A new energy automobile risk identification model based on capacity abnormal degradation characteristics

Abstract

The application relates to the technical field of risk identification, and discloses a new energy automobile risk identification model based on capacity abnormal recession characteristics, which comprises a data acquisition module, a data processing module, a data analysis module and a risk identification module; the data acquisition module is used for collecting basic data; the data processing module is used for preprocessing the basic data and obtaining target data; the data analysis module is used for extracting safety elements from the target data and converting the safety elements into quantitatively represented safety characteristics; and when the safety elements are extracted, first, a reference battery cell is selected according to a selection strategy, and a median pressure difference vector V p of the reference battery cell is calculated, a velocity vector V p of V p is taken as a safety element S f ; and the risk identification module is used for determining a risk probability according to the safety characteristic value. The application can realize quantitative identification of new energy automobile risks, and the risk identification precision is high.

Description

Technical Field

[0001] This invention relates to the field of risk identification technology, and specifically to a risk identification model for new energy vehicles based on the characteristics of abnormal capacity decline. Background Technology

[0002] With the advent of the era of electric vehicles, ensuring "safety" has become a primary concern. As a core component of new energy vehicles, the safety of the power battery directly affects the operational safety of the entire vehicle. Many risks associated with new energy vehicles also stem from inherent risks in the power battery system.

[0003] Therefore, the need for accurate quantitative estimation of risks in new energy vehicles and power battery systems, as well as the protection of safety structures, is becoming increasingly prominent. Accurate quantitative estimation of the safety status of power batteries is crucial for improving the durability, safety, and reliability of new energy vehicles.

[0004] Currently, domestic and international assessment methods for the safety of new energy vehicle batteries primarily rely on qualitative analysis of risks, such as expert scoring, which depends on personal experience and subjective judgment, with limited quantitative methods. Furthermore, the national standard GB / T32960.3-2016 defines alarm fault levels as manufacturer-defined. Related data assessments have revealed that such customized solutions cannot accurately identify existing and potential risks of new energy vehicle power batteries in real-world applications, thus failing to guarantee the safety of new energy vehicles. In addition, the increasing diversity of functions and the growing variability and complexity of operating environments in existing new energy vehicles make risk assessment based on vehicle operational big data increasingly difficult. The complexity and strong coupling of operational data render conventional data analysis methods ineffective in accurately completing risk analysis.

[0005] Furthermore, due to the diverse types of faults and risks associated with power batteries, such as abnormal connections, abnormal self-discharge, abnormal capacity, and abnormal internal resistance, the occurrence of these faults will directly cause abnormal fluctuations in the operating data of the power battery. Moreover, the similarity between these fluctuations is very high, which makes it difficult to identify the fault type and risk type, thus making it impossible to accurately eliminate potential risks. As a result, there are still significant hidden dangers in the operation safety of new energy vehicles. Summary of the Invention

[0006] The present invention aims to provide a risk identification model for new energy vehicles based on the characteristics of abnormal capacity decline, which can effectively quantify and identify the risks of new energy vehicles with high accuracy.

[0007] The basic solution provided by this invention is: a risk identification model for new energy vehicles based on the characteristics of abnormal capacity decline, including a data acquisition module, a data processing module, a data analysis module, and a risk identification module;

[0008] The data acquisition module is used to collect basic data, which is the historical operating data of the power battery; the data processing module is used to preprocess the basic data and obtain the target data.

[0009] The data analysis module is used to extract safety elements from the target data and convert the safety elements into quantitatively represented safety features. When extracting safety elements, a reference cell is first selected according to a selection strategy, and the median voltage difference vector V of the reference cell is calculated. p Take V p velocity vector V p As a safety element S f ;

[0010] The risk identification module is used to determine the risk probability based on the security feature values.

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

[0012] This invention provides a risk identification model for new energy vehicles based on abnormal capacity degradation characteristics. Based on historical operating data of power batteries, the model extracts safety elements and transforms safety features through a data analysis module. Specifically, by constructing and identifying abnormal capacity degradation characteristics through the data analysis module, the model can calculate and quantify the risk characteristics of abnormal capacity degradation during vehicle operation, thereby achieving an accurate determination of the overall condition of the vehicle.

[0013] In particular, this solution first selects the characteristic of abnormal capacity degradation as the benchmark feature for risk quantification assessment. This feature is an important abnormal performance characteristic of the power battery system. Risk identification based on this feature dimension can effectively capture risks and achieve accurate risk identification in a specific dimension. Furthermore, in the complex operating environment of new energy vehicles, facing historical operating data with a lot of information coupling, redundancy, and errors, compared with the conventional approach of intuitively analyzing numerical anomalies, the data-driven safety feature extraction method based on abnormal capacity degradation characteristics and combined with mechanistic knowledge provided in this solution can achieve a quantitative description of the vehicle's risk status from a specific dimension, and is relatively unaffected by the coupling and redundancy in direct data.

