Self-adaptive SOC stable interval identification method based on voltage consistency standard deviation

Abstract

The invention relates to an adaptive SOC stable interval identification method based on a voltage consistency standard deviation, and belongs to the technical field of power battery fault diagnosis. Aiming at the problems of false and inconsistent voltage and high fault misjudgment rate caused by an OCV-SOC nonlinear relation between high and low SOC intervals when consistency analysis is carried out in a full SOC range in the prior art, the method is realized through the following steps: firstly, collecting and cleaning battery operation data; calculating the mean value of the row direction standard deviation of the voltage in each box according to SOC box division as an average consistency score; and finally, selecting the candidate interval with the longest continuous range as the most stable SOC interval, and extracting voltage data of the interval for subsequent fault diagnosis. According to the method, the SOC working interval with the most stable voltage can be automatically identified, nonlinear interference is effectively inhibited, and the accuracy and stability of battery consistency evaluation and fault diagnosis are remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of power battery fault diagnosis technology, and relates to an adaptive SOC stability interval identification method based on voltage consistency standard deviation. Background Technology

[0002] With the widespread application of electric vehicles, the performance and safety of power batteries, as their core components, have attracted much attention. State of charge (SCC) is a key parameter for measuring the remaining capacity of a battery, and its accuracy directly affects the effectiveness of battery management. Currently, existing power battery consistency analysis and fault diagnosis methods typically perform calculations directly within the full SCC range. However, there is a significant nonlinear relationship between the battery's open-circuit voltage and SCC. In the high and low SCC ranges, the slope of the open-circuit voltage-SCC curve is large. At this point, even small SCC measurement errors or dynamic fluctuations can cause significant voltage deviations, thus amplifying the apparent inconsistencies between cells. This spurious inconsistency further interferes with voltage-time-series-based algorithms, such as dynamic time warping and Z-score anomaly detection, leading to model misjudgments and false detections, severely reducing the stability and accuracy of fault identification.

[0003] On the other hand, the specific shape of the open-circuit voltage-charge-state curve is mainly determined by the chemical system of the battery cell, including the types of positive and negative electrode materials, reaction mechanisms, and polarization characteristics. The open-circuit voltage-charge-state curves corresponding to different chemical systems exhibit significant differences in the range of their flattened slope regions. Therefore, relying solely on manual experience to fixate on selecting a specific charge-state interval for data screening will lack universality across different chemical systems or diverse application scenarios, failing to guarantee the effectiveness of subsequent consistency evaluations or fault identification results.

[0004] Currently, most commonly used methods rely on manual experience to define the state-of-charge interval, lacking an interval selection mechanism that can adapt to data characteristics and has good generalization ability. This limitation severely restricts the promotion and application of intelligent diagnostic methods based on voltage consistency indices across vehicle models, systems, and real-world complex operating conditions.

[0005] To address the aforementioned technical issues, existing technologies urgently need an algorithm that can automatically identify the state-of-charge range with the most stable voltage and optimal consistency based on actual battery operating data, in order to significantly improve the robustness and universality of battery consistency analysis and fault diagnosis methods. Summary of the Invention

[0006] In view of this, the purpose of this invention is to provide an adaptive SOC stability interval identification method based on voltage consistency standard deviation. This method automatically identifies the SOC interval with the most stable and consistent voltage changes among battery cells by statistically analyzing the dispersion of cell voltage under different SOCs, thereby providing a more stable input data interval for subsequent fault diagnosis or model training. This method can be extended to different vehicle models, different cell formulations and operating conditions, has low computational complexity, and is easy to embed into BMS or offline analysis platforms.

