A method, system, and vehicle for early warning and diagnosis of abnormal battery capacity consistency.

Abstract

This invention discloses a method, system, and vehicle for early warning and diagnosis of abnormal battery capacity consistency. The method includes: acquiring historical charging data of a target power battery system within a preset time period; segmenting the data into multiple continuous charging segments based on a first set condition; selecting charging segments where the remaining battery power is higher than a set power threshold to form a first set; if the number of charging segments in the first set exceeds a set threshold, identifying charging segments that meet a second set to form a second set; further selecting charging segments from the second set that meet a third set to form a third set; if the number of charging segments in the third set meets a fourth set condition, determining whether there is an abnormality in the battery capacity of the target power battery system based on the voltage change trend of the charging segments in the third set, and issuing an early warning if so. This invention enables early online warning of abnormal battery capacity consistency using only regular charging data from the vehicle's cloud platform.

Description

Technical Field

[0001] This invention belongs to the field of power battery diagnostic technology, and more specifically, relates to a method, system, and vehicle for early warning diagnosis of abnormal battery capacity consistency. Background Technology

[0002] With the rapid development of the new energy vehicle industry, lithium-ion power battery systems, as their core power source, directly affect the vehicle's range, safety, and lifespan through their performance and reliability. Among them, lithium iron phosphate batteries have become the mainstream choice in commercial vehicles and some passenger vehicles due to their excellent thermal stability, long cycle life, and relatively low cost.

[0003] However, lithium iron phosphate batteries also face significant challenges in application. Their inherent characteristics, such as a flat voltage plateau and low energy density, make accurately assessing their health status and performance consistency particularly difficult. In practical use, due to factors such as manufacturing tolerances, differences in operating temperature, and uneven charge and discharge processes, the capacity decay rate of hundreds or thousands of individual cells within a battery pack is often inconsistent, resulting in "capacity consistency" degradation. As vehicle mileage increases, this inconsistency continues to worsen, leading to a faster decline in system usable capacity, a shorter driving range, and in severe cases, even overcharging and over-discharging, creating safety hazards. Therefore, early and accurate online warnings for abnormal battery capacity consistency are of crucial engineering significance for improving battery system management, ensuring vehicle safety, and extending the overall lifespan of the battery pack.

[0004] Currently, the industry lacks efficient, convenient, and reliable solutions for assessing battery health, especially for consistency monitoring of lithium iron phosphate batteries. Common traditional methods each have their own significant limitations:

[0005] 1) Discharge test method, voltage method, internal resistance method, etc. usually require the battery to be under specific test conditions (such as constant current, long-term static placement, and suitable temperature), which is difficult to meet in actual vehicle operation and cannot achieve real-time online monitoring.

[0006] 2) The ampere-hour integration method is susceptible to errors in initial state of charge estimation and cumulative errors in current measurement, and its accuracy will drift significantly after long-term use;

[0007] 3) Data-driven methods based on battery models or machine learning, while highly accurate, heavily rely on a large amount of complete laboratory calibration data or long-term operational data for model training. This results in a long development cycle, weak generalization ability, and poor adaptability to different battery models.

[0008] In summary, existing technologies cannot effectively and timely diagnose abnormal capacity consistency of lithium iron phosphate batteries using only data naturally generated during the daily operation of the vehicle, without affecting the normal use of the vehicle or relying on specific testing conditions.

[0009] Therefore, there is an urgent need for an online early warning and diagnosis method that can utilize existing vehicle-reported data, adapt to complex actual working conditions, and has a high accuracy rate.

[0010] The information disclosed in the background section of this invention is intended only to enhance the understanding of the general background of this invention and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art. Summary of the Invention

[0011] The purpose of this invention is to propose a method, system, and vehicle for early warning and diagnosis of battery capacity inconsistency, solving the problem that existing methods for assessing capacity inconsistency in lithium iron phosphate batteries rely on specific test conditions or large amounts of training data, making it difficult to provide real-time online warnings. This invention enables early online warnings of battery capacity inconsistency using only regular charging data from the vehicle's cloud platform, without requiring additional testing.

[0012] To achieve the above objectives, in a first aspect, the present invention proposes a method for early warning and diagnosis of abnormal battery capacity consistency, comprising:

[0013] Acquire historical charging data of the target power battery system within a preset time period;

[0014] The historical charging data is segmented based on the first set condition to obtain multiple consecutive charging segments;

[0015] The charging segments in which the remaining power battery charge is higher than a set power threshold are identified from each of the charging segments and constitute the first set;

[0016] If the number of charging segments in the first set is greater than a set threshold, then the charging segments that meet the second set condition are identified from the first set to form the second set.

