Judgment method for automatically determining voltage drop and K value quasi value of battery cell

Abstract

The invention relates to the technical field of battery cell detection, and discloses a determination method for automatically determining a battery cell voltage drop and a K value standard value, which comprises the following steps of: 1, importing a fitting curve function of the battery cell voltage drop and the K value standard value; step 2, measuring the current battery cell voltage V1, and recording the measurement time T1; step 3, obtaining the initial voltage V0 and the test time T0 of the battery cell; calculating the voltage difference value delta V and the storage time delta T of the battery cell; step 4, acquiring a voltage drop standard value Y according to the delta T by fitting a curve function, and judging whether the battery cell is qualified or not; and step 5, obtaining a standard value and an actual value of the K value according to the delta V and the delta T by fitting a curve function, and judging whether the battery cell is qualified or not. According to the invention, more consistent standard values of the voltage drop and the K value can be automatically calculated, repeated measurement and calculation are avoided, generated errors and errors are effectively avoided, automatic calibration and detection are realized, the product quality and the detection efficiency are improved, and the production is more intelligent.

Description

Technical Field

[0001] The present invention relates to the technical field of battery cell detection, and particularly to a determination method for automatically determining the voltage drop of a battery cell and the reference value of the K value. Background Art

[0002] During the battery production process, the quality control of single battery cells is a core indicator to ensure battery safety. Especially before assembly production, it is necessary to detect the voltage of single battery cells to ensure that they meet the production standards. Currently, the quality inspection process requires operators to calculate and set the corresponding voltage drop and K value standards according to the storage time of the battery cells. However, the voltage drop and K value of the battery cells do not have a simple proportional relationship with the storage time. In the initial stage of storage, the voltage drop and K value change significantly, and as the storage time prolongs, the change gradually tends to be stable. Therefore, to facilitate calculation, a method of dividing fixed cycle durations is usually adopted to gradiently formulate the standards for voltage drop and K value. For example, the corresponding standard values are set with fixed cycles such as 5 days, 10 days, 20 days, 30 days, etc.

[0003] Although this method can help operators quickly set the standard values during operation to detect whether the battery cells are qualified. However, before each batch of battery cells is put into production, operators need to reset the standard values according to the current storage time of the battery cells, which is inefficient and there is also a risk of setting errors. Moreover, since the current standard values are set in a gradient form according to fixed cycle durations, although the change tends to be stable as the storage time prolongs, within the same gradient cycle (such as battery cells with a storage time between 30 days and 40 days), there may be significant differences in the actual performance of the battery cells on the first day and the last day. This results in that during detection, some unqualified battery cells cannot be effectively screened out and thus flow into the production line, affecting the quality and safety of the final product. Summary of the Invention

[0004] The present invention aims to provide a determination method for automatically determining the voltage drop of a battery cell and the reference value of the K value to solve the problems of low efficiency and high risk of false detection in the existing battery cell quality inspection.

[0005] To achieve the above object, the present invention adopts the following technical solutions. A determination method for automatically determining the voltage drop of a battery cell and the reference value of the K value includes the following steps:

[0006] Step 1, import the fitting curve function of the voltage drop standard value and K value of the battery cell;

[0007] Step 2, measure the current voltage V1 of the battery cell and record the measurement time T1;

[0008] Step 3, obtain the initial voltage V0 of the battery cell and the test time T0; calculate the voltage difference ΔV of the battery cell and the storage time ΔT;

[0009] Step 4: Obtain the standard value Y of the pressure drop according to ΔT through the fitting curve function, and determine whether the battery cell is qualified.

[0010] Step 5: Obtain the standard value and actual value of the K value according to ΔV and ΔT through the fitting curve function, and determine whether the battery cell is qualified.

[0011] The principle and advantages of this solution are as follows:

[0012] Based on the self-discharge characteristics of battery cells, this solution measures the voltage change of the battery cells during the storage time, calculates the K value (pressure drop rate) to evaluate the self-discharge performance of the battery cells, and then evaluates the performance indicators of individual battery cells to achieve the purpose of automated detection.

[0013] The existing test requirements need to set the standard values of the pressure drop and the K value first, and determine whether the battery cells meet the requirements by comparing with the standard values. To improve the detection efficiency and facilitate the operator to quickly set the standard values before each batch of battery cells are put on the line, the conventional method is to divide them according to the storage time of the battery cells at a fixed cycle and formulate a stepped standard value accordingly, so that the operator can quickly make a comparison and set them.