[0007] To achieve the above objectives, the present invention provides the following technical solution: An adaptive SOC stability interval identification method based on voltage consistency standard deviation, the method includes the following steps: S1: Collect electric vehicle operation data, including voltage data and other characteristics of each cell in the battery pack, and perform data filtering and cleaning; S2: The battery cell voltage data is divided into bins according to the SOC value. The row standard deviation of the cell voltage matrix at each moment in each bin is calculated. Then, the mean of all standard deviations is calculated to obtain the average consistency score in the bin. S3: Apply Savitzky-Golay filtering to the average consistency score-SOC curve for smoothing. S4: Calculate the average of the smoothed scores, define a stability threshold, consider SOC bins below the threshold as stable intervals, and then merge the obtained stable intervals according to continuity to obtain several candidate stable intervals. S5: Select the longest continuous range from the candidate intervals as the most stable and complete SOC interval, and extract voltage data based on the determined most stable and complete SOC interval.

[0008] Optionally, S1 specifically includes: S11: Collect normal operation data of a certain electric vehicle, including parameters such as individual cell voltage, total voltage, total current, state of charge (SOC), timestamp, and charging / discharging status; S12: Perform data processing and cleaning on the extracted unit voltages, fill in missing values ​​and filter out outliers. S13: Establish a power battery operation database based on the collected operating data of the power battery pack and individual battery cells.

[0009] Optionally, S2 specifically includes: S21: Divide the battery cell voltage data into boxes according to the SOC value, such as each 1% SOC is a box. S22: Within each SOC box, calculate the row standard deviation of the individual unit voltage matrix at each time step; For example, currently , For a fixed value, there are a total of Individual cells, the individual cell voltage matrix of the current SOC box. Represented as:

[0010]

[0011] in, Indicates the current SOC box number The moment, i.e. the th moment Second-rate hour; Indicates the first At the [time]th moment Individual cell voltage values. Indicates the first The row standard deviation of the individual voltage matrix at each time point; Indicates the first At the [time]th moment Individual cell voltage values; Indicates the first At that moment The average value of the voltage of each individual cell.

[0012] S23: Within each SOC bin, calculate the mean of the standard deviations at each time step. :

[0013] Calculate the average consistency score for each SOC box. :

[0014] Average consistency score The smaller the value, the closer the voltage between the cells and the higher the consistency.

[0015] Optionally, S3 specifically includes: S31: SOC is plotted on the x-axis; average voltage consistency score. Plot the ordinate and draw the graph. curve; S32: Yes The curve is smoothed by applying Savitzky-Golay filtering to obtain the smoothed curve. curve.

[0016] Optionally, S4 specifically includes: S41: Based on the smoothed The curve is used to calculate the average voltage consistency score for each SOC box. average ; S42: Define the stability threshold :

[0017] in, The multiple represents the average voltage consistency score for each SOC box, and the stability threshold represents the average voltage consistency score for each SOC box. average of times.

[0018] S43: Average voltage consistency score is below the stability threshold The SOC box is considered to be at a stable interval point; S44: Merge the obtained stable interval points according to continuity to obtain several candidate stable intervals.

[0019] Optionally, S5 specifically includes: S51: Select the longest continuous interval from the candidate intervals as the most stable and complete SOC interval. :

[0020] in, This represents the starting point of the complete SOC interval, i.e. ; This represents the cutoff point of the complete SOC interval, i.e. ; This represents the longest segment of the complete SOC interval, i.e. The segment with the largest value is denoted as . .

[0021] S52: Based on the determined most stable complete SOC interval Voltage data within the determined SOC range is extracted from the raw data; S53: The voltage data obtained can be directly used in subsequent DTW or feature extraction algorithms for power battery fault diagnosis and analysis.

[0022] The beneficial effects of this invention are as follows: (1) This invention effectively avoids the false inconsistency problem caused by the nonlinear characteristics of the open-circuit voltage-state-of-charge curve between high and low state-of-charge intervals by adaptively identifying the most stable state-of-charge operating range. This method significantly improves the accuracy and stability of battery voltage consistency assessment and fault diagnosis results, making the identification of abnormal cells more accurate and reliable.

[0023] (2) The present invention has excellent versatility and adaptability. The method does not rely on a specific battery chemistry system or human experience settings, and can be applied to different vehicle models, different cell formulations and diverse actual operating conditions. It realizes cross-platform and cross-system consistency evaluation and diagnosis, which greatly expands its application scope.