[0017] The charging segments that meet the third set are identified from the second set and a third set is formed.

[0018] If the number of charging segments in the third set meets the fourth preset condition, then based on the voltage change trend of the charging segments in the third set, it is determined whether there is an inconsistency in the battery capacity of the target power battery system. If so, an early warning is issued.

[0019] Optionally, the historical charging data includes:

[0020] Multiple charging times, the voltage of each individual cell at each charging time, the cell number with the highest individual voltage, the remaining power capacity of the power battery, and the vehicle's mileage.

[0021] Optionally, the first setting condition includes:

[0022] If the time difference between two adjacent charging moments is less than a set time threshold, and the corresponding change in vehicle mileage is less than a set mileage threshold, then the charging data of the two charging moments will be divided into the same charging segment.

[0023] Optionally, the second setting condition includes:

[0024] In all charging records of the first set, the cell with the highest single-cell voltage is the cell with the largest proportion of charging records of the same cell, and is regarded as a suspected abnormal cell, and the proportion is greater than a first proportion threshold.

[0025] The suspected abnormal cell is the cell with the highest single-cell voltage for a period of time that is greater than the second proportion threshold during the entire charging segment.

[0026] Optionally, the third setting condition includes:

[0027] The charging segment in which the voltage difference between the suspected abnormal cell and the average voltage of the other cells is greater than a set voltage difference threshold.

[0028] Optionally, the fourth setting condition includes:

[0029] The proportion of the number of charging segments in the third set to the number of the second set is greater than the third proportion threshold, and the proportion of the number of charging segments in the third set to the number of the first set is greater than the fourth proportion threshold.

[0030] Optionally, determining whether there is an anomaly in the battery capacity of the target power battery system based on the voltage change trend of the charging segments in the third set includes:

[0031] Select the last few adjacent charging segments in time sequence from the third set;

[0032] For each of the charging segments, calculate the first voltage change slope of the suspected abnormal cell at the last multiple consecutive time sampling points, and the second voltage change slope of the average voltage of the remaining cells.

[0033] If at least one of the charging segments simultaneously meets the following conditions, then the consistency anomaly is determined to exist:

[0034] The slope of the first voltage change is greater than zero;

[0035] The slope of the second voltage change is less than zero and less than a set slope threshold;

[0036] The average voltage of the remaining cells showed a continuous decreasing trend in the last several consecutive sampling points;

[0037] The cell with the highest single-cell voltage in the last charging segment of the first set is the suspected abnormal cell.

[0038] Optionally, the set power threshold is 98%.

[0039] Secondly, this invention proposes an early warning and diagnostic system for abnormal battery capacity consistency, comprising:

[0040] The acquisition module is used to acquire historical charging data of the target power battery system within a preset time period;

[0041] The segmentation module is used to segment the historical charging data based on a first set condition to obtain multiple continuous charging segments.

[0042] The first identification module is used to identify charging segments from each of the charging segments where the remaining power battery charge is higher than a set power threshold, and to form a first set;

[0043] The second identification module is used to identify charging segments that meet the second set conditions from the first set if the number of charging segments in the first set is greater than a set number threshold, and to form a second set.

[0044] The third identification module is used to identify charging segments that meet the third set conditions from the second set, and form the third set;

[0045] The judgment and early warning module is used to determine whether there is an inconsistency in the battery capacity of the target power battery system based on the voltage change trend of the charging segments in the third set if the number of charging segments in the third set meets the fourth preset condition. If so, an early warning is issued.

[0046] Thirdly, the present invention provides a vehicle comprising the early warning and diagnostic system for abnormal battery capacity consistency described in the second aspect.

[0047] The beneficial effects of this invention are as follows: This invention can perform automated analysis using only existing conventional historical charging data in the vehicle's cloud, without relying on specific laboratory testing conditions or additional sensor deployments. This achieves low-cost, non-invasive online monitoring of battery consistency anomalies. The method filters statistically significant anomalies from massive amounts of data layer by layer, focusing on the core dynamic characteristic of "voltage trend divergence" at the charging end for final judgment. This significantly improves the accuracy and reliability of early warnings, effectively avoiding false alarms caused by occasional fluctuations or noise. The final output warning information can be accurate down to the specific vehicle and individual battery cell, providing a direct basis for targeted battery maintenance, balancing, or replacement. This helps to mitigate risks in advance, extend the overall lifespan of the battery system, and ensure vehicle operational safety.