[0014] Although this method greatly improves the setting efficiency of the operator, there are also disadvantages, and these disadvantages continue to emerge in the actual production detection, resulting in the outflow of unqualified products. Therefore, it is difficult to improve the product qualification rate, and even directly leads to product scrapping. More seriously, it directly flows into the client, causing a market impact.

[0015] Regarding the problem that the product qualification rate is difficult to break through reflected at the result end, how to identify unqualified products during detection and effectively avoid unqualified products from flowing into the production line has become a difficult problem in the detection operation. No matter how the detection accuracy or the number of rechecks of the detection personnel is improved, this problem cannot be effectively solved.

[0016] After multi-directional analysis and research, we determined the technical problem to be the difference in the standard values, rather than the omissions in the detection process or the detection method. Thus, it was realized that the existing standard value setting is not exact. With the continuous update of battery technology and the continuous changes in the environment and requirements, the previous "qualified products" may no longer meet the current technical requirements. As a result, even if we set and detect according to the existing standard value setting method, they are presented as unqualified at the product end, ultimately affecting the product qualification rate. Through in-depth analysis, it was found that even though the storage time was divided, the difference between the battery cells at the two ends of a period of time was still relatively large. Under the current technical requirements, this difference can no longer be ignored, and it will also affect the quality of the battery cells. However, if the same standard value is still used for detection, it will lead to misjudging unqualified products as qualified.

[0017] For the problems found, this solution abandons the existing periodic and staged fast calibration method, but sets a more suitable functional relationship. Although the calculation is more complex, it can achieve automatic calculation, greatly reducing the workload of testers and ensuring accuracy and precision. By directly importing the fitting curve function into the test equipment, this solution can automatically obtain the current voltage and test time of the battery cell, as well as the initial voltage and initial test time, so as to automatically calculate the corresponding difference. Combining with the imported fitting curve function, it can automatically obtain the standard values of the fitting voltage drop and K value, thus no longer needing to set the corresponding standard values according to the segmented period form, avoiding repeated measurement and calculation by operators, and effectively avoiding the generated errors and mistakes, realizing automatic calibration and detection, and no longer requiring operators to participate in the determination and setting every time they go online, improving the accuracy and efficiency of battery cell detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic flow framework diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The following is further detailed through specific embodiments:

[0020] Embodiment 1

[0021] In this embodiment, the determination method for automatically determining the standard values of the voltage drop and K value of the battery cell imports the fitting curve functions of the voltage drop and K value of the battery cell, automatically matches the corresponding standard values, improves the accuracy of setting the standard values, reduces operation interference, improves the detection efficiency, and reduces the unqualified rate. In this embodiment, as shown in the attached Figure 1 figures, it includes the following steps:

[0022] S1. Import the fitting curve functions of the standard values of the voltage drop and K value of the battery cell.

[0023] In this embodiment, the voltage drop Y at different storage times T can be deduced through historical data and mathematical models, and the experimental data is fitted to obtain the functional relationship between the voltage drop and the storage time. Finally, the determination coefficients A and the degree Z are optimized by methods such as the least square method, and the fitting curve function is obtained therefrom to describe the law of the voltage change of the battery over time during storage, expressed as

[0024] Y = A * T^Z (Equation 1);

[0025] K = Y / T (Equation 2);

[0026] Wherein, Y represents the voltage drop (voltage decrease value); A represents a coefficient (related to factors such as the initial state of the battery, material properties, temperature, etc.); T represents the storage time; Z represents the order of the equation, reflecting the variation law of the voltage drop with time (such as linear, exponential or power-law relationship); K represents the K value, which is the ratio of the voltage drop to the storage time, reflecting the rate of voltage decrease per unit time, and is used to determine whether the self-discharge performance of the battery meets the standard.

[0027] After obtaining the corresponding fitting curve function, import the fitting curve functions of the cell voltage drop and the standard value of K into the device or MES system.

[0028] S2, measure the current cell voltage V1 and record the measurement time T1.

[0029] When the cell is put on line, the device automatically measures the cell voltage value V1 and records the current test time T1.

[0030] S3, obtain the initial voltage V0 of the cell and the test time T0; calculate the cell voltage difference ΔV and the storage time ΔT.

[0031] In this embodiment, directly retrieve the initial voltage value V0 of the cell and the test time T0 corresponding to the initial voltage through the device or MES system, and calculate the voltage difference and time difference respectively in the following way.

[0032] ΔV = V1 - V0; ΔT = T1 - T0.

[0033] Thus, the storage time of the cell and the corresponding voltage drop value can be automatically judged.

