Multi-parameter inconsistency detection method for power battery pack based on incremental capacity analysis

Abstract

The present invention specifically discloses a method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis, and relates to the field of battery detection technology. The method comprises: S1, performing short-time pulse constant current discharge on a parallel module of a power battery pack containing n cells, and recording the terminal voltage change data of the parallel module of the battery pack in the test step; S2, using the coulomb counting method to calculate the total charge change of the parallel module of the battery pack in the test step; S3, using the incremental capacity analysis method to calculate the instantaneous rate at which the capacity of the parallel module of the battery pack changes with its terminal voltage in the test step; S4, using a statistical analysis method to extract the statistical characteristic vector of the capacity change rate; S5, using a statistical learning method to analyze the statistical characteristic vector; S6, using a machine learning method to perform multi-parameter consistency diagnosis. This method solves the problem of long detection time and high cost of the existing lithium battery pack parameter consistency detection method.

Description

Technical Field

[0001] The present invention relates to the technical field of battery detection, and in particular to a method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis. Background Art

[0002] Lithium-ion batteries, due to their advantages in power, energy density, and cycle life, have become the most widely used energy storage device in applications such as electric vehicles and energy storage power stations. In practical applications, individual cells are typically connected in parallel and then in series to increase capacity and voltage, thereby meeting high power and energy requirements. Parallel-connected cells share the same terminal voltage and are therefore often considered a single "large battery." However, due to factors such as electrochemical material properties, manufacturing process variations, and operating environment variations, lithium-ion battery packs exhibit inherent and insurmountable variations in cell parameters (including capacity, ohmic internal resistance, and contact impedance). Under cyclic charge and discharge conditions, these variations exhibit a dynamic cumulative effect: with increasing cycle counts, the polarization characteristics of individual cells diverge, leading to uneven current distribution and localized SOC shifts within the pack. This phenomenon results in a dual risk mechanism: electrically, this manifests as an increased probability of overcharge / overdischarge of specific cells, and thermodynamically, it increases the risk of localized thermal runaway, ultimately accelerating the degradation of the battery pack's overall available capacity and narrowing its safety margin. Based on this, building a rapid detection system for multi-parameter consistency of battery packs has become a key technical path to achieve battery pack health status diagnosis and balanced management, which is necessary to extend the system service life and ensure operational safety.

[0003] At present, the main methods for online detection of battery pack consistency at home and abroad include impedance detection method, EIS detection method, magnetic field detection method, etc. The first two methods are difficult to implement by embedding sensors in the battery pack, so online detection cannot be achieved. In addition, the magnetic field detection method is limited by the detection environment and is difficult to apply to online detection in actual operation scenarios. In addition, the inconsistency of multiple parameters within the battery pack can also be detected by the unbalanced current of a single cell. However, the cost of adding a current sensor to each single cell is too high and the current sensor has application limitations and is not suitable for all actual situations. Therefore, the existing detection methods are affected by environmental factors, complex working conditions and their own limitations, and cannot perform fast and effective multi-parameter inconsistency detection within the battery pack. Summary of the Invention

[0004] The purpose of the present invention is to propose a multi-parameter inconsistency detection method for a power battery pack based on incremental capacity analysis. By measuring the voltage change at the terminals of the parallel modules of the power battery pack under short-time pulse constant current discharge, the instantaneous rate at which the battery pack capacity changes with voltage is analyzed, thereby realizing rapid detection of multi-parameter inconsistency of the battery pack.

[0005] To achieve the above objectives, the present invention proposes a method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis, which includes the following steps:

[0006] Step S1: Perform short-time pulse constant current discharge on a parallel power battery pack module containing n cells, and record terminal voltage change data of the parallel power battery pack module during the test step;

[0007] Step S2: Calculate the total charge change of the parallel modules of the battery pack in the test step using the Coulomb counting method;

[0008] Step S3, using an incremental capacity analysis method to calculate the instantaneous rate of change of the capacity of the parallel modules of the battery pack with their terminal voltage in the test step;

[0009] Step S4: extracting the statistical characteristic vector of the capacity change rate using a statistical analysis method;

[0010] Step S5: Analyze the statistical feature vector using a statistical learning method;

[0011] Step S6: Use machine learning methods to perform multi-parameter consistency diagnosis, and implement rapid detection of multi-parameter inconsistency of parallel modules of the power battery pack based on the diagnosis results.