[0048] The system of the present invention has other features and advantages that will be apparent from or will be set forth in detail in the accompanying drawings and following detailed description, which together serve to explain the objective principles of the invention. Attached Figure Description

[0049] The above and other objects, features and advantages of the present invention will become more apparent from the accompanying drawings, in which like reference numerals generally denote like parts.

[0050] Figure 1 A flowchart illustrating the steps of the early warning diagnostic method for abnormal battery capacity consistency according to the present invention is shown.

[0051] Figure 2 The graph shows the trend of the voltage (Vn) of the suspected abnormal cell and the average voltage (Vavg) of the remaining cells over time during the charging process of a power battery according to Embodiment 1 of the present invention.

[0052] Figure 3 A schematic diagram of an early warning and diagnostic system for abnormal battery capacity consistency according to Embodiment 2 of the present invention is shown. Detailed Implementation

[0053] The invention will now be described in more detail with reference to the accompanying drawings. While preferred embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0054] like Figure 1As shown, a method for early warning and diagnosis of abnormal battery capacity consistency according to the present invention includes:

[0055] Acquire historical charging data of the target power battery system within a preset time period;

[0056] Based on the first set condition, the historical charging data is segmented to obtain multiple continuous charging segments.

[0057] The charging segments with remaining battery power exceeding a set threshold are identified from each charging segment and constitute the first set.

[0058] If the number of charging segments in the first set is greater than the set number threshold, then the charging segments that meet the second set condition are identified from the first set and constitute the second set.

[0059] The charging segments that meet the third set conditions are identified from the second set and a third set is formed.

[0060] If the number of charging segments in the third set meets the fourth set condition, then based on the voltage change trend of the charging segments in the third set, it is determined whether there is an inconsistency in the battery capacity of the target power battery system. If so, an early warning is issued.

[0061] Specifically, the early warning and diagnostic method for abnormal battery capacity consistency provided by this invention is based on an automated diagnostic logic constructed from historical charging data, multi-level filtering, and intelligent analysis. The method begins with a data preparation phase, where the system acquires all historical charging data uploaded to the cloud by the target power battery system within a preset period (e.g., one month). This data typically includes key information such as timestamps, individual cell voltages, total remaining charge (SOC) of the battery pack, and cumulative vehicle mileage. Subsequently, the data preprocessing and fragmentation phase begins. Based on a first set condition—typically, the time interval between adjacent data points is extremely short and the vehicle mileage remains almost unchanged—the continuous data stream is divided into multiple independent charging segments. Each charging segment represents a complete charging event, thus providing clear time boundaries for subsequent analysis.

[0062] The core of the diagnostic process is a three-tiered, progressive screening and analysis structure designed to gradually focus on genuine anomalies from massive amounts of data. First, in the initial screening, the system traverses all charging stages, selecting only segments where the remaining battery capacity exceeds a set threshold (e.g., SOC > 98%) to form the first set. This step aims to lock onto the critical observation window at the end of the charging process, as the battery chemical state tends towards saturation at this point, and the capacity differences between individual cells are maximized in the form of voltage differentiation. To ensure the analysis is statistically significant, the system checks whether the number of segments in the first set meets the minimum requirement (a set threshold).

[0063] Once the data foundation meets the requirements, the second layer of screening begins. The system conducts in-depth analysis of the first set and identifies more targeted abnormal segments based on a second set of conditions, forming the second set. This condition is usually associated with a specific suspected abnormal cell; for example, it requires that the cell consistently exhibits the highest voltage in a sufficiently high proportion of the end-of-charge records, which initially indicates potential capacity degradation.

[0064] Furthermore, in the third layer of screening, the system refines the second set based on a third set of conditions to generate a third set. These conditions typically involve quantitative calculations, such as requiring that the difference between the voltage of a suspected abnormal cell and the average voltage of the remaining cells in the battery pack must exceed a significant threshold, thereby transforming a simple voltage spike into a quantifiable and significant deviation, thus strengthening the abnormal signal.

