Power battery safety state recognition system based on equivalent circuit least square error

Abstract

The application relates to power battery safety detection technology and discloses a power battery safety state recognition system based on equivalent circuit least square error, which comprises a processor module and a data acquisition module, a data processing module and a state recognition module connected with the processor module respectively, the data acquisition module is used for collecting operation data of the power battery to form a data set and then sending the data set to the data processing module; the data processing module is used for pre-processing the data set to obtain a selected data set; the processor module is used for extracting safety elements according to the selected data set and performing safety feature quantization to obtain safety quantization features; and the state recognition module is used for recognizing the safety state of the power battery according to the safety quantization features. The application has the beneficial effects of improving the accuracy of power battery fault analysis and risk traceability results and guaranteeing the operation safety of new energy vehicles.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power battery safety detection technology, and particularly relates to a power battery safety state recognition system based on equivalent circuit least square error. BACKGROUND

[0002] With the development of new energy automobile technology, the number of new energy automobiles on the market is also increasing, and the use rate of new energy automobiles is gradually increasing. Therefore, the problems of new energy automobiles have gradually been exposed, especially in recent years, the occurrence of new energy automobile fire accidents has caused great trouble to the use of new energy automobiles. There are many reasons for the new energy automobile fire, but the most important one is the safety state of the power battery. The power battery is the power source of the new energy automobile, and a large amount of heat will be generated during its working process, which will greatly threaten the safety of the power battery.

[0003] Therefore, to ensure the driving safety of new energy automobiles, the most important thing is to accurately analyze and identify the safety state of the power battery. At present, the means for identifying the safety state of the power battery are all based on the analysis of the historical operation data of the power battery. However, due to the changeable driving environment and complex scene of new energy automobiles, the operation data representing the safety state has the characteristics of multi-dimension, redundancy, heterogeneity and strong coupling, which leads to the fact that the accuracy of the ordinary analysis method based on voltage and current cannot be effectively guaranteed. Therefore, there is an urgent need for a method that can accurately identify the safety state of the power battery. SUMMARY

[0004] The present application aims to provide a power battery safety state recognition system based on equivalent circuit least square error to improve the accuracy of the power battery safety state recognition result.

[0005] To achieve the above-mentioned purpose, the present application adopts the following technical solution: a power battery safety state recognition system based on equivalent circuit least square error, comprising a processor module, and a data acquisition module, a data processing module and a state recognition module connected with the processor module respectively;

[0006] The data acquisition module is used for collecting the operation data of the power battery to form a data set and sending the data set to the data processing module;

[0007] The data processing module is used for pre-processing the data set to obtain a selected data set;

[0008] The processor module comprises a storage unit and an analysis unit. The storage unit is used for storing the pre-processed selected data set. The analysis unit is used for extracting safety elements from the selected data set and quantifying safety features to obtain safety quantitative features.

[0009] a state recognition module configured to recognize a safety state of the power battery according to the safety quantification feature.

[0010] The principle and advantages of the scheme are as follows: in actual application, basic operation data of the power battery are collected, the data are cleaned and preprocessed to improve the effectiveness of the data, the least square fitting error is derived and calculated according to the collected data and an equivalent circuit of the power battery, the safety element vector is obtained by performing nonlinear mapping and signal amplification on the fitting error, the absolute risk probability of the consistency safety feature is obtained according to the safety element, and the safety state of the power battery is analyzed and the safety risk thereof is traced according to the safety element image. Compared with the prior art, the scheme has the advantages that the power battery operation safety state equation is constructed based on the linear relationship between voltage and current in the equivalent circuit model of the power battery, the equation hyperparameters are solved by combining standard data and the least square method fitting, the obtained least square fitting error is taken as the safety element, the safety feature of the consistency state between different battery cells is quantified, the safety state of the power battery can be accurately and objectively reflected, the safety state of the power battery can be accurately analyzed, and the safety of the power battery is improved, thereby improving the operation safety of the new energy vehicle.

[0011] Preferably, as an improvement, the state recognition module is further configured to obtain a safety element image according to the safety quantification feature, and trace the safety risk of the power battery according to the safety element image.

[0012] Preferably, as an improvement, the pre-processing of the data set includes data signal boundary value limitation, interference pulse identification and marking, time interval point identification and marking, and mean value filtering.

[0013] Preferably, as an improvement, when the safety element is extracted, the equivalent circuit of the power battery is transformed to obtain the least square fitting error, and the safety element vector is obtained by performing nonlinear mapping and signal amplification on the least square fitting error.

