) T ,...,u(k-1) T ] T

[0066] Y(k)=[y(1),y(2),...,y(k)] T

[0067] The parameters of the corresponding extended state-space model are:

[0068]

[0069]

[0070] H(k)=C(k)F(k-1),L(k)=C(k)G(k-1),G w2 (k)=C(k)G w1 (k-1).

[0071] According to the above extended model, the unbiased expectation of the initial value of the state of charge at time k can be calculated by using the posterior x(k|k) of the state estimation result at time k and the inverse of the state model parameter F(k-1). The specific formula is as follows:

[0072] According to formula (5), the unbiased expectation of the expanded state at time k and the unbiased expectation of the state at the initial time can be written as follows:

[0073]

[0074] So we have:

[0075]

[0076] At this point, the initial value based on the current state estimation result can be calculated

[0077] S3, with other parameters unchanged, re-execute the sequential estimation of the state of charge using the unbiased expected value of the initial value of the state of charge until the current moment;

[0078] Using the unbiased expectation of the initial state of charge, under the condition that other parameters remain unchanged, the extended Kalman filter algorithm is re-executed until the current moment, the state of charge estimate from the initial moment to the current moment is smoothed, and the current moment state of charge estimate after smoothing based on the unbiased expectation of the initial state of charge is used as a more reasonable current moment estimate.

[0079] S4. Taking the state of charge estimation result based on the unbiased estimation of the initial value at the current moment as the most reasonable estimation value at the current moment, and executing the state of charge estimation again at the next moment.

[0080] After the lithium battery state of charge completes the smoothing of the variational initial value at each moment, the extended Kalman filter can be used again to obtain the current state estimation result with smaller error. Then, based on the state estimation result with smaller error, the state of charge at the next moment is estimated. The above steps are repeated to calculate the unbiased expectation of the initial value based on the state of charge estimation value at the next moment, and the extended Kalman filter is repeatedly executed to obtain the state estimation value at the next moment.

[0081] The following is an example to illustrate the process and accuracy of the extended Kalman filter algorithm based on variational thinking proposed in the present invention for estimating the state of charge of a lithium battery.

[0082] The example uses four 18650 lithium iron phosphate (LiFePO 4 ) The battery was subjected to a battery state of charge charge and discharge cycle experiment. The rated capacity was 2400mAh, and the initial discharge capacity was 2365mAh, 2390mAh, 2338mAh, and 2349mAH respectively. Constant current charging was performed at 2.4A. When the battery terminal voltage rose to 4.2V, constant current charging was considered to be completed. After 2 minutes of storage, constant current discharge was performed at 2.4A until the lithium battery terminal voltage dropped to 3V, and discharge was completed. Among them, Figure 3 It is the relationship between the state of charge and the open circuit voltage of the lithium battery.

[0083] The specific steps are as follows:

[0084] (1) Based on the historical fitting function of the lithium battery voltage measurement value and the state of charge, the state of charge prediction value obtained according to the n-order RC model is corrected using the measured open circuit voltage to obtain the estimated result of the state of charge at the current moment.

[0085] (2) After estimating the state of charge of the lithium battery based on the RC state space model of the lithium battery using the extended Kalman filter algorithm, the initial state expansion form of the state space equation is used to show the impact of the initial value of the state of charge on the current moment. Then, based on the expansion form and the inverse function matrix of the state space model, the unbiased expected value of the initial state of charge based on the estimated value at the current moment is calculated.

[0086] (3) Using the unbiased expectation of the initial state of charge, and under the condition that other parameters remain unchanged, re-execute the extended Kalman filter algorithm until the current moment, smooth the estimated state of charge from the initial moment to the current moment, and use the current state of charge estimated value smoothed based on the unbiased expectation of the initial state of charge as a more reasonable current moment estimated value. After the lithium battery state of charge completes the smoothing of the variational initial value at each moment, the extended Kalman filter is used again to obtain the current moment state estimation result with a smaller error, and then the state of charge at the next moment is estimated based on the state estimation result with a smaller error, and the above steps are repeated to calculate the initial value unbiased expectation based on the state of charge estimate at the next moment, and the extended Kalman filter is repeatedly executed to obtain the state estimation value at the next moment.

[0087] The comparison of the estimation results and the error of the experiment are shown in Figure 4 , Figure 5 As shown. Figure 4 (a) is the estimated result using a relatively inaccurate initial state of charge setting value, Figure 4(b) shows the estimation results based on different initial state of charge settings. Figure 5 (a) is the estimated error using a less accurate initial SOC setting, Figure 5 (b) is the estimation error based on different initial state of charge setting values. This experiment uses mean absolute error (MAE) and mean square error (MSE) as evaluation indicators for state of charge estimation. The comparison results of battery state of charge estimation errors of the extended Kalman filter algorithm under different initial value conditions are shown in Table 1.

[0088] Table 1: Comparison of battery state of charge estimation errors based on the extended Kalman filter algorithm under different initial value conditions

[0089]

[0090]

[0091] Figure 6 The present invention is applied to an automobile example, wherein (a) is the result of the state of charge estimation based on different initial values, and (b) is the estimation error of the state of charge based on different initial values.

[0092] From the error comparison of the lithium battery health state estimation results, it can be seen that the method proposed in the present invention proves the influence of inaccurate initial value on the state of charge estimation results. At the same time, the introduction of variational initialization method proposed in the present invention has a better initial state of charge setting effect, thereby improving the overall state of charge estimation accuracy, and has the characteristics of more accurate initial value and reduced estimation error.

