A real-time evaluation method for lithium battery SOC that takes into account both accuracy and cost

Abstract

The present invention relates to a real-time evaluation method for lithium battery SOC that takes into account both accuracy and cost. The method evaluates the SOC of a battery cell based on the ampere-hour integration (Ah) method and the current and voltage data of a working battery cell collected in real time. When the accumulated working time reaches the "ampere-hour integration working time", the double extended Kalman filter algorithm is used to correct the SOC obtained by the ampere-hour integration method. When the accumulated calibration time reaches the "Kalman working time", the calibration is stopped until the battery cell ends its operation. In the present invention, the error threshold is used to determine the ampere-hour integration working time (T Ah ), so the SOC real-time evaluation error will not exceed the error threshold specified in the project; due to the strong correction ability of the double-EKF algorithm, the Kalman working time T ekf Generally speaking, it is small, so compared with the single Ah method, the increase in calculation time is not large; it has good practical value in the application of SOC of underwater vehicle electric power system.

Description

Technical Field

[0001] The present invention belongs to a lithium battery SOC real-time evaluation method, and relates to a lithium battery SOC real-time evaluation method that takes both accuracy and cost into consideration. Background Art

[0002] Power lithium batteries have the characteristics of high safety, long cycle life, excellent power characteristics, low cost, and simple later maintenance. They are currently the most ideal power source for underwater vehicles used in industrial science. According to the state of charge (SOC) of the lithium battery, the battery management system reasonably allocates battery energy to ensure that the vehicle successfully completes the established mission. SOC represents the ratio of the available power of the battery in the current state to the available power of the battery in the fully charged state. This is a state quantity that cannot be directly measured. It needs to be evaluated by the corresponding algorithm.

[0003] Currently, among all SOC evaluation methods, the ampere-hour integral (Ah) method can well balance the accuracy of the method and the computational complexity, so it is quite popular in engineering applications. However, this method is an open-loop algorithm, which is easily affected by the external environment and measurement noise, and the evaluation error will gradually accumulate with the working time, resulting in a significant reduction in the accuracy of SOC evaluation at the end of discharge. The Kalman filter algorithm is a model-based SOC evaluation method and a closed-loop control algorithm. It can be adjusted in real time according to the evaluation error, so it has a higher prediction accuracy. However, correspondingly, the algorithm requires real-time evaluation of circuit model parameters, EKF update and calibration calculation steps, and the computational burden is heavy, which limits the application of this method in the engineering field.

[0004] Therefore, it is extremely urgent to establish a real-time SOC evaluation method for lithium batteries that takes into account both accuracy and cost based on existing methods and apply it to practical engineering. Summary of the invention

[0005] Technical issues to be solved

[0006] In order to avoid the shortcomings of the prior art, the present invention proposes a lithium battery SOC real-time evaluation method that takes both accuracy and cost into consideration.

[0007] Technical Solution

[0008] A lithium battery SOC real-time evaluation method taking into account both accuracy and cost is characterized by the following steps:

[0009] Step 1: Use the Ah method to evaluate the SOC of the battery cell in real time:

[0010]

[0011] Where: k and k-1 represent the experimental data sampling time points; i Lrepresents the sampling current of the working battery cell at time k-1; η represents the coulomb efficiency; Δt represents the data sampling time interval; C n Indicates the maximum available capacity of the battery;

[0012] Step 2: When the accumulated working time reaches the “ampere-hour integrated working time”, i.e. T Ah ", double-EKF algorithm is used to correct the SOC obtained by the ampere-hour integration method; when the calibration cumulative time reaches "Kalman working time T EKF ", stop calibration and return to step 1;

[0013] Step 3: During the working time of the battery cell, repeat steps 1 and 2 to evaluate the SOC of the lithium battery in real time.

[0014] The "ampere-hour integrated working time is T Ah ” and “Kalman working time T EKF The determination of " is as follows:

[0015] 1. Add random noise to the current data obtained from actual battery operation as a measure of T Ah and T EKF Current data of

[0016] 2. Use the noisy current data to evaluate the battery SOC in real time through the ampere-hour integration method, and compare it with the experimental results to calculate the deviation ΔSOC between the two;

[0017] 3. When ΔSOC>3%, record the cumulative time as T Ah_1 ; until the battery is discharged to the lower cut-off voltage, repeat steps 2 and 3, and record the time as T Ah_i

[0018] 4. Use the double-EKF algorithm instead of the Ah method to evaluate the battery SOC in real time according to the current data of the noise and the voltage data obtained from the actual operation of the battery, and compare it with the experimental results to calculate the deviation ΔSOC between the two;

[0019] 5. When ΔSOC < 1%, record the cumulative time as T EKF_1 ; Repeat steps 4 and 5 until the battery is discharged to the lower cut-off voltage, and record the time as T EKF_i ;

[0020] 6. To ensure the accuracy of SOC prediction, take T Ah_i The smallest value among them is taken as T Ah , take T EKF_i The largest value among them is taken as T EKF .