[0024] (3) The algorithm of this invention has a simple and efficient structure, low computational complexity, and excellent engineering feasibility. This method does not rely on complex models or massive sample data. It can be embedded in the battery management system for real-time analysis or deployed on an offline platform for in-depth processing, providing a practical and efficient solution for early anomaly detection at the cell level.

[0025] (4) By performing feature extraction and analysis only within the identified stable state of charge range, this invention can effectively suppress interference caused by nonlinear amplification effects and highlight true abnormal characteristics. This provides a higher quality data foundation for the health monitoring and safety management of power batteries, and helps to improve the overall reliability of electric vehicles.

[0026] (5) Finally, by providing an objective and automated stable interval identification mechanism, this invention reduces the reliance on the experience of professionals and reduces the error introduced by subjective judgment, making the battery consistency analysis and fault diagnosis process more standardized and repeatable, which is conducive to the promotion and application of this technology.

[0027] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0028] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 This is a flowchart illustrating the overall method of the present invention; Figure 2 Here is a diagram of the algorithm framework for an example; Figure 3 The unit voltage curve is the original data from Example 1 or Example 2; Figure 4 The original data voltage DTW distance curve in Example 1 or Example 2; Figure 5 The average consistency score curve and the smoothed average consistency score curve for each SOC box in Example 1 or Example 2; Figure 6This is a comparison chart of the SOC range before and after screening in Example 1 or Example 2; Figure 7 The voltage DTW distance curve is the data after selecting a specified SOC range in Example 1 or Example 2. Detailed Implementation

[0029] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0030] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0031] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0032] The technical solution provided by this invention can perform statistical analysis on the state of charge (SOC) distribution of the power battery based on measured data during the operation of electric vehicles, automatically identify the stable SOC range with optimal battery voltage consistency, and ensure that subsequent feature extraction, anomaly detection and dynamic time warping (DTW) calculation are all performed within the range of stable voltage response, avoiding misjudgment caused by the nonlinear characteristics of high and low SOC, and achieving more accurate and universal power battery fault diagnosis.

[0033] Please see Figures 1-2 An adaptive SOC stability interval identification method based on voltage consistency standard deviation specifically consists of the following steps: S1: Collect electric vehicle operating data, including voltage data and other characteristics of each cell in the battery pack, and perform data filtering and cleaning, specifically including the following steps: S11: Collect normal operation data of a certain electric vehicle, including parameters such as individual cell voltage, total voltage, total current, state of charge (SOC), timestamp, and charging / discharging status; S12: Perform data processing and cleaning on the extracted unit voltages, fill in missing values ​​and filter out outliers. Data cleaning involves removing rows with a total voltage of 0, and missing data values ​​can be filled using the average value interpolation method between adjacent cells.

[0034] S13: Establish a power battery operation database based on the collected operating data of the power battery pack and individual battery cells.

[0035] S2: The battery cell voltage data is binned according to the SOC value. The row standard deviation of the cell voltage matrix at each moment in each bin is calculated. Then, the mean of all standard deviations is calculated to obtain the average consistency score within the bin. This includes the following steps: S21: Divide the battery cell voltage data into boxes according to the SOC value, such as each 1% SOC is a box. S22: Within each SOC box, calculate the row standard deviation of the individual unit voltage matrix at each time step; For example, currently , For a fixed value, there are a total of Individual cells, the individual cell voltage matrix of the current SOC box. Represented as:

[0036]

[0037] in, Indicates the current SOC box number The moment, i.e. the th moment Second-rate hour; Indicates the first At the [time]th moment Individual cell voltage values. Indicates the first The row standard deviation of the individual voltage matrix at each time point; Indicates the first At the [time]th moment Individual cell voltage values; Indicates the first At that moment The average value of the voltage of each individual cell.

[0038] S23: Within each SOC bin, calculate the mean of the standard deviations at each time step. :

[0039] Calculate the average consistency score for each SOC box. :

[0040] Average consistency score The smaller the value, the closer the voltage between the cells and the higher the consistency.