[0012] Preferably, in step S1, short-time pulse constant current discharge is performed, the current pulse amplitude adopts the standard discharge current of the battery pack to be tested 0.25C, the current pulse width adopts 30 seconds, and the terminal voltage acquisition frequency adopts 100 Hz.

[0013] Preferably, in step S2, the calculation formula for the total charge change is as follows:

[0014]

[0015] Where Q is the total charge discharged by the parallel modules of the tested power battery pack over time, I(t) is the current that changes with time t, and t1 and t2 are the starting and ending times of the unit short-time pulse constant current discharge process, respectively.

[0016] Preferably, in step S3, the incremental capacity analysis method calculation formula is as follows:

[0017]

[0018] Where, IC is the incremental capacity, and V is the terminal voltage of the parallel modules of the power battery pack under test.

[0019] Preferably, in step S4, the statistical analysis method adopts integral calculation. The change of the battery charge is related to the voltage interval distribution. The total IC curve is divided into 6 segments, and then the integral area is calculated respectively. Then the 6-dimensional statistical characteristic vector is:

[0020]

[0021] Wherein, x is the statistical eigenvector, z1 is the area of ​​the interval between (V0, V1) and the corresponding IC curve, z2 is the area of ​​the interval between (V1, V2) and the corresponding IC curve, z3 is the area of ​​the interval between (V2, V3) and the corresponding IC curve, z4 is the area of ​​the interval between (V3, V4) and the corresponding IC curve, z5 is the area of ​​the interval between (V4, V5) and the corresponding IC curve, z6 is the area of ​​the interval between (V5, V6) and the corresponding IC curve, V0, V1, V2, V3, V4, V5, and V6 are the terminal voltage values ​​corresponding to different moments respectively.

[0022] Preferably, in step S5, the specific steps of the statistical method are as follows:

[0023] Step S51: constructing an equivalent circuit model representing the actual performance state of the tested power lithium battery cell, wherein the equivalent circuit model includes an equivalent voltage source, an equivalent ohmic internal resistance, a polarization internal resistance, and a polarization capacitance;

[0024] Step S52: for a parallel power battery pack module comprising n cells, construct an equivalent circuit model of the parallel power battery pack module based on the equivalent circuit model, and generate terminal voltage change data of the parallel power battery pack module under different parameter inconsistency conditions according to step S1;

[0025] Step S53: Process the terminal voltage change data according to steps S2 to S4 to obtain IC curve data under conditions where different parameters of the parallel modules of the power battery pack are inconsistent;

[0026] Step S54: Assume that there are m IC curves, x1, x2…, x m Represent the statistical eigenvectors corresponding to each IC curve, and obtain the sample matrix X:

[0027] X=[x1 x2 … x m ] T ;

[0028] Among them, x m is the statistical eigenvector corresponding to the mth IC curve, where m is a positive integer;

[0029] Step S55: Perform Z-SCORE normalization on the sample matrix X, and then perform principal component analysis. According to the calculation results, analyze the inconsistent trend of the parameters.

[0030] Preferably, in step S6, a machine learning method is used to perform multi-parameter consistency diagnosis, and the specific process is: construct corresponding multi-parameter inconsistency degree labels for the eigenvectors in the sample matrix X, which are divided into four levels: normal, mild, moderate, and severe, to obtain a labeled data sample library D; use D to train a Gaussian naive Bayes algorithm model to establish a multi-parameter consistency diagnosis model; finally, use the trained diagnostic model to quickly diagnose the multi-parameter consistency of the tested power battery pack.

[0031] Therefore, the present invention proposes a method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis, which has the following beneficial effects:

[0032] The detection method of the present invention rapidly detects multi-parameter inconsistencies in a power lithium battery pack by measuring and analyzing the instantaneous rate at which module capacity changes with terminal voltage during short-duration pulse constant-current discharge of parallel modules. This method boasts the advantages of rapid detection of inconsistent parameters in a battery pack, high accuracy, and scalability to multi-cell modules. This method addresses the long detection time and high cost associated with existing lithium battery pack parameter consistency detection methods, providing technical support for pre-shipment testing, maintenance, safety assessment, and cascade utilization of lithium battery packs.