[0034] S4, through the fitting curve function, obtain the standard value Y of the voltage drop according to ΔT and determine whether the cell is qualified.

[0035] In this embodiment, substitute ΔT into Equation 1, then the standard value Y of the voltage drop can be automatically calculated, and determine whether the cell is qualified according to the magnitude of the Y value and ΔV.

[0036] When Y ≥ ΔV, the cell is qualified; when Y < ΔV, the cell is unqualified.

[0037] S5, through the fitting curve function, obtain the standard value and actual value of K according to ΔV and ΔT, and determine whether the cell is qualified.

[0038] Substitute ΔV and ΔT into Equation 2, calculate the standard value K of K and the actual value K1, and determine whether the cell is qualified according to the magnitude of the standard value K and the actual value K1. By comparing the actual value with the standard value, it can be judged whether the self-discharge performance of the cell meets the requirements or standards. In this embodiment, the actual value K1 = ΔV / T to reflect the self-discharge performance of the cell under actual conditions.

[0039] When K ≥ K1, the battery cell is qualified; when K < K1, it indicates that the actual value exceeds the standard value range, and there may be quality problems with the battery cell (such as internal short circuit, electrolyte decomposition, etc.), thus determining that the battery cell is unqualified.

[0040] In this embodiment, by automatically calculating the difference, the accurate pressure drop and K value standards can be obtained according to the imported fitting curve function and directly applied to the sorting and judgment of battery cells, without the need for separate setting and judgment. Moreover, the obtained standard values are more accurate, can better fit the standard values determined according to the storage duration of the battery cells, avoid using the same pressure drop and K value standards for battery cells with large differences in storage time, prevent unqualified battery cells from flowing out, and effectively improve product quality. At the same time, through the fitting curve function, the calculation accuracy can be improved, the possibility of errors and mistakes can be reduced, and the test can be made more intelligent.

[0041] Embodiment 2

[0042] In this embodiment, it also includes periodically adjusting the function parameters A and Z according to real-time data and outputting the adjusted fitting curve function to adjust the standard values of the pressure drop and K value.

[0043] In the actual production process, the parameters A and Z may be different due to battery individual differences or changes in experimental conditions. Therefore, in this embodiment, the parameters A and Z are dynamically adjusted, and the adjustment method is as follows:

[0044] First, real-time obtain data such as the storage time and pressure drop of the battery to form multiple data points. Initialize the model parameters, namely A and Z, based on historical data.

[0045] Define the loss function as

[0046]

[0047] In the formula, N is the number of data points; Y i is the actual pressure drop value of the i-th data point; T i is the storage time of the i-th data point; A·T i Z is the predicted pressure drop value of the model for the i-th data point. Measure the error between the model prediction value and the actual value through the loss function, and minimize the loss function to optimize the model parameters and make the prediction result closer to the true value.

[0048] Calculate the partial derivatives of the loss function with respect to A and Z:

[0049] To reflect the influence of the change of A on the loss function.

[0050] To reflect the influence of the change of Z on the loss function.

[0051] In the formula, Partial derivatives with respect to A and Z respectively.

[0052] Update the parameters according to the calculated partial derivatives:

[0053]

[0054] Where K is the number of iterations; α is the learning rate, controlling the step size of parameter update.

[0055] Repeat the calculation of the partial derivatives of the loss function with respect to A and Z until the loss function converges, complete the dynamic update of the parameters, and improve the accuracy and adaptability of the battery voltage drop and K value prediction.

[0056] In this embodiment, by optimizing A and Z, the parameters can better fit the actual data and reduce the prediction error. At the same time, it can capture the complex relationship between the voltage drop and the storage time more accurately. With the update of new data, it can dynamically adjust the model parameters according to the differences of the batteries to adapt to the changes in the battery state, making the prediction curve more consistent with the actual data.

[0057] Embodiment III

[0058] In this embodiment, during the actual production process, the voltage drop of the battery may also be affected by various factors, such as temperature, state of charge, degree of aging, etc., resulting in the model being unable to accurately describe the voltage drop behavior in complex scenarios. Therefore, in this embodiment, variables such as temperature and state of charge are incorporated into the model, and the fitting curve function is established as:

[0059] Y = A * T^Z · e B·Temp ·C SOC ;

[0060] Where Y is the voltage drop; T is the storage time; Temp is the temperature; SOC is the state of charge; A is the model parameter, reflecting the initial amplitude of the voltage drop; B is the temperature coefficient, reflecting the influence of temperature on the voltage drop; C is the SOC coefficient, reflecting the influence of the state of charge on the voltage drop; Z is the power parameter, reflecting the change rate of the voltage drop over time.