[0021] The random noise adopts a random normal distribution noise with a mean of 0 and a variance of 0.5A.

[0022] The coulombic efficiency η is 1.

[0023] The sampling time frequency value is 1.

[0024] The double-EKF algorithm uses two extended Kalman filter algorithms to respectively identify the battery model parameters and the current SOC of the battery in real time; wherein the battery model parameters use a first-order RC equivalent circuit model.

[0025] The ampere-hour integrated working time T Ah The deviation between the SOC evaluation value and the true value is greater than 3%.

[0026] The Kalman working time T EKF The deviation between the SOC evaluation value and the true value is less than 1%.

[0027] Beneficial Effects

[0028] The present invention proposes a lithium battery SOC real-time evaluation method that takes into account both accuracy and cost. The method evaluates the SOC of a battery cell based on the ampere-hour integration (Ah) method and the real-time collection of current and voltage data of a working battery cell. When the accumulated working time reaches the "ampere-hour integration working time", the double extended Kalman filter algorithm is used to correct the SOC obtained by the ampere-hour integration method. When the accumulated calibration time reaches the "Kalman working time", the calibration is stopped until the battery cell ends its operation.

[0029] The beneficial effects are as follows:

[0030] 1. The present invention uses the error threshold to determine the ampere-hour integrated working time (T Ah ), so the SOC real-time evaluation error will not exceed the error threshold specified by the project;

[0031] 2. In the present invention, due to the strong correction capability of the double-EKF algorithm, the Kalman working time T ekf It is generally small, so the increase in calculation time is not large compared to the Ah method alone;

[0032] 3. The present invention has good practical value in the application of SOC of the electric power system of underwater vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 Schematic diagram of the method of the present invention

[0034] Figure 2 T in the present invention Ah and T EKF Determine method flow chart DETAILED DESCRIPTION

[0035] The present invention will now be further described with reference to the embodiments and the accompanying drawings:

[0036] A real-time evaluation method for lithium battery SOC that takes into account both accuracy and cost includes the following steps:

[0037] Step 1: Collect and record the current and voltage data of the working battery cells in real time;

[0038] Step 2: Evaluate the SOC of the battery cell in real time based on the ampere-hour integration (Ah) method;

[0039] Step 3: When the accumulated working time reaches the “ampere-hour integrated working time (T Ah )”, double-extended Kalman filter algorithm (double-EKF) is used to correct the SOC obtained by the ampere-hour integration method; when the calibration cumulative time reaches the “Kalman working time (T EKF )”, the calibration stops;

[0040] Step 4: Repeat steps 2 and 3 until the battery cell is fully operational.

[0041] In step 1, the acquisition frequency of the battery cell voltage and current is fixed, and the frequency value is 1.

[0042] In step 2, the method for real-time evaluation of battery SOC based on the ampere-hour integration method is as follows:

[0043]

[0044] In the formula,

[0045] k and k-1 represent the sampling time points of experimental data;

[0046] i L represents the current at time k-1;

[0047] η represents the Coulomb efficiency, which is 1 here;

[0048] Δt represents the data sampling time interval;

[0049] C n Indicates the maximum available battery capacity.

[0050] In step 3, double-EKF means using two extended Kalman filter algorithms to respectively identify the battery model parameters and the current SOC of the battery in real time. Among them, the battery model parameters use a first-order RC equivalent circuit model.

[0051] In step 3, the “ampere-hour integrated working time (T Ah )” and “Kalman working time (T EKFThe method for determining ")" is determined by the project accuracy threshold. In the present invention, when the deviation between the SOC evaluation value and the true value is greater than 3% (maximum error), the double-EKF algorithm needs to be used for calibration; when the deviation between the SOC evaluation value and the true value is less than 1% (negligible error value), the error is considered negligible, and the ampere-hour integration method can continue to be used to estimate the SOC of the battery cell in real time. Determine T Ah and T EKF The specific process is as follows.