[0041] S3: Apply Savitzky-Golay filtering to the average consistency score-SOC curve for smoothing, specifically including the following steps: S31: SOC is plotted on the x-axis; average voltage consistency score. Plot the ordinate and draw the graph. score - SOC curve; S32: Yes The curve is smoothed by applying Savitzky-Golay filtering to obtain the smoothed curve. curve.

[0042] S4: Calculate the average of the smoothed scores, define a stability threshold, and consider SOC bins below this threshold as stable intervals. Then, merge the obtained stable intervals according to continuity to obtain several candidate stable intervals. This includes the following steps: S41: Based on the smoothed The curve is used to calculate the average voltage consistency score for each SOC box. average ; S42: Define the stability threshold :

[0043] in, The multiple represents the average voltage consistency score for each SOC box, and the stability threshold represents the average voltage consistency score for each SOC box. average of times.

[0044] S43: Average voltage consistency score is below the stability threshold The SOC box is considered to be at a stable interval point; S44: Merge the obtained stable interval points according to continuity to obtain several candidate stable intervals.

[0045] S5: Select the longest continuous range from the candidate intervals as the most stable complete SOC interval, and extract voltage data based on the determined most stable complete SOC interval. This includes the following steps: S51: Select the longest continuous interval from the candidate intervals as the most stable and complete SOC interval. :

[0046] in, a This represents the starting point of the complete SOC interval, i.e. ; b This represents the cutoff point of the complete SOC interval, i.e. ; This represents the longest segment of the complete SOC interval, i.e. b - a The segment with the largest value is denoted as . .

[0047] S52: Based on the determined most stable complete SOC interval Voltage data within the determined SOC range is extracted from the raw data; S53: The voltage data obtained can be directly used in subsequent DTW or feature extraction algorithms for power battery fault diagnosis and analysis.

[0048] The DTW algorithm calculates the DTW distance between the normalized individual cell voltage time series at each time point. To reduce the impact of abnormal battery cells on the average voltage, the normalized median voltage value is compared with the normalized individual cell voltage values ​​to achieve battery fault detection.

[0049] The standard score (Z-score) of the distance between each individual voltage curve and the median voltage curve is calculated as follows:

[0050] in, Indicates monomer i The distance between the voltage curve and the median voltage curve. This represents the arithmetic mean of the distances between all individual voltage curves and the median voltage curve. This represents the standard deviation of the distance between all individual voltage curves and the median voltage curve. This represents the standard fraction of the distance between each individual voltage curve and the median voltage curve.

[0051] To demonstrate the effectiveness of the present invention, the present invention prepared the operating data of a faulty vehicle. Within the complete cycle, there were two abnormal units that showed different degrees of interference fluctuations on the DTW curve. Figure 3 The DTW algorithm is applied to the original voltage data to obtain the individual unit voltage curves. Figure 4 The original data shows the voltage DTW distance curve. After adaptive SOC stability interval identification, the following is obtained: The curve is then smoothed using Savitzky-Golay filtering, and a stability threshold is defined here. Take 0.9, Figure 5 The average consistency score curve and the smoothed average consistency score curve for each SOC bin. Figure 6 To compare the SOC ranges before and after filtering, the SOC interval was determined to be (9, 49). The SOC was extracted from the original voltage data. Voltage data within the range are processed using the DTW algorithm. Figure 7 The voltage-to-displacement (DTW) distance curve of the data after obtaining data within a specified SOC range significantly suppresses local interference fluctuations caused by high and low SOCs in abnormal cells. Simultaneously, the abnormal increase in DTW distance at later time points is effectively highlighted, making the identification of abnormal cells more intuitive. Therefore, this invention can adaptively identify the most stable SOC operating range based on battery operating data, effectively suppressing the interference of nonlinear amplification effects in the high and low SOC ranges on consistency analysis. By performing feature extraction and fault diagnosis only within a stable SOC range, this invention can significantly improve the accuracy and stability of power battery consistency assessment and fault identification, thereby achieving high-precision, real-time health monitoring of electric vehicle power batteries and accurate location of abnormal cells.