[0014] Secondly, this scheme constructs the characteristics of abnormal capacity degradation by extracting safety elements. This scheme selects the velocity vector of the median voltage difference vector as a safety element. This safety element is actually an equivalent representation of the abnormal capacity degradation characteristics. Compared to the conventional approach of judging capacity status through SOC value, this scheme's feature construction (i.e., safety element construction) goes further, starting from the operating mechanism of abnormal capacity degradation. This scheme finds that in the actual operating environment, when abnormal capacity degradation occurs, the cell exhibits an accelerated charging and discharging speed. This phenomenon corresponds to an accelerated rate of voltage increase and decrease, and the velocity vector of the median voltage difference vector is selected as the characteristic representation of the "fast charging and fast discharging" phenomenon. Therefore, the abnormal capacity degradation characteristics can be accurately and equivalently represented by voltage, and further converted into a quantitatively represented safety feature, thus accurately and intuitively displaying the abnormal capacity status of the battery.

[0015] Furthermore, compared to the SOC value scheme, the most basic quantity of this safety element is voltage. The data acquisition difficulty is lower than that of the SOC value (conventional SOC value acquisition requires setting up a test to fully charge and discharge the power battery, the test conditions are difficult to unify and the test time is long, while voltage can be directly acquired). The acquisition accuracy is higher than that of the SOC value (due to the limitations of message protocol standards and test equipment, the acquisition accuracy of SOC is at least 1%, while voltage can be accurate to mV), which helps to improve the detail and accuracy of risk analysis. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the model structure of an embodiment of the new energy vehicle risk identification model based on the characteristics of abnormal capacity decay according to the present invention.

[0017] Figure 2 This is a schematic diagram of cell risk in an embodiment of a new energy vehicle risk identification model based on abnormal capacity degradation characteristics according to the present invention. Detailed Implementation

[0018] The following detailed explanation illustrates the specific implementation methods:

[0019] The basic implementation examples are as follows: Figure 1 As shown: A risk identification model for new energy vehicles based on the characteristics of abnormal capacity decline includes a data acquisition module, a data processing module, a data analysis module, and a risk identification module;

[0020] The data acquisition module is used to collect basic data, which is the historical operating data of the power battery. In this embodiment, the corresponding historical operating data is obtained by parsing the message logs of the power battery system conforming to the GB32960 protocol as the basic data. The data selection conforms to the standard, and the basic data is reliable. Moreover, this embodiment analyzes the power batteries of new energy vehicles operating in complex environments with variable operating conditions. The corresponding historical operating data has characteristics such as multidimensionality, redundancy, heterogeneity, and strong coupling, making it difficult to analyze. The following steps of this solution can accurately and purposefully achieve early determination of sampling anomaly risks from such complex data.

[0021] The data processing module is used to preprocess the basic data and obtain the target data. The preprocessing operations of the data processing module include limiting the boundary values ​​of the data signal, identifying and marking interference pulses, identifying and marking time discontinuities, and performing mean filtering on the basic data.

[0022] Specifically, the limit data signal boundary value is defined as follows: based on the first boundary threshold, remove the voltage signal data and current signal data of the historical operation data that exceed the first boundary threshold. In this embodiment, the first boundary threshold is set to [2.5, 4.25]. This setting can eliminate the interference of abnormal excess data on subsequent risk tracing and help improve the accuracy of risk tracing.

[0023] The identification and marking of interference pulses involves: identifying and judging voltage data in historical operating data; if the difference between the current frame's voltage data and the previous frame exceeds a second boundary threshold, then the data in this frame is marked; the second boundary threshold is set to 3 seconds. d ;s d This represents the standard deviation of the voltage change rate. Identifying and marking time discontinuities involves identifying and judging the timestamp data in historical operation data. If the difference between the current frame's timestamp and the previous frame exceeds a specified threshold, then this frame is marked. The specified threshold can be set to 120 seconds. Mean filtering of the base data involves applying a mean filter to the historical operation data to reduce noise in the base data.

[0024] The data analysis module is used to extract safety elements from the target data and convert the safety elements into quantitatively represented safety features. When extracting safety elements, a reference cell is first selected according to a selection strategy, and the median voltage difference vector V of the reference cell is calculated. p Take V p velocity vector V p As a safety element S f .