[0061] Then the loss function is

[0062]

[0063] And perform gradient calculations on each parameter in the way of finding partial derivatives in Embodiment II, respectively represented as

[0064]

[0065] Update the parameters according to the calculated partial derivatives to optimize the parameters.

[0066] Initialize the values of A, Z, B, and C, and for each data point (Ti , Temp i , SOC i , Y i ) Perform the following calculations:

[0067] First, calculate the predicted value, which is expressed as

[0068]

[0069] Calculate the loss function as

[0070]

[0071] Then, perform gradient calculations respectively based on this to obtain the partial derivatives of each parameter; update the parameters according to the partial derivatives, which is expressed as

[0072]

[0073] Repeat the iteration until the loss function converges or reaches the maximum number of iterations.

[0074] In this embodiment, by fully considering the temperature and SOC, the voltage drop behavior of the battery can be described more accurately. And through dynamic measurement, the parameter model can adapt to the battery performance changes under different temperature and state of charge conditions, making the predicted parameter values closer to the actual values, significantly improving the prediction accuracy and applicability of the model, and providing a more reliable basis for battery detection and quality control.

[0075] The above are only the embodiments of the present invention. Specific technical solutions and / or common knowledge such as characteristics well known in the art are not described in detail herein. It should be noted that for those skilled in the art, without departing from the technical solution of the present invention, several deformations and improvements can still be made, and these should also be regarded as the protection scope of the present invention, and these will not affect the implementation effect of the present invention and the practicality of the patent. The protection scope required by this application should be subject to the content of its claims, and the specific implementation manners described in the specification can be used to interpret the content of the claims.

Claims

1. A method for automatically determining a cell voltage drop and a K value, characterized in that: The following steps are involved: Step 1, import the fitting curve function of the battery cell voltage drop and the standard value of K value; Step 2, measure the current cell voltage V1 and record the measurement time T1; Step 3, obtaining the cell initial voltage V0 and the test time T0; Calculate the cell voltage difference ΔV and storage time ΔT; Step 4, obtaining the voltage drop standard value Y according to ΔT by fitting the curve function, and determining whether the battery cell is qualified; Step 5: Obtain the standard value and actual value of K value according to ΔV and ΔT by fitting the curve function to determine whether the battery cell is qualified.

2. The method for automatically determining the voltage drop and K value of a battery cell according to claim 1, characterized in that: The fitting curve function is expressed as Y = A*T^Z (Formula 1); K = Y / T (Formula 2); Wherein, Y represents the voltage drop; A represents the coefficient; T represents the storage time; Z represents the number of equations; K represents the K value.

3. The method for automatically determining the voltage drop and K value of a battery cell according to claim 1, characterized in that: In step 3, ΔV=V1-V0; ΔT=T1-T0.

4. The method for automatically determining the voltage drop and K value of a battery cell according to claim 2, characterized in that: In step 4, substitute ΔT into equation 1 to calculate the voltage drop standard value Y, and determine whether the battery cell is qualified based on the value of Y and the size of ΔV.

5. The method for automatically determining the voltage drop and K value of a battery cell according to claim 2, characterized in that: In step 5, ΔV and ΔT are substituted into formula 2 to calculate the standard value K and the actual value K1 of the K value, and whether the battery cell is qualified is determined based on the size of the standard value K and the actual value K1.

6. The method for automatically determining the voltage drop and K value of a battery cell according to claim 4, characterized in that: When Y≥ΔV, the battery cell is qualified; when Y<ΔV, the battery cell is unqualified.

7. The method for automatically determining the voltage drop and K value of a battery cell according to claim 5, characterized in that: The actual value K1 = ΔV / T.

8. The method for automatically determining the voltage drop and K value of a battery cell according to claim 5, characterized in that: When K≥K1, the battery cell is qualified; when K<K1, the battery cell is unqualified.

9. The method for automatically determining the voltage drop and K value of a battery cell according to claim 2, characterized in that: It also includes periodically adjusting function parameters A and Z according to real-time data, outputting the adjusted fitting curve function, and adjusting standard values ​​of pressure drop and K value.

10. The method for automatically determining the voltage drop and K value of a battery cell according to claim 9, characterized in that: The adjustment method is as follows: Where K is the number of iterations; α is the learning rate, which controls the step size of parameter update; are the partial derivatives of A and Z respectively.

Images