A method for risk assessment of abnormal connection of power batteries

Abstract

The present invention relates to the field of battery anomaly assessment methods, and particularly to a method for assessing the risk of abnormal connection of power batteries, including obtaining the operation historical data of a vehicle for preprocessing; intercepting the charging state data, calculating the median pressure difference, constructing a sign function, calculating the cumulative value of the connection anomaly characteristic value of each battery cell, and extracting the safety elements of the connection anomaly risk characteristics; using variance entropy to quantify the safety characteristics of internal resistance consistency; constructing a risk quantification characteristic, performing discrete integration on the risk quantification characteristic on the time scale to obtain a risk curve, and extracting the absolute risk probability; defining the moment when the absolute risk probability exceeds a specified threshold as a high-risk point and drawing the image of the safety elements within the local time before and after the high-risk point, and combining the local characteristics of the change of the safety elements in the image to achieve the determination of the safety state. The present invention realizes the intelligent identification of the safety state and the determination of the abnormal connection fault mode based on the operation historical data of the vehicle.

Description

Technical Field

[0001] The present invention relates to the field of battery anomaly assessment methods, and particularly to a method for assessing the risk of abnormal connection of power batteries. Background Art

[0002] With the adjustment of the national energy structure, the rapid development of technologies such as batteries, motors, and electronic controls, the transformation of consumer demands, and the introduction and improvement of relevant industrial policy supports. In recent years, the production and sales volume of new energy vehicles in China have been increasing continuously. However, the actual operating environment of new energy vehicles is changeable and the scenarios are complex. The data characterizing their safety status has characteristics such as multi-dimensionality, redundancy, heterogeneity, and strong coupling, which brings great challenges to exploring and mining the vehicle safety characteristics contained in the operating big data. It is difficult to achieve large-scale application of the state assessment of the power batteries of new energy vehicles with strong coupling complex systems through mechanism analysis. Therefore, a data-driven safety feature extraction method combining mechanism knowledge can quantitatively describe and batch apply the vehicle risk status from a specific dimension.

[0003] Since vehicles will upload operating data according to GB-32960 during their service life, and voltage, current and other characterization signals are a kind of characterization quantity of the comprehensive state of the battery, it is possible to analyze and mine the safety characteristics of the vehicle's operating historical data to further quantify the difference degree of the state of the power battery during its service life. However, considering the physical characteristics of the battery system itself and factors such as the design and acquisition accuracy of sensors, there must be information coupling, redundancy, and errors between different signals, making it difficult to accurately extract and quantify safety characteristics; in addition, the causes of battery failures are diverse, and the operation of judging the causes of failures is complex, and it is impossible to directly, quickly, and accurately judge which anomaly it is. Summary of the Invention

[0004] The present invention aims to provide a method for assessing the risk of abnormal connection of power batteries, so as to complete the calculation and safety quantification of the abnormal connection risk characteristics during the vehicle operation process by constructing and identifying the abnormal connection risk characteristics, and thereby achieve the accurate determination and fault identification of the vehicle's comprehensive state, and solve the problem that it is difficult to accurately extract and quantify safety characteristics.

[0005] The method for assessing the risk of abnormal connection of power batteries in this solution includes:

[0006] Step 1, obtaining the operating historical data of the vehicle and preprocessing the operating historical data;

[0007] Step 2: Intercept the charging state data of the preprocessed operation history data, calculate the median voltage difference of the charging state data, construct a sign function based on the median voltage difference, the range voltage, and the connection anomaly characteristic representation coefficient, retain the part representing the connection anomaly characteristic in the median voltage difference according to the sign function and the set model, calculate the cumulative value of the connection anomaly characteristic value of each battery cell according to the preset model, sort the cumulative values, and take the largest cumulative value as the safety factor of the connection anomaly risk characteristic of the reference battery cell;

[0008] Step 3: Calculate the safety quantification characteristic based on the variance entropy of the connection anomaly risk characteristic safety factor;

[0009] Step 4: Construct a risk quantification characteristic according to the safety quantification characteristic, perform discrete integration on the risk quantification characteristic on the time scale to obtain a risk curve, take the slope value of the risk curve as the safety state identification, and take the amplitude corresponding to the slope value as the absolute risk probability corresponding to the internal resistance consistency safety characteristic;

[0010] Step 5: Define the moment when the absolute risk probability exceeds the specified threshold as a high-risk point and draw the image of the safety factor within the local time before and after the high-risk point, and combine the local characteristics of the change of the safety factor in the image to achieve the determination of the safety state.