[0025] Specifically, the selection strategy includes the following steps:

[0026] Step 1: Based on the target data, calculate the median voltage difference V of the single-cell voltage at any moment d ; Specifically, the median voltage difference V d refers to the difference between the single-cell voltage at any moment and its median, that is, the difference between the single-cell voltage of each battery cell and the median of the voltage data of all battery cells in the battery system at each moment.

[0027] Step 1.1: Calculate the maximum voltage difference V dmax and the minimum voltage difference V dmin of the median voltage difference at any moment;

[0028] Step 1.2: Set V d of the median voltage difference to 0 at the position where 40 < SOC < 70;

[0029] Step 1.3: Set to 0 the part of V d of the median voltage difference that is greater than 0 during discharge and the part that is less than 0 during charging;

[0030] Step 1.4: Set to 0 the part of V d of the median voltage difference that is greater than α· dmax during discharge, and set to 0 the part of V d of the median voltage difference that is less than α· dmin during charging, where α ∈ (0, 1).

[0031] In the above steps 1.1 to 1.4, combined with the abnormal manifestations of the capacity anomaly characteristics (that is, corresponding to fast charging and fast discharging), the voltage is higher at high SOC and lower at low SOC. Based on this, the data in the range of SOC [40, 70] is set to 0 to reduce the error caused by abnormal intermediate data; during discharge, since the abnormal battery cell discharges fast, the median voltage difference should be less than 0, so the part greater than 0 is set to 0, which is also to reduce the error caused by data anomalies. Similarly, α in step 1.4 is equivalent to an abnormal determination threshold. Based on the maximum and minimum median voltage differences as the boundary, on the basis of step 1.3, some possible anomalies with small anomalies are excluded. Based on the above steps, this solution can effectively exclude the errors that may be caused by data anomalies, including relatively small errors, which helps to improve the accuracy and reliability of subsequent reference battery cell selection and safety factor extraction.

[0032] Step 2: Amplify the signal of V d of the median voltage difference, and calculate the discrete integral of the median voltage difference of each battery cell in the power battery, and obtain the integral value vector V dm of each battery cell;

[0033] Step 3: Sort V dm of each battery cell, and take the battery cell where the maximum V dm value is located as the reference battery cell.

[0034] The reference cell selected through the above steps is the cell in the power battery system that best reflects the fluctuation of characteristics (identification elements). Compared with the general analysis of all cells, this solution effectively limits the analysis scope to a single reference cell, greatly reducing the workload of data processing. Moreover, the selected cell is highly representative and can effectively make up for the analysis accuracy gap after the analysis is simplified.

[0035] When transforming safety elements, variance entropy is used to quantify the safety elements, and the quantified safety characteristics λ = E are obtained. 2 (Sf) / ( 2 ), and 0≤λ≤1. Specifically, the closer λ is to 1, that is, the smaller p is, the smaller the characteristic fluctuation is on the time scale, and the safer the battery state is.

[0036] The risk identification module is used to determine the risk probability based on the numerical values ​​of safety features. When determining the risk probability based on the numerical values ​​of safety features, the risk identification module first lets p = 1 - λ be the risk quantification feature. It then performs a discrete integral on the risk quantification feature over a time scale to obtain a discrete integral function, and takes the slope value of the function curve as the risk probability judgment. The magnitude of the slope value of the function curve is taken as the risk probability value, and the larger the magnitude, the higher the degree of risk.

[0037] Furthermore, the risk probability determined by the risk identification module is the early risk probability of an abnormal capacity degradation fault. This early risk probability refers to the risk probability during the evolution of an abnormal capacity degradation fault or in its early stages of formation. In other words, this solution can achieve early predictive risk identification of abnormal capacity degradation faults. Specifically, this solution establishes directional risk identification features based on the characteristics of abnormal capacity degradation. These features actually describe the intuitive mechanism of abnormal capacity degradation. The velocity vector of the median voltage difference vector can intuitively quantify whether the "fast charging and fast discharging" phenomenon of the abnormal capacity degradation mechanism exists, rather than being limited to simple voltage and current values. It can accurately capture the evolution process of abnormal capacity degradation from a mechanistic perspective, thereby achieving early and accurate identification of the fault.

[0038] In addition, as attached Figure 2The image shown is an example of cell risk analysis obtained using this model. The moment when the calculated risk probability value exceeds the probability threshold is designated as the high-risk point. This allows for the location of individual cell voltage, current, and other information from 2000 sampling points before and after the high-risk moment. Combined with the image, it can be quickly determined that the problematic cell (cell #64) exhibits abnormal capacity degradation (manifested as a higher voltage in the high SOC segment and a lower voltage in the low SOC segment compared to other cells during discharge; the voltage drop of this cell clearly intersects with that of other cells throughout the discharge process; simultaneously, during charging, due to abnormal capacity degradation, the problematic cell reaches the charging cutoff voltage faster under the same charging conditions, causing the BMS to control the charging process to end, resulting in the overall voltage of other cells being lower than that of the problematic cell at the end of charging). Furthermore, the model can accurately identify the point of occurrence of the anomaly. In addition to accurately identifying risks, this model also helps in risk tracing and determination.