[0033] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a flow chart of a method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis according to the present invention;

[0035] Figure 2 A schematic diagram of the terminal voltage test results of a dual-cell parallel module provided by an embodiment of the present invention;

[0036] Figure 3 A schematic diagram of an IC curve of a dual-cell parallel module provided by an embodiment of the present invention;

[0037] Figure 4 A schematic diagram of the principal component analysis results of multi-parameter inconsistency of a dual-cell parallel module provided by an embodiment of the present invention;

[0038] Figure 5 A schematic diagram of an IC curve of a set of tested dual-cell parallel modules provided by an embodiment of the present invention;

[0039] Figure 6 A schematic diagram of the principal component analysis results of multi-parameter inconsistencies of a set of tested dual-cell parallel modules provided by an embodiment of the present invention;

[0040] Figure 7 This is a principal component analysis result of the inconsistency of a multi-cell module provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] To make the technical solutions, advantages, and objectives of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below. The described embodiments are part of the embodiments of the present invention, not all of them. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] Unless otherwise defined, technical or scientific terms used in the present invention shall have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs.

[0043] like Figure 1 FIG. 1 is a flow chart of a method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis according to the present invention, comprising the following steps:

[0044] Step 1: During the short-time constant current pulse discharge process of the lithium battery pack to be tested, the terminal voltage data outside the battery pack is obtained by using a voltage measuring device with a test step of 30s. The terminal voltage data of multiple groups of lithium battery packs are measured to obtain a series of data U1, U2, U3, ..., U n ,U1,U2,U3,…,U n Respectively represent the terminal voltage data measured by different lithium battery packs;

[0045] Step 2: According to the formula Calculate the total charge change of the battery pack during the test step. I(t) is a constant current pulse with an amplitude of 0.25C, the standard discharge current of the battery pack under test. t1 and t2 are the start and end times of the unit short-duration pulse constant current discharge process, respectively.

[0046] Step 3: Perform mathematical operations on the derivative of the discharge curve based on U and Q obtained in Steps 1 and 2, calculate the instantaneous rate of change of the volume with voltage in the test step, and obtain a series of IC curves;

[0047] Step 4: Extract features from the instantaneous rate graph obtained in step 3 and use integral calculation to calculate the change in battery charge during the test step. Divide the total IC curve into several sub-areas for statistical analysis; specifically, divide the IC curve into 6 equal parts, and then calculate the integral area of ​​each part to obtain the feature vector:

[0048]

[0049] Wherein, x is the statistical eigenvector, z1 is the area of ​​the interval enclosed by (V0, V1) and the corresponding IC curve, z2 is the area of ​​the interval enclosed by (V1, V2) and the corresponding IC curve, z3 is the area of ​​the interval enclosed by (V2, V3) and the corresponding IC curve, z4 is the area of ​​the interval enclosed by (V3, V4) and the corresponding IC curve, z5 is the area of ​​the interval enclosed by (V4, V5) and the corresponding IC curve, z6 is the area of ​​the interval enclosed by (V5, V6) and the corresponding IC curve, V0 is the terminal voltage value corresponding to the time t=0s, V1 is the terminal voltage value corresponding to the time t=5s, V2 is the terminal voltage value corresponding to the time t=10s, V3 is the terminal voltage value corresponding to the time t=15s, V4 is the terminal voltage value corresponding to the time t=20s, V5 is the terminal voltage value corresponding to the time t=25s, and V6 is the terminal voltage value corresponding to the time t=30s.