[0065] After completing the data focusing described above, the method performs statistical significance verification (i.e., determines whether the fourth predefined condition is met), for example, requiring that the number of fragments in the third set accounts for a certain proportion in the second set and even the first set. This ensures that the observed anomalous patterns are not accidental but statistically repeatable and stable.

[0066] Finally, the system enters the dynamic trend analysis and decision-making stage. Based on a high-purity third set of data, especially the most recent charging segments, the system analyzes the instantaneous voltage change trend at the final moment of charging. The core diagnostic criterion is capturing trend divergence: when the voltage of a suspected abnormal cell is still rising (positive slope), while the average voltage of other cells has begun to clearly decline (negative slope, and the trend is continuous), this divergence is considered dynamic evidence of capacity inconsistency. Once this trend condition is met, the system determines that there is a consistency anomaly and automatically generates warning information accurate to the specific vehicle and suspected individual cell number, thus completing the entire process from data to diagnostic decision. The entire method relies solely on existing cloud monitoring data, achieving online, non-intrusive intelligent early warning for latent battery faults.

[0067] The charging segments in the first, second, and third sets are all arranged in time sequence.

[0068] In one example, historical charging data includes:

[0069] Multiple charging times, the voltage of each individual cell at each charging time, the cell number with the highest individual voltage, the remaining power capacity of the power battery, and the vehicle's mileage.

[0070] In one example, the first preconditions include:

[0071] If the time difference between two adjacent charging moments is less than a set time threshold, and the corresponding change in vehicle mileage is less than a set mileage threshold, then the charging data of the two charging moments will be divided into the same charging segment.

[0072] Specifically, in the data preprocessing stage of this method, one of the key steps is to restore and divide the continuously reported, timestamped raw charging data stream into several independent charging events, i.e., charging segments. To achieve this, the first setting condition is a simple and effective judgment rule based on physical logic. Specifically, the system sequentially examines each pair of adjacent charging data records (each record corresponds to a specific charging time). For any two adjacent records, the system simultaneously checks two core parameters: the time difference between them and the difference in their corresponding vehicle running distance (ODO). Only when both of these differences are less than their respective predefined thresholds, i.e., the time difference is less than a set time threshold (e.g., 10 minutes) and the mileage change is less than a set mileage threshold (e.g., 0.1 kilometers), does the system determine that these two records belong to the same uninterrupted charging process, thus merging them into the same charging segment. The underlying logic of this rule is that in a real vehicle charging scenario, if the charging process is not interrupted by driving behavior, then between adjacent data collection points, the vehicle should hardly move (mileage remains unchanged), and the collection interval is usually very short. Conversely, if the time interval is too long or the mileage increases significantly, it means that charging has stopped, and the vehicle may have entered a driving or long-term stationary state. Through this dual criterion, the method can accurately distinguish different charging events, laying a reliable data foundation for subsequent analysis based on complete charging segments and effectively avoiding analytical biases caused by data splicing errors.

[0073] In one example, the second condition includes:

[0074] In all charging records of the first set, the cell with the highest single-cell voltage is the cell with the largest proportion of charging records of the same cell, and is considered a suspected abnormal cell, and the proportion is greater than the first proportion threshold.

[0075] A suspected abnormal cell is a cell that is the highest single-cell voltage cell for a greater than the second proportion threshold during the entire charging process.

[0076] Specifically, the second condition defines a rigorous dual criterion, aiming to accurately identify specific individual cells within the battery pack most likely to experience capacity degradation from both macroscopic statistical regularities and microscopic process performance dimensions, and thereby screen out charging segments with high analytical value. Its logical flow consists of two sequential steps: First, the system statistically analyzes all charging records of the first set (i.e., all high SOC charging segments), calculating the frequency of each cell appearing as the highest voltage cell. Cells with the highest frequency, exceeding a preset first proportion threshold (e.g., 50%), are marked as suspected abnormal cells. This step searches for recurring abnormal patterns from numerous charging events, ensuring that the target cell's anomaly has statistical significance across cycles, rather than being a random event. Subsequently, the condition is further tightened; it does not simply accept all charging segments containing the cell, but requires that the cell's voltage performance within a single charging segment also be consistent. Specifically, only charging segments in which the suspected abnormal cell exists as the highest-voltage single cell for a period exceeding a second threshold (e.g., 90%) are ultimately adopted and included in the second set. This step is a microscopic verification of the statistical results, excluding segments where the cell's voltage spikes only at the end of charging, ensuring that the analyzed charging segments fully reflect the charging process in which the cell is in an abnormal state from start to finish. Through this macroscopic statistical targeting and microscopic process confirmation of a continuous dual screening mechanism, the second set of conditions efficiently eliminates a large amount of interfering data, constructing a set of charging segments with strong abnormal signals and high consistency, laying a highly reliable data foundation for subsequent quantitative analysis and final diagnosis.