[0011] The beneficial effects of this solution are as follows:

[0012] Through the representation characteristics of battery connection anomalies in voltage signal data, the data expression and automatic recognition of fault information are effectively realized. Furthermore, by constructing steps of data processing, safety factor extraction, safety characteristic quantification, safety state identification, and risk traceability determination, a technical system for new energy vehicle safety state identification and fault mode determination is constructed to realize intelligent identification of safety states and determination of connection anomaly fault modes based on vehicle operation history data.

[0013] Furthermore, the preprocessing includes the following steps:

[0014] a) Signal boundary value limitation: Remove abnormal data in voltage and current signal data that exceeds the specified threshold;

[0015] b) Interference pulse identification and marking: If the difference between the current frame voltage data and the previous frame exceeds the specified threshold, mark this frame of data;

[0016] c) Time discontinuity point identification and marking: If the difference between the current frame timestamp data and the previous frame exceeds the specified threshold, mark this frame of data;

[0017] d) Mean filtering.

[0018] The beneficial effects are as follows: Through the preprocessing of the operation historical data, the signal boundary value limitation can remove individual significantly abnormal data, identify and mark interference pulses and time discontinuity points, facilitating subsequent processing. It can standardize the operation historical data from multiple aspects, effectively prevent extremely individual abnormal data from interfering with the processing and analysis results, facilitate subsequent processing, and improve the accuracy of subsequent data processing.

[0019] Further, in step 2, according to the magnitude relationship between the median pressure difference and the product of the range voltage and the connection anomaly characteristic representation coefficient, a set value is obtained to construct a sign function, which is denoted as Vdm. Specifically:

[0020]

[0021] Wherein, Vaa is the range voltage, which is the difference between the maximum value of the single-cell voltage and the minimum value of the single-cell voltage at any moment, and α is the connection anomaly characteristic representation coefficient, and α ∈ (0, 1).

[0022] The beneficial effects are as follows: By constructing a sign function with multiple parameters, it can accurately retain the characteristics and information indicating the abnormality of the vehicle battery in the operation historical data, improving the integrity and accuracy of the risk assessment of vehicle battery connection anomalies.

[0023] Further, in step 2, the set model is expressed as:

[0024]

[0025] Wherein, Vd represents the median voltage.

[0026] The beneficial effects are as follows: Setting the part with weak representation ability to 0 can effectively reduce the interference error of the data itself fluctuation accumulation effect on the determination result.

[0027] Further, in step 2, the preset model is expressed as:

[0028]

[0029] Wherein, SC i is the connection anomaly characteristic value of the i-th battery cell, n is the number of sampling points, I is the total current, and V di is the median pressure difference of the i-th battery cell, and ε is the marking vector with the sign function value of 1.

[0030] The beneficial effects are as follows: Through the design of the preset model, by calculating the cumulative value of the connection anomaly characteristic values of each battery cell, it can amplify the characterization situation when the battery cell connection is abnormal, improving the accuracy of the risk assessment of battery connection anomalies.

[0031] Further, in step 3, the connection anomaly risk characteristic safety element is denoted as Sf, the safety quantification characteristic is denoted as λ, and substituting the connection anomaly risk characteristic safety element into the variance entropy calculation formula, the safety quantification characteristic is obtained as:

[0032] λ = E 2 (Sf) / E(Sf 2 );

[0033] where 0 ≤ λ ≤ 1.

[0034] The beneficial effect is that quantifying the connection anomaly risk characteristics of the battery can more intuitively represent the connection anomaly risk situation of the battery.

[0035] Further, in step 4, the risk quantification characteristic is denoted as p = 1 - λ;

[0036] The risk curve is denoted as: Sp = ∑p.

[0037] The beneficial effect is that through the representation of the risk quantification characteristic and the risk curve, the risk situation can be intuitively, accurately and quickly represented. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic block diagram of an embodiment of the method for evaluating the connection anomaly risk of the power battery of the present invention;

[0039] Figure 2 It is an image of a high-risk battery cell in an embodiment of the method for evaluating the connection anomaly risk of the power battery of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0040] The following is a further detailed description through specific embodiments.

[0041] Embodiment

[0042] As Figure 1 shown, the method for evaluating the connection anomaly risk of the power battery includes:

[0043] Step 1: Obtain the operation history data of the vehicle. The operation history data is obtained by retrieving from the server or memory after the vehicle reports, and preprocess the operation history data. The preprocessing includes the following steps:

[0044] a) Signal boundary value limitation: Remove abnormal data where the voltage and current signal data exceed the specified threshold;

[0045] b) Interference pulse identification and marking: If the difference between the current frame voltage data and the previous frame exceeds the specified threshold, mark this frame of data;

[0046] c) Time discontinuity identification and marking: If the difference between the current frame timestamp data and the previous frame exceeds the specified threshold, the frame data is marked;

[0047] d) Mean filtering: Mean filtering is an existing algorithm and will not be described in detail here.