[0050] Step 5: Use statistical learning methods to analyze the statistical feature vectors in step 4. Here, the principal component analysis method is used. Assume that the above steps have obtained m IC curves, x1, x2…, x m Represent the statistical eigenvectors corresponding to each IC curve, and obtain the sample matrix X:

[0051] X=[x1 x2 … x m ] T ;

[0052] Among them, x m is the statistical eigenvector corresponding to the mth IC curve, where m is a positive integer;

[0053] After Z-SCORE standardization operation is performed on the sample matrix, principal component analysis calculation is performed, and according to the calculation results, the inconsistent trend of the parameters is analyzed;

[0054] Step 6: Match the sample matrix X with its corresponding multi-parameter inconsistency labels, specifically categorizing them into four levels: normal, mild, moderate, and severe. In this embodiment, the parameter ψ is within the range of ψ < 5%, 5% ≤ ψ < 10%, 10% ≤ ψ < 20%, and ψ ≥ 20%. Battery packs within these ranges correspond to these four levels, with ψ representing the difference in the same parameter. After obtaining the labeled data sample library D, use D to train a Gaussian Naive Bayes algorithm model to establish a multi-parameter inconsistency diagnosis model. Finally, use the trained diagnostic model to rapidly diagnose the multi-parameter consistency of the tested power battery pack.

[0055] like Figure 2 As shown in the figure, three parallel battery packs are constructed with two batteries respectively. These three battery packs include normal consistency, inconsistent internal resistance, and inconsistent capacity. Figure 2It can be seen that the terminal voltage of the battery pack varies. However, the battery module consists of a large number of batteries connected in parallel, and the individual cells have the same terminal voltage. When there is an inconsistency in the parameters of a single cell, the voltage difference caused is very small. It is difficult to detect the inconsistency of the battery pack parameters through the terminal voltage change. Therefore, it is necessary to construct other features for diagnosis.

[0056] like Figure 3 As shown in the figure, by measuring the data of 20 battery packs with inconsistent parameters and then performing incremental capacity analysis, it can be seen from the results that when the inconsistency between ohmic internal resistance and capacity deepens, the IC image features undergo large abnormal changes. Therefore, the inconsistency of battery pack parameters can be detected based on this.

[0057] like Figure 4 As shown in FIG, the principal component analysis results show that the battery packs with inconsistent parameters have obvious characteristic changes as the degree of inconsistency increases. Therefore, the IC graph distribution characteristics can be used to quickly detect inconsistent battery pack parameters.

[0058] like Figure 5-6 As shown, two batteries were selected for parallel simulation experiments. The parameters of the four groups of tested battery packs were group 1: 1520mAh, 1900mAh, group 2: 1520mAh, 1748mAh, group 3: 10.4mΩ, 10.6mΩ, group 4: 10.4mΩ, 15.6mΩ. The consistency of parameters not specified was good. The four groups of battery packs were tested, measured and subjected to incremental capacity analysis; the Gaussian naive Bayes algorithm model was used to judge the parameters, and the diagnostic results of the four battery packs were as follows: the capacity parameter inconsistency was severe, the capacity parameter inconsistency was medium, the inconsistency was normal, and the internal resistance parameter inconsistency was severe. From the results of the multi-parameter inconsistency analysis, the diagnostic results for the multi-parameter inconsistency of the battery pack were relatively accurate, and the battery pack could be diagnosed quickly and accurately. Therefore, the present invention has a good application effect.

[0059] like Figure 7 As shown, the number of parallel cells verified in the examples of the present invention is small, while the number of parallel cells in actual applications is large. From the analysis results of multiple batteries, it can be seen that the present invention is still applicable in the case of multiple batteries, which expands the scope of application of the present invention.

[0060] It is worth noting that the contents not elaborated in detail in the present invention are all prior art and are well known to those skilled in the art.

[0061] Therefore, the present invention provides a multi-parameter inconsistency detection method for a power battery pack based on incremental capacity analysis. By measuring and analyzing the instantaneous rate at which the module capacity changes with the terminal voltage during the short-time pulse constant current discharge of the parallel modules of the power lithium battery pack, rapid detection of multi-parameter inconsistencies in the battery pack is achieved, solving the problems of long detection time and high cost in the existing lithium battery pack parameter consistency detection methods.