[0052] 1) To increase the reliability of the calibration time, a certain degree of random noise is added to the current data obtained from the actual operation of the battery as a factor in determining T Ah and T EKF The present invention uses random normal distribution noise with a mean of 0 and a variance of 0.5A.

[0053] 2) Based on the current data with noise added in i, the battery SOC is evaluated in real time by the ampere-hour integration method, and compared with the experimental results to calculate the deviation ΔSOC between the two;

[0054] 3) When ΔSOC>3%, record the cumulative time as T Ah_1 ;

[0055] 4) The double-EKF algorithm is used instead of the Ah method. The battery SOC is evaluated in real time according to the current data with noise added in i and the voltage data obtained from the actual operation of the battery. The results are compared with the experimental results to calculate the deviation ΔSOC between the two.

[0056] 5) When ΔSOC < 1%, record the cumulative time as T EKF_1 ;

[0057] 6) Repeat steps ii and iii, and record the time as T Ah_2 ;

[0058] 7) Repeat steps iv, v, and record the time as T EKF_2 ;

[0059] 8) Repeat in sequence and record T Ah_1 , T EKF_1 , T Ah_2 , T EKF_2 , ... until the battery is discharged to the lower cut-off voltage;

[0060] To ensure the SOC prediction accuracy, take T Ah_i The smallest value among them is taken as T Ah , take T EKF_i The largest value among them is taken as T EKF .

Claims

1. A real-time evaluation method for lithium battery SOC that takes into account both accuracy and cost, characterized in that Here are the steps: Step 1: Use the Ah method to evaluate the SOC of the battery cell in real time: Where: k and k-1 represent the experimental data sampling time points; i L, k-1 represents the sampling current of the working battery cell at time k-1; η represents the coulomb efficiency; Δt represents the data sampling time interval; C n Indicates the maximum available capacity of the battery; Step 2: When the accumulated working time reaches "Ah integral working time (T) Ah ", double-EKF algorithm is used to correct the SOC obtained by the ampere-hour integration method; when the cumulative calibration time reaches "Kalman working time T EKF ", stop calibration and return to step 1; The "ampere-hour integrated working time" is T Ah " and "Kalman working time T EKF The determination of T is as follows:

1. Add random noise to the current data obtained from the actual operation of the battery as a reference to determine T Ah and T EKF Current data of 2. Use the noisy current data to evaluate the battery SOC in real time through the ampere-hour integration method, and compare it with the experimental results to calculate the deviation ΔSOC between the two; 3. When ΔSOC>3%, record the cumulative time as T Ah_1 ; until the battery is discharged to the lower cut-off voltage, repeat steps 2 and 3, and record the time as T Ah_i 4. Use the double-EKF algorithm instead of the Ah method to evaluate the battery SOC in real time according to the current data of the noise and the voltage data obtained from the actual operation of the battery, and compare it with the experimental results to calculate the deviation ΔSOC between the two; 5. When ΔSOC < 1%, record the cumulative time as T EKF_1 ; Repeat steps 4 and 5 until the battery is discharged to the lower cut-off voltage, and record the time as T EKF_i ; 6. To ensure the accuracy of SOC prediction, take T Ah_i The smallest value among them is taken as T Ah , take T EKF_i The largest value among them is taken as T EKF ; Step 3: During the working time of the battery cell, repeat steps 1 and 2 to evaluate the SOC of the lithium battery in real time.

2. The lithium battery SOC real-time evaluation method taking into account both accuracy and cost according to claim 1 is characterized by: The random noise adopts a random normal distribution noise with a mean of 0 and a variance of 0.5A.

3. The lithium battery SOC real-time evaluation method taking into account both accuracy and cost according to claim 1 is characterized by: The coulombic efficiency η is 1.

4. The lithium battery SOC real-time evaluation method taking into account both accuracy and cost according to claim 1 is characterized by: The sampling time frequency value is 1.

5. The lithium battery SOC real-time evaluation method taking into account both accuracy and cost according to claim 1, characterized in that: The double-EKF algorithm uses two extended Kalman filter algorithms to respectively identify the battery model parameters and the current SOC of the battery in real time; wherein the battery model parameters use a first-order RC equivalent circuit model.

6. The lithium battery SOC real-time evaluation method taking into account both accuracy and cost according to claim 1, characterized in that: The ampere-hour integrated working time T Ah The deviation between the SOC evaluation value and the true value is greater than 3%.

7. The lithium battery SOC real-time evaluation method taking into account both accuracy and cost according to claim 1, characterized in that: The Kalman working time T EKF The deviation between the SOC evaluation value and the true value is less than 1%.

Images