[0077] In one example, the third condition includes:

[0078] A charging segment where the voltage difference between a suspected abnormal cell and the average voltage of the other cells exceeds a set voltage difference threshold.

[0079] Specifically, the third setting condition is a crucial quantitative verification step in the entire diagnostic process. Its core lies in transforming the suspected anomalies identified in the previous step based on frequency and time proportion into a precisely measurable and objectively comparable physical quantity of evidence. Specifically, this condition performs a precise calculation on the charging segments already screened by the second setting condition (i.e., those segments dominated by the statistically most suspicious cells that consistently exhibit the highest voltage during a single charge). For each such charging segment, the system calculates the difference between the real-time voltage of the suspected anomalous cell and the average voltage of all other cells in the battery pack. This voltage difference directly quantifies the degree of deviation of the suspected cell's electrical state from the battery group under the same charging time and current excitation. This condition requires that only charging segments with voltage differences consistently or significantly exceeding a preset voltage difference threshold (e.g., 0.15V) can be ultimately retained and included in the third set. This threshold setting is significant; as a rigid technical barrier, it effectively distinguishes normal voltage fluctuations from substantial voltage deviations indicating capacity decay. Through this condition, the system achieves a leap from identifying abnormal behavioral patterns to quantifying the severity of anomalies. It not only filters out suspicious cells that, while having the highest potential, have only a slight advantage or may be caused by measurement noise or minor inconsistencies, thus further purifying the analysis sample set, but more importantly, it provides a solid, objective basis for the final diagnostic conclusion based on clear physical quantities, ensuring that subsequent early warning decisions are not based on vague trends, but on significant electrical signals that exceed reasonable fluctuation ranges.

[0080] In one example, the fourth condition includes:

[0081] The number of charging segments in the third set accounts for a greater proportion of the number of charging segments in the second set than the third proportion threshold, and the number of charging segments in the third set accounts for a greater proportion of the number of charging segments in the first set than the fourth proportion threshold.

[0082] Specifically, in this method, the fourth precondition plays a crucial role in verifying statistical significance. Its purpose is to use rigorous statistical rules to assess the credibility of the third set (i.e., abnormal charging segments containing significant voltage differences) obtained after the first three rounds of screening, determining whether the observed abnormal patterns are generalized rather than isolated phenomena under specific conditions. This condition sets two proportion thresholds that must be met simultaneously, forming a rigorous framework for cross-validation: First, it calculates the proportion of charging segments in the third set relative to the second set (comprising charging segments dominated by suspected abnormal cells), requiring this proportion to be greater than the third proportion threshold. This validation, known as internal consistency validation, ensures that a sufficiently high proportion of the charging events identified as highly suspicious exhibit quantified, significant voltage differences, thus confirming the effectiveness of the initial screening logic and the concentration of abnormal behavior in the target cells. Second, and more importantly, it simultaneously calculates the proportion of charging segments in the third set relative to the initial first set (all high-SOC charging segments), requiring this proportion to be greater than the fourth proportion threshold. This verification, known as global significance verification, requires that significant anomalies not only appear in the suspected subset but also constitute a significant proportion of all available charging events within the entire observation window. This dual verification mechanism is crucial: if only internal consistency is met while the global proportion is low, it may indicate that the anomaly is triggered only under specific conditions (such as a certain charging mode or ambient temperature) and is not a widespread systemic risk; conversely, if the global proportion meets the standard but internal consistency is insufficient, it may mean that the screening logic is not precise enough. Only when both proportions exceed the predetermined thresholds is the capacity inconsistency characteristic exhibited by the suspected abnormal cell considered to have sufficient statistical significance and universality, representing a stable and evolving fault trend, thus allowing the process to proceed to the final dynamic trend analysis stage. This step greatly enhances the anti-interference capability and reliability of the early warning system, effectively filtering out false alarms caused by data noise, transient operating conditions, or accidental fluctuations.