[0093] In summary of the above embodiments, the lithium battery state of charge estimation method based on variational initialization and extended Kalman filtering proposed in the present invention updates the system initial state of charge setting value by introducing the state of charge estimation result of the current period of time. This iterative method improves the accuracy of the system initial value, thereby ensuring the accuracy of the model parameters. As the system model becomes more accurate, the state of charge estimation result is closer to the actual result. The improved algorithm has greatly improved the accuracy of state of charge estimation, and overcomes the problem of inaccurate system model parameters caused by inaccurate initial state of charge setting of the battery when using a constant current charging method, which leads to inaccurate state of charge estimation. It has great guiding significance and reference value.

[0094] The lithium battery state of charge estimation method disclosed in the above embodiment can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiment can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. The computer program is written to a storage medium and runs on a battery management system. The battery management system includes a monitoring module for real-time monitoring of the current lithium battery charge and discharge information, and a computing module for reading the current lithium battery charge and discharge information and calculating the lithium battery state of charge; at least one monitoring module and at least one computing module are encapsulated on the mainboard. The battery management system can automatically execute the lithium battery state of charge estimation method disclosed in the above embodiment, which will not be repeated here.

[0095] When the computer instruction or computer program is loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instruction can be stored in a computer-readable storage medium, or transmitted from a computer-readable storage medium to another computer-readable storage medium. For example, the computer instruction can be transmitted from a website site, a computer, a server, or a data center to another website site, a computer, a server, or a data center by a wired (e.g., infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server, a data center, etc. that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0096] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0097] The matters not described in detail in the present invention are all known technologies to those skilled in the art.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for estimating the state of charge of a lithium battery, characterized in that: include: S1. Construct a state space equation according to the parameters of the lithium battery equivalent RC model, and estimate the current state of charge based on the state space model; S2. Execute an extended Kalman filter algorithm according to the estimated state of charge at the current moment and the state space equation to calculate an unbiased expectation of the initial value of the state of charge; S3, using the unbiased expected value of the initial value of the state of charge, continuing to execute the extended Kalman filter algorithm until the current moment; Smoothing the estimated state of charge from the initial moment to the current moment; S4. Taking the state of charge estimation result based on the unbiased estimation of the initial value at the current moment as the most reasonable estimation value at the current moment, and executing the state of charge estimation again at the next moment.

2. The method for estimating the state of charge of a lithium battery according to claim 1, characterized in that: The internal structure of the lithium battery is equivalent to the power supply U OC , internal resistance R0 and two RC combinations R1C1 and R2C2 in series; the equivalent RC model of the lithium battery satisfies the following equation: U L (t)=U OC (t)-U1(t)-U2(t)-R0I(t) Where U L (t) is the open circuit voltage; I(t) is the main circuit current; U1(t) and U2(t) are the voltages of the two RC combinations; Represents the divided voltage of the first group of capacitors and resistors in parallel; Indicates the partial voltage of the second group of capacitors and resistors in parallel; U OC (t) represents the discharge voltage.

3. The method for estimating the state of charge of a lithium battery according to claim 2, characterized in that: Open circuit voltage U OC The fitting curve with the state of charge SOC can be described as the following high-order polynomial: You OC =a6SOC 6 +a5SOC 5 +a4SOC 4 +a3SOC 3 +a2SOC 2 +a1SOC+a0 Wherein, parameters a0~a6 are obtained by fitting the experimental measurement data.

4. The method for estimating the state of charge of a lithium battery according to claim 3, characterized in that: The state space equation is as follows: In the formula, w(k) represents the system noise at time k; v(k) represents the measurement noise at time k; u(k) represents the input current at time k; C＝[0-1-1] D=R0 In the formula, △t represents the interval time.

5. The method for estimating the state of charge of a lithium battery according to claim 4, characterized in that: Before calculating the unbiased expectation of the initial value of the state of charge, the state space equation is converted into the following extended state space equation form: Where x(0) is the initial value of the state; F(k-1), G(k-1), G w2 (k-1) is the state space model parameter at time k-1 after expansion.

6. The method for estimating the state of charge of a lithium battery according to claim 5, characterized in that: Unbiased expectation of extended state at time k The unbiased expectation of the initial state Written as the following equation: At this point, the initial value based on the current state estimation result is calculated 7. The method for estimating the state of charge of a lithium battery according to claim 1, characterized in that: Step S4 specifically includes: After the lithium battery state of charge completes the smoothing of the variational initial value at each moment, the extended Kalman filter is used again to obtain the current state estimation result with smaller error, and then the state of charge at the next moment is estimated based on the state estimation result with smaller error, and then steps S2 to S3 are repeated to calculate the unbiased expectation of the initial value based on the state of charge estimation value at the next moment, and the extended Kalman filter is repeatedly executed to obtain the state estimation value at the next moment.

8. A battery management system, characterized in that: include: At least one monitoring module, used to monitor the current charging and discharging information of the lithium battery in real time; At least one calculation module, used to read the current lithium battery charge and discharge information and calculate the lithium battery state of charge; At least one monitoring module and at least one computing module are encapsulated on a mainboard; the mainboard is also provided with a memory for storing computer program instructions; when the computing module executes the computer program instructions, the lithium battery state of charge estimation method as described in any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The storage medium stores at least one executable instruction. When the executable instruction is executed on the electronic device, the electronic device executes the lithium battery state of charge estimation method according to any one of claims 1 to 6.

Images