[0052] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An adaptive SOC stability interval identification method based on voltage consistency standard deviation, characterized in that: The method includes the following steps: S1: Collect electric vehicle operation data, including the voltage data of each cell in the battery pack, and perform data filtering and cleaning; S2: The battery cell voltage data is divided into bins according to the state of charge (SOC) value. The row standard deviation of the cell voltage matrix at each moment in each SOC bin is calculated. Then, the mean of all row standard deviations is calculated to obtain the average consistency score in the SOC bin. S3: Smooth the average consistency score-SOC curve; S4: Calculate the average of the smoothed average consistency scores, define a stability threshold, identify SOC bins below the stability threshold as stable interval points, and merge consecutive stable interval points to obtain several candidate stable intervals. S5: Select the longest continuous segment from the candidate stable intervals as the most stable complete SOC interval, and extract voltage data based on the determined most stable complete SOC interval.

2. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 1, characterized in that: S1 specifically includes: S11: Collect normal operation data of electric vehicles, including individual cell voltage, total voltage, total current, state of charge (SOC), timestamp, and charging / discharging status; S12: Perform data processing and cleaning on the extracted unit voltages, fill in missing values, and filter out outliers. S13: Establish a power battery operation database based on the collected operating data of the power battery pack and individual battery cells.

3. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 1, characterized in that: S2 specifically includes: S21: The battery cell voltage data is sorted into bins according to the SOC value; S22: Within each SOC box, calculate the row standard deviation of the individual unit voltage matrix at each time step; S23: Within each SOC bin, calculate the mean of the row standard deviations obtained at each time step to obtain the average consistency score of the SOC bin.

4. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 3, characterized in that: In S22, the individual voltage matrix of the current SOC box T Represented as: in, Indicates the current SOC box number n At that moment, Indicates the first n At the [time]th moment m The voltage value of the first individual cell; the first n Row standard deviation of the individual voltage matrix at each time point The calculation formula is: in, Indicates the first n At the [time]th moment i Individual cell voltage values, Indicates the first n At that moment m The average value of the voltage of each individual cell.

5. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 3, characterized in that: In S23, the formula for calculating the average consistency score for each SOC bin is as follows: in, Indicates the first j The standard deviation of the direction of travel at each moment. n This represents the total number of time points within the current SOC box.

6. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 1, characterized in that: In step S3, Savitzky-Golay filtering is used to smooth the average consistency score-SOC curve.

7. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 1, characterized in that: S4 specifically includes: S41: Calculate the smoothed average voltage consistency score for each SOC box. score 'Average ; S42: Define the stability threshold Threshold for: Where N is a preset multiple; S43: Average voltage consistency score score 'Below the stability threshold' Threshold The SOC box was identified as a stable interval point; S44: Merge the identified stable interval points according to their continuity to obtain the candidate stable interval.

8. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 1, characterized in that: S5 specifically includes: S51: Select the segment with the longest duration from the candidate stable intervals as the most stable complete SOC interval. ,in a' The SOC value is the starting point of the interval. b ' is the SOC value at the interval cutoff point; S52: Extract SOC values ​​from the raw data within the range Voltage data within the range; S53: Use the extracted voltage data for subsequent fault diagnosis and analysis.

9. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 8, characterized in that: In S53, the fault diagnosis analysis includes applying the Dynamic Time Warping (DTW) algorithm or the feature extraction algorithm.

10. The adaptive SOC stability interval identification method based on voltage consistency standard deviation according to claim 9, characterized in that: The application of the Dynamic Time Warping (DTW) algorithm includes: calculating the DTW distance between each individual voltage time series and the median voltage time series, and calculating the standard score (Z-score) of the distance between each individual voltage curve and the median voltage curve, the expression of which is: in, Indicates monomer i The DTW distance between the voltage curve and the median voltage curve. This represents the arithmetic mean of the distances between all individual voltage curves and the median voltage curve. This represents the standard deviation of the distance between all individual voltage curves and the median voltage curve.

Images