[0083] In one example, based on the voltage change trend of the charging section in the third set, determining whether there is an inconsistency in the battery capacity of the target power battery system includes:

[0084] Select the last few adjacent charging segments in terms of timing from the third set;

[0085] For each charging segment, calculate the first voltage change slope of the suspected abnormal cells at the last multiple consecutive time sampling points, and the second voltage change slope of the average voltage of the remaining cells.

[0086] If at least one charging segment simultaneously meets the following conditions, then a consistency anomaly is determined to exist:

[0087] The slope of the first voltage change is greater than zero;

[0088] The slope of the second voltage change is less than zero and less than the set slope threshold;

[0089] The average voltage of the remaining cells showed a continuous decreasing trend in the last several consecutive sampling points;

[0090] The cell with the highest single-cell voltage in the last charging segment of the first set is a suspected abnormal cell.

[0091] Specifically, this step is the core of the final decision and dynamic feature verification in the early warning and diagnosis process. Based on the high-purity abnormal samples (third set) extracted from multiple rounds of screening and statistical verification in the early stages, it makes the most physically-based final judgment on capacity consistency anomalies by analyzing the instantaneous voltage dynamic behavior at the very end of the charging process. This process first focuses on the latest evidence in time, selecting multiple adjacent charging segments that occurred last in time from the third set for analysis. This ensures that the diagnostic conclusion reflects the latest state of the battery system and has real-time early warning value. For each selected charging segment, the method accurately calculates two key trend indicators: one is the slope of voltage change of the suspected abnormal cell in the last few consecutive sampling points at the end of charging (first voltage change slope), and the other is the slope of voltage change of the average voltage of all other cells in the battery pack at the same time in the last few consecutive sampling points at the end of charging (second voltage change slope). The physical insight behind this design lies in the following: During the final stages of charging, if a cell reaches full charge earlier due to capacity decay, its voltage will continue to rise during the final stages of constant or reduced current charging, exhibiting a positive slope. Simultaneously, the average voltage of a group of cells with normal capacity will naturally decline after reaching saturation due to reduced polarization or adjustments in the charging strategy, exhibiting a negative slope. The simultaneous occurrence of these two opposing trends constitutes the decisive dynamic characteristic for diagnosing capacity inconsistency: voltage trend divergence.

[0092] To translate this feature into a reliable criterion, the method sets a series of stringent conditions that must be simultaneously met by at least one recent charging segment: First, the slope must be greater than zero, confirming that the voltage of the abnormal cell is indeed rising; second, the slope must not only be less than zero but also less than a preset negative slope threshold, ensuring that the voltage decline trend of normal cells is sufficiently significant, eliminating minor fluctuations caused by measurement noise or small balancing currents; furthermore, the average voltage of normal cells must show a continuous downward trend across the last few sampling points, meaning the voltage value at each subsequent point is no higher than the previous point, further reinforcing the clarity and monotonicity of the downward trend and avoiding interference from intermittent fluctuations. Finally, as a cross-validation of state persistence, the condition requires that in the most recent charging segment in the first set (all high SOC charging segments), the highest-voltage cell is still the suspected abnormal cell. This confirms that the abnormal dominance of the cell persists in the most recent observable charging event, ensuring the timeliness and accuracy of the warning. Only when all these conditions are met simultaneously does the system ultimately determine that the target power battery system has a definite capacity consistency anomaly. This comprehensive set of criteria, based on dynamic trend divergence, quantitative slope threshold, continuous decline verification, and latest state review, deeply integrates electrochemical mechanisms and data science, enabling early warning decisions to have both high physical credibility and anti-interference capabilities, thus achieving a closed loop from data to accurate diagnosis.

[0093] In one example, the battery threshold is set to 98%.

[0094] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the invention. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present invention can be combined with each other.

[0095] Example 1

[0096] This embodiment provides a method for early warning and diagnosis of abnormal battery capacity consistency, including:

[0097] First, all historical charging data for vehicles with Vehicle Identification Number (VIN) "xxx" within a one-month monitoring period is acquired. The data package includes multiple charging times and the corresponding cell voltage, highest voltage cell number, remaining battery charge (SOC), and vehicle mileage (ODO) for each time. Data is segmented according to a first set condition, with a set time threshold of 10 minutes and a set mileage threshold of 0.1 kilometers. Continuous data points that meet the condition of "time difference less than 10 minutes and ODO change less than 0.1 kilometers" are grouped into the same charging segment, thus obtaining a series of independent charging event data.