[0039] This embodiment provides a new energy vehicle risk identification model based on abnormal capacity decay characteristics. By constructing and identifying abnormal capacity decay characteristics through a data analysis module, it can complete the calculation and safety quantification of abnormal capacity decay risk characteristics during vehicle operation, thereby achieving accurate judgment of the overall vehicle status.

[0040] Furthermore, this solution specifically selects the velocity vector of the median voltage difference vector as a safety element. Starting from the operational mechanism of abnormal capacity degradation, specifically, when a problematic cell in the power battery system experiences abnormal capacity degradation, its energy storage decreases, causing it to reach the charging cutoff voltage faster under the same charging conditions. At this point, the BMS system will control the charging process to end, resulting in the voltage of the remaining cells being lower than that of the problematic cell at the end of charging. During discharge, because the problematic cell has a higher initial voltage but lower energy storage, its voltage drops faster under the same discharge conditions. In summary, cells with abnormal capacity degradation exhibit a "fast charging, fast discharging" phenomenon in terms of voltage. Based on this operational mechanism, this solution specifically selects the velocity vector of the median voltage difference vector as a characteristic representation of the "fast charging, fast discharging" phenomenon, thus accurately representing the characteristics of abnormal capacity degradation. When facing highly complex and strongly coupled operational data, the numerical relationship mined by this feature (safety element) is also a fundamental mechanistic relationship, less susceptible to the influence of complex data appearances, enabling precise analysis under complex conditions.

[0041] The above descriptions are merely embodiments of the present invention. Commonly known structures and characteristics of the solutions are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent.

Claims

1. A risk identification model for new energy vehicles based on abnormal capacity degradation characteristics, characterized in that, It includes a data acquisition module, a data processing module, a data analysis module, and a risk identification module; The data acquisition module is used to collect basic data, which is the historical operating data of the power battery; the data processing module is used to preprocess the basic data and obtain the target data. The data analysis module is used to extract safety elements from the target data and convert the safety elements into quantitatively represented safety features. When extracting safety elements, a reference cell is first selected according to a selection strategy, and the median voltage difference vector of the reference cell is calculated. ,Pick velocity vector As a safety element ; The selection strategy includes the following steps: Step 1: Based on the target data, calculate the median voltage difference of each individual cell at any given time. ; Step 1.1: Calculate the maximum pressure difference value of the median pressure difference at any given time. and minimum pressure difference ; Step 1.2: Set the median pressure difference exist Set the position to 0; Step 1.3: Adjust the median pressure difference. During the discharge process, the portion greater than 0 and during the charging process, the portion less than 0 is set to 0; Step 1.4: Adjust the median pressure difference. During the discharge process, greater than Set part of the median pressure difference to 0. Less than during charging Part of it is set to 0, where ; Step 2: Analyze the median pressure difference. The signal is amplified, and the median voltage difference discrete integral of each cell in the power battery is calculated to obtain the integral value vector of each cell. ; Step 3: For each cell Sort the data and take the largest value. The cell containing the value is the reference cell; The risk identification module is used to determine the risk probability based on the security feature values.

2. The new energy vehicle risk identification model based on abnormal capacity degradation characteristics according to claim 1, characterized in that, When transforming safety elements, variance entropy is used to quantify the safety elements and obtain the quantified safety characteristics. ,and .

3. The new energy vehicle risk identification model based on abnormal capacity degradation characteristics according to claim 2, characterized in that, When the risk identification module determines the risk probability based on the security feature values, it first sets... To quantify risk characteristics, a discrete integral function is obtained by performing discrete integration on the risk quantification characteristics over the time scale, and the slope value of the function curve is taken to determine the risk probability.

4. The new energy vehicle risk identification model based on abnormal capacity degradation characteristics according to claim 3, characterized in that, The magnitude of the slope of the function curve is used as the risk probability value, and the larger the magnitude, the higher the risk level.

5. A new energy vehicle risk identification model based on abnormal capacity degradation characteristics according to claim 1, characterized in that, The preprocessing operations of the data processing module include limiting the boundary values ​​of the data signal, identifying and marking interference pulses, identifying and marking time discontinuities, and performing mean filtering on the basic data.

6. The new energy vehicle risk identification model based on abnormal capacity degradation characteristics according to claim 1, characterized in that, The risk probability determined by the risk identification module is the early risk probability of an abnormal capacity degradation failure.

Images