[0048] Step 2, intercept the charging state data of the pre-processed running history data, calculate the median voltage difference of the charging state data, expressed as Vd, and the calculation method of the median voltage difference is to subtract the median voltage from the single cell voltage at any time. Construct a sign function based on the median voltage difference, the extreme voltage difference and the connection abnormality characteristic characterization coefficient, and obtain the set value to construct a sign function based on the size relationship between the median voltage difference and the product of the extreme voltage difference and the connection abnormality characteristic characterization coefficient. The sign function is expressed as Vdm, which is specifically expressed as:

[0049]

[0050] Wherein, Vaa is the extreme difference voltage, which is the difference between the maximum value of the single cell voltage and the minimum value of the single cell voltage at any time, and α is the connection abnormality characteristic characterization coefficient, α∈(0,1).

[0051] According to the sign function and the setting model, the part of the median voltage difference that represents the abnormal connection characteristics is retained, and the setting model is expressed as:

[0052]

[0053] Wherein, Vd represents the median voltage.

[0054] The cumulative value of the abnormal connection characteristic value of each battery cell is calculated according to the preset model. The preset model is expressed as:

[0055]

[0056] Among them, SC i is the abnormal connection characteristic value of the ith battery cell, n is the number of sampling points, I is the total current, V di is the median voltage difference of the ith battery cell, and ε is a marker vector with a sign function value of 1. The accumulated values ​​are sorted, and the largest accumulated value is taken as the connection abnormality risk characteristic safety factor of the reference battery cell, and the connection abnormality risk characteristic safety factor is represented as Sf.

[0057] Step 3: Based on the variance entropy calculation of the connection abnormal risk feature safety factor, the safety quantitative feature is expressed as λ, and the connection abnormal risk feature safety factor is substituted into the variance entropy calculation formula to obtain the safety quantitative feature:

[0058] λ=E 2 (Sf) / E(Sf 2 );

[0059] Among them, 0 ≤ λ ≤ 1. By limiting the range of the values of the safety quantization features, it is possible to uniformly evaluate and assess different data subsequently while retaining the information representing the safety features in the data, so as to accurately evaluate the specific connection anomaly risk of the battery subsequently.

[0060] Step 4: Construct a risk quantization feature based on the safety quantization feature. Represent the risk quantization feature as p = 1 - λ, and perform a discrete integral on the risk quantization feature over the time scale to obtain a risk curve, which is represented as: Sp = ∑p.

[0061] Take the slope value of the risk curve for safety state identification, and take the amplitude corresponding to the slope value as the absolute risk probability corresponding to the internal resistance consistency safety feature.

[0062] Step 5: Define the moment when the absolute risk probability exceeds the specified threshold as a high-risk point and draw an image of the safety elements within the local time before and after the high-risk point. Combine the local characteristics of the changes in the safety elements in the image to achieve safety state determination. For example, judge the safety state from the manifestations of the local characteristics of the changes in the safety elements in the image, such as the connection anomaly risk probability, the large voltage fluctuation of abnormal battery cells during driving, and the higher voltage of abnormal battery cells during charging.

[0063] Taking the single-cell voltage, current and other information of a total of 2000 sampling points before and after the high-risk moment included in the high-risk image information as an example for simulation, the result is obtained as Figure 2 shown. It can be seen that compared with other battery cells, the abnormal battery cell has a larger fluctuation amplitude during driving, and during parking charging, the voltage is significantly higher than that of the other battery cells, that is, there is an obvious phenomenon of higher charging and lower discharging of the voltage. This is because the DC internal resistance of the battery cell with a connection anomaly increases, resulting in an increased current oscillation amplitude during the state switching between discharging and braking power feedback during driving, and during parking charging, the ohmic internal resistance of the abnormal battery cell increases, and the voltage division increases under the condition of a constant current.