[0062] 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 the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A multi-parameter inconsistency detection method for a power battery pack based on incremental capacity analysis, characterized in that: The following steps are involved: Step S1: Perform short-time pulse constant current discharge on a parallel power battery pack module containing n cells, and record terminal voltage change data of the parallel power battery pack module during the test step; Step S2: Calculate the total charge change of the parallel modules of the battery pack in the test step using the Coulomb counting method; Step S3, using an incremental capacity analysis method to calculate the instantaneous rate of change of the capacity of the parallel modules of the battery pack with their terminal voltage in the test step; Step S4: extracting the statistical characteristic vector of the capacity change rate using a statistical analysis method; Step S5: Analyze the statistical feature vector using a statistical learning method; Step S6: Use machine learning methods to perform multi-parameter consistency diagnosis, and implement rapid detection of multi-parameter inconsistency of parallel modules of the power battery pack based on the diagnosis results.

2. The method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis according to claim 1, characterized in that: In step S1, short-time pulse constant current discharge is performed, the current pulse amplitude adopts the standard discharge current of the battery pack to be tested 0.25C, the current pulse width adopts 30 seconds, and the terminal voltage acquisition frequency adopts 100Hz.

3. The method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis according to claim 2, characterized in that: In step S2, the calculation formula for the total charge change is as follows: Where Q is the total charge discharged by the parallel modules of the tested power battery pack over time, I(t) is the current that changes with time t, and t1 and t2 are the starting and ending times of the unit short-time pulse constant current discharge process, respectively.

4. The method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis according to claim 3, characterized in that: In step S3, the incremental capacity analysis method is calculated using the following formula: Where, IC is the incremental capacity, and V is the terminal voltage of the parallel modules of the power battery pack under test.

5. The method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis according to claim 4, characterized in that: In step S4, the statistical analysis method uses integral calculation. The change in battery charge is related to the voltage interval distribution. The total IC curve is divided into 6 segments, and the integral area is calculated for each segment. The formula for the 6-dimensional statistical eigenvector is as follows: Among them, x is the statistical characteristic vector, z1 is the area of ​​the interval enclosed by (V0, V1) and the corresponding IC curve, z2 is the area of ​​the interval enclosed by (V1, V2) and the corresponding IC curve, z3 is the area of ​​the interval enclosed by (V2, V3) and the corresponding IC curve, z4 is the area of ​​the interval enclosed by (V3, V4) and the corresponding IC curve, z5 is the area of ​​the interval enclosed by (V4, V5) and the corresponding IC curve, z6 is the area of ​​the interval enclosed by (V5, V6) and the corresponding IC curve, V0, V1, V2, V3, V4, V5, and V6 are the terminal voltage values ​​corresponding to different moments respectively.

6. The method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis according to claim 5, characterized in that: In step S5, the specific steps of the statistical method are as follows: Step S51: constructing an equivalent circuit model representing the actual performance state of the tested power lithium battery cell, wherein the equivalent circuit model includes an equivalent voltage source, an equivalent ohmic internal resistance, a polarization internal resistance, and a polarization capacitance; Step S52: for a parallel power battery pack module comprising n cells, construct an equivalent circuit model of the parallel power battery pack module based on the equivalent circuit model, and generate terminal voltage change data of the parallel power battery pack module under different parameter inconsistency conditions according to step S1; Step S53: Process the terminal voltage change data according to steps S2 to S4 to obtain IC curve data under conditions where different parameters of the parallel modules of the power battery pack are inconsistent; Step S54: Assume that there are m IC curves, x1, x2, ..., x m Represent the statistical eigenvectors corresponding to each IC curve, and obtain the sample matrix X: X=[x1 x2…x m ] T ; Among them, x m is the statistical eigenvector corresponding to the mth IC curve, where m is a positive integer; Step S55: Perform Z-SCORE normalization on the sample matrix X, and then perform principal component analysis. According to the calculation results, analyze the inconsistent trend of the parameters.

7. The method for detecting multi-parameter inconsistency of a power battery pack based on incremental capacity analysis according to claim 6, characterized in that: In step S6, multi-parameter consistency diagnosis is performed using a machine learning method. The specific process is as follows: construct corresponding multi-parameter inconsistency degree labels for the eigenvectors in the sample matrix X, which are divided into four levels: normal, mild, moderate, and severe, to obtain a labeled data sample library D; use D to train a Gaussian naive Bayes algorithm model to establish a multi-parameter consistency diagnosis model; finally, use the trained diagnostic model to quickly diagnose the multi-parameter consistency of the tested power battery pack.

Images