[0098] Next, the power threshold is set to 98%, and segments with SOC > 98% are selected from all charging segments, totaling N0 = 6 segments, forming the first set. The preset quantity threshold is 5. Since N0 (6) > 5, the conditions for in-depth analysis are met. Subsequently, the second setting condition is executed: in the first set, it is found that cell No. 57 is the highest voltage single cell in 75% (r1 = 0.75) of the records. This proportion exceeds the preset first proportion threshold of 50%, so it is identified as a suspected abnormal cell (Vn = 57). Based on this, the second proportion threshold is set to 90%, and segments in which cell No. 57 is the highest voltage single cell for more than 90% of the time in its entire charging segment are selected, resulting in N1 = 4 charging segments, forming the second set. Then, based on the third set condition, the characteristic voltage difference (Vdiff) of each segment in the second set is calculated. The voltage difference threshold is set to 0.15V, and segments with Vdiff greater than 0.15V are selected (the actual maximum difference max_vdiff=0.177V). All 4 segments meet the condition, so N2=4, forming the third set.

[0099] Statistical verification was then performed (fourth pre-set condition): the proportion of the third set (N2=4) in the second set (N1=4) was calculated as R2=1.0, and the proportion in the first set (N0=6) was R3≈0.667. Both proportions exceeded the preset third and fourth proportion thresholds (both 0.5), thus passing the statistical significance test. Finally, dynamic trend analysis was performed: the three most recent charging segments in time sequence were selected from the third set, and the slopes of the last multiple sampling points at the end of charging were calculated. The slope threshold for judging the trend strength was set to -0.0015. The analysis data showed that the voltage change slope (Kt) of cell 57 was 0.0085, 0.0045, and 0.0065 in the three most recent segments, all greater than zero; while the average voltage change slope (Kavg) of the remaining cells was -0.002, which was not only less than zero but also less than the threshold of -0.0015. Meanwhile, the average voltage of the remaining cells showed a continuous downward trend across the sampling points (all relevant flag bits were 1), and the highest voltage cell in the last charging segment of the first set was confirmed to be cell number 57 (last_Vn=57). Since all final judgment conditions were met simultaneously, it was determined that cell number 57 in the power battery system of vehicle "xxx" has a definite capacity consistency anomaly, and a precise warning was generated for this vehicle and this specific cell.

[0100] Table 1 shows the cell parameter data records related to the diagnosis of capacity consistency anomalies during the charging process of the power battery.

[0101]

[0102] Table 1. Cell Voltage and Consistency Status Parameters for Power Battery Charging Section

[0103] The characteristics of the vehicle's charging terminal voltage data are as follows: Figure 2 As shown, under a certain current excitation, the voltage Vn of the low-capacity cell will still rise, while the average voltage Vavg of other high-capacity cells has shown a downward trend, indicating an abnormality in capacity consistency.

[0104] Example 2

[0105] like Figure 3 As shown, this embodiment provides an early warning and diagnostic system for abnormal battery capacity consistency, including:

[0106] The acquisition module is used to acquire historical charging data of the target power battery system within a preset time period;

[0107] The segmentation module is used to segment historical charging data based on a first set condition to obtain multiple continuous charging segments.

[0108] The first identification module is used to identify charging segments from each charging segment where the remaining power battery charge is higher than a set power threshold, and to form a first set;

[0109] The second identification module is used to identify charging segments that meet the second set conditions from the first set if the number of charging segments in the first set is greater than a set number threshold, and to form a second set.

[0110] The third identification module is used to identify charging segments that meet the third set conditions from the second set, and form the third set;

[0111] The judgment and early warning module is used to determine whether there is an inconsistency in the battery capacity of the target power battery system based on the voltage change trend of the charging segments in the third set if the number of charging segments in the third set meets the fourth set condition. If so, an early warning is issued.

[0112] Example 3

[0113] This embodiment proposes a vehicle that includes the early warning and diagnostic system for abnormal battery capacity consistency described in Embodiment 2.