[0064] In this embodiment, when the abnormal fault characteristic of the power battery connection is manifested as an abnormal battery cell voltage with obvious high charging and low discharge phenomenon, since the power battery system forms an electrical connection between the battery cell and the module through a connector (bolting, welding, etc.), during the operation of the vehicle, vehicle vibration, collision and other phenomena will cause problems such as loose bolts or broken welding points, which will lead to virtual connection of the connector. The data is manifested as an abnormal increase in the DC internal resistance of the battery cell adjacent to the virtual connection position, which is obtained by the calculation formula of the DC internal resistance DCR=ΔU / ΔI. It can be seen that during the charging process, the voltage of the abnormal battery cell will be larger, and during the discharge process, when the vehicle switches between the acceleration and braking states, the voltage fluctuation amplitude of the abnormal battery cell will be more severe. Based on the above-mentioned fault characteristic principle, the connection abnormality feature based on the operation history data is designed to realize data-driven vehicle abnormality feature recognition, thereby completing the determination of the vehicle safety state and the identification of the fault mode.

[0065] Therefore, this embodiment constructs a new energy vehicle safety status identification and failure mode determination technology system by constructing data processing, safety factor extraction, safety feature quantification, safety status identification, and risk tracing and determination steps, so as to realize intelligent identification of safety status and connection abnormality failure mode determination based on vehicle operation history data, accurately evaluate the battery connection abnormality information from the operation history data, and make the battery connection abnormality risk assessment more complete and accurate.

[0066] The above is only an embodiment of the present invention, and the common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for risk assessment of abnormal connection of power batteries, characterized in that, Including: Step 1: Obtain the operation history data of the vehicle and preprocess the operation history data; Step 2: Intercept the charging state data of the preprocessed operation history data, calculate the median voltage difference of the charging state data, construct a sign function based on the median voltage difference, the range voltage, and the connection anomaly characteristic representation coefficient, retain the part representing the connection anomaly characteristic in the median voltage difference according to the sign function and the set model, calculate the cumulative value of the connection anomaly characteristic value of each battery cell according to the preset model, sort the cumulative values, and take the largest cumulative value as the safety factor of the connection anomaly risk characteristic of the reference battery cell; Step 3: Calculate the safety quantification characteristic based on the variance entropy of the connection anomaly risk characteristic safety factor; Step 4: Construct a risk quantification characteristic according to the safety quantification characteristic, perform discrete integration on the risk quantification characteristic on the time scale to obtain a risk curve, take the slope value of the risk curve as the safety state identification, and take the amplitude corresponding to the slope value as the absolute risk probability corresponding to the internal resistance consistency safety characteristic; Step 5: Define the moment when the absolute risk probability exceeds the specified threshold as a high-risk point and draw an image of the safety factor within the local time before and after the high-risk point, and combine the local characteristics of the change of the safety factor in the image to achieve the determination of the safety state.

2. The method for evaluating the risk of abnormal connection of power batteries according to claim 1, wherein: The preprocessing includes the following steps: a) Signal boundary value limitation: Remove abnormal data where the voltage and current signal data exceed the specified threshold; b) Interference pulse identification and marking: If the difference between the current frame voltage data and the previous frame exceeds the specified threshold, mark this frame of data; c) Time discontinuity point identification and marking: If the difference between the current frame timestamp data and the previous frame exceeds the specified threshold, mark this frame of data; d) Mean filtering.

3. The method for evaluating the risk of abnormal connection of power batteries according to claim 2, wherein: In the said step 2, according to the magnitude relationship between the median pressure difference and the product of the range voltage and the connection anomaly characteristic representation coefficient, a set value is obtained to construct and form a sign function, and the sign function is expressed as Vdm , specifically: ； Among them, Vd represents the median voltage, Vaa is the range voltage, and the range voltage is the difference between the maximum value of the single-cell voltage and the minimum value of the single-cell voltage at any moment, α is the connection anomaly characteristic representation coefficient, α ∈(0, 1).

4. The method for evaluating the risk of abnormal connection of power batteries according to claim 3, characterized in that: In Step 2, the set model is expressed as: ； Among them, Vd represents the median voltage.

5. The method for evaluating the risk of abnormal connection of power batteries according to claim 3, characterized in that: In Step 2, the preset model is expressed as: ； Among them, is the connection anomaly eigenvalue of the i th battery cell, n is the number of sampling points, I is the total current, is the median voltage difference of the i th battery cell, ε is the marking vector with the sign function value of 1.

6. The method for evaluating the risk of abnormal connection of power batteries according to claim 4, wherein: In step 3, the security element of the connection anomaly risk feature is represented as Sf , the security quantification feature is represented as λ , and substituting the security element of the connection anomaly risk feature into the variance entropy calculation formula, the security quantification feature is obtained as: ； Among them, 0 ≤ λ ≤ 1.

7. The method for evaluating the risk of abnormal connection of power batteries according to claim 6, wherein: In step 4, the risk quantification feature is expressed as p = 1 - λ ; The risk curve is expressed as: Sp = ∑ p .

Images