[0114] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A method for early warning and diagnosis of abnormal battery capacity consistency, characterized in that, include: Acquire historical charging data of the target power battery system within a preset time period; The historical charging data is segmented based on the first set condition to obtain multiple consecutive charging segments; The charging segments in which the remaining power battery charge is higher than a set power threshold are identified from each of the charging segments and constitute the first set; If the number of charging segments in the first set is greater than a set threshold, then the charging segments that meet the second set condition are identified from the first set to form the second set. The charging segments that meet the third set are identified from the second set and a third set is formed. If the number of charging segments in the third set meets the fourth set condition, then based on the voltage change trend of the charging segments in the third set, it is determined whether there is an inconsistency in the battery capacity of the target power battery system. If so, an early warning is issued. The first set conditions include: If the time difference between two adjacent charging moments is less than a set time threshold, and the corresponding change in vehicle mileage is less than a set mileage threshold, then the charging data of the two charging moments will be divided into the same charging segment. The second setting conditions include: In all charging records of the first set, the cell with the highest single-cell voltage is the cell with the largest proportion of charging records of the same cell, and is regarded as a suspected abnormal cell, and the proportion is greater than a first proportion threshold. The suspected abnormal cell is the charging segment in which the proportion of time it is the highest single-cell voltage cell in the entire charging segment is greater than the second proportion threshold. The third setting condition includes: The charging segment in which the voltage difference between the suspected abnormal cell and the average voltage of the other cells is greater than a set voltage difference threshold. The fourth setting condition includes: The proportion of the number of charging segments in the third set to the number of the second set is greater than the third proportion threshold, and the proportion of the number of charging segments in the third set to the number of the first set is greater than the fourth proportion threshold.

2. The method for early warning and diagnosis of abnormal battery capacity consistency according to claim 1, characterized in that, The historical charging data includes: Multiple charging times, the voltage of each individual cell at each charging time, the cell number with the highest individual voltage, the remaining power capacity of the power battery, and the vehicle's mileage.

3. The method for early warning and diagnosis of abnormal battery capacity consistency according to claim 1, characterized in that, The determination of whether there is an inconsistency in the battery capacity of the target power battery system based on the voltage change trend of the charging section in the third set includes: Select the last few adjacent charging segments in time sequence from the third set; For each of the charging segments, calculate the first voltage change slope of the suspected abnormal cell at the last multiple consecutive time sampling points, and the second voltage change slope of the average voltage of the remaining cells. If at least one of the charging segments simultaneously meets the following conditions, then the consistency anomaly is determined to exist: The slope of the first voltage change is greater than zero; The slope of the second voltage change is less than zero and less than a set slope threshold; The average voltage of the remaining cells showed a continuous decreasing trend in the last several consecutive sampling points; The cell with the highest single-cell voltage in the last charging segment of the first set is the suspected abnormal cell.

4. The method for early warning and diagnosis of abnormal battery capacity consistency according to claim 1, characterized in that, The set power threshold is 98%.

5. A warning and diagnostic system for abnormal battery capacity consistency, characterized in that, include: The acquisition module is used to acquire historical charging data of the target power battery system within a preset time period; The segmentation module is used to segment the historical charging data based on a first set condition to obtain multiple continuous charging segments. The first identification module is used to identify charging segments from each of the charging segments where the remaining power battery charge is higher than a set power threshold, and to form a first set; The second identification module is used to identify charging segments that meet the second set conditions from the first set if the number of charging segments in the first set is greater than a set number threshold, and to form a second set. The third identification module is used to identify charging segments that meet the third set conditions from the second set, and form the third set; The judgment and warning module is used to determine whether there is an inconsistency in the battery capacity of the target power battery system based on the voltage change trend of the charging segments in the third set if the number of charging segments in the third set meets the fourth set condition. If so, a warning is issued. The first set conditions include: If the time difference between two adjacent charging moments is less than a set time threshold, and the corresponding change in vehicle mileage is less than a set mileage threshold, then the charging data of the two charging moments will be divided into the same charging segment. The second setting conditions include: In all charging records of the first set, the cell with the highest single-cell voltage is the cell with the largest proportion of charging records of the same cell, and is regarded as a suspected abnormal cell, and the proportion is greater than a first proportion threshold. The suspected abnormal cell is the charging segment in which the proportion of time it is the highest single-cell voltage cell in the entire charging segment is greater than the second proportion threshold. The third setting condition includes: The charging segment in which the voltage difference between the suspected abnormal cell and the average voltage of the other cells is greater than a set voltage difference threshold. The fourth setting condition includes: The proportion of the number of charging segments in the third set to the number of the second set is greater than the third proportion threshold, and the proportion of the number of charging segments in the third set to the number of the first set is greater than the fourth proportion threshold.

6. A vehicle, characterized in that, The vehicle includes the early warning and diagnostic system for abnormal battery capacity consistency as described in claim 5.

Images