[0014] Preferably, as an improvement, the safety quantification feature obtained by performing safety feature quantification is quantified by using variance entropy, and the safety quantification feature is obtained by substituting the safety element vector into a variance entropy calculation formula.

[0015] Preferably, as an improvement, the least square fitting error is:

[0016] Loss=ΔV 2 ΔI 2 -(ΔV T ΔI) 2 , wherein ΔV is a voltage median pressure difference, and ΔI is a mean current difference.

[0017] Preferably, as an improvement, the safety factor vector is Sf=e αLoss wherein, a is a loss amplification coefficient constant.

[0018] Preferably, as an improvement, the variance entropy calculation formula is λ=E 2 (Sf) / E(Sf 2 wherein, 0≤λ≤1.

[0019] Preferably, as an improvement, the safety state recognition includes the following steps:

[0020] Let p=1-λ be the risk quantification feature, and the risk curve Sp=∑p is obtained by discretely integrating the risk quantification feature on the time scale.

[0021] The slope value of the risk curve is taken for safety state recognition, and the amplitude of the identified slope value is taken as the absolute risk probability corresponding to the consistency safety feature.

[0022] Preferably, as an improvement, the safety risk of the power battery is traced back to the time point when the absolute risk probability exceeds the probability threshold, and the safety factor image in the local time before and after the high-risk point is drawn, and the safety risk of the power battery is traced back according to the local feature of the safety factor change in the image. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 It is a structure schematic view of the first embodiment of the safety state recognition system of the power battery based on the least square error of the equivalent circuit of the application.

[0024] Figure 2 It is a recognition flowchart of the first embodiment of the safety state recognition system of the power battery based on the least square error of the equivalent circuit of the application.

[0025] Figure 3 It is a schematic view of risk mode 1 of the first embodiment of the safety state recognition system of the power battery based on the least square error of the equivalent circuit of the application.

[0026] Figure 4 It is a schematic view of risk mode 2 of the first embodiment of the safety state recognition system of the power battery based on the least square error of the equivalent circuit of the application.

[0027] Figure 5 It is a schematic view of risk mode 3 of the first embodiment of the safety state recognition system of the power battery based on the least square error of the equivalent circuit of the application. DETAILED DESCRIPTION

[0028] The following will be further described in detail through specific embodiments:

[0029] The labels in the drawings of the specification include: a data acquisition module 1, a data processing module 2, a processor module 3, a storage unit 4, an analysis unit 5, and a state recognition module 6.

[0030] Embodiment one:

[0031] This embodiment is basically as shown in the accompanying Figure 1 The power battery safety state recognition system based on the least square error of the equivalent circuit includes a processor module 3, and a data acquisition module 1, a data processing module 2, and a state recognition module 6 connected with the processor module 3 respectively.

[0032] The data acquisition module 1 is used to collect the running data of the power battery to form a data set and then send it to the data processing module 2.

[0033] The data processing module 2 is used to pre-process the data set to obtain a selected data set.

[0034] The processor module 3 includes a storage unit 4 and an analysis unit 5. The storage unit 4 is used to store the selected data set after processing. The analysis unit 5 is used to extract safety elements from the selected data set and quantize safety features to obtain safety quantization features.

[0035] The state recognition module 6 is used to recognize the safety state of the power battery according to the safety quantization features.

[0036] As shown in the accompanying Figure 2 After the least square fitting error of the equivalent circuit of the power battery is derived and calculated based on the collected historical running data of the power battery, the safety element vector is obtained by nonlinear mapping and signal amplification of the fitting error, and the absolute risk probability of the consistent safety feature is obtained according to the safety element to complete the safety state analysis, judgment, and risk tracing of the power battery. The specific process is as follows:

[0037] First, the historical running data of the power battery of the target vehicle is collected by using the data acquisition module 1, and then the collected data is pre-processed by using the data processing module 2. Specifically, first, the data signal boundary value is limited to remove abnormal data of voltage and current signal data exceeding the specified threshold value; then, interference pulse recognition and marking are performed, if the difference between the current frame voltage data and the previous frame exceeds the interference pulse judgment threshold, the frame data is marked; then, time breakpoint recognition and marking are performed, if the difference between the current frame timestamp data and the previous frame exceeds the time jump judgment threshold, the frame data is marked; finally, the data is subjected to mean filtering.

[0038] Specifically, in this step, the threshold range of the voltage signal data is 1000-6000 mV; the threshold range of the current signal data is -1000-1000 A; the interference pulse determination threshold is 3 times the standard deviation; and the time jump determination threshold is 150 seconds. Through the limitation of the above threshold values, the pre-processing of the data can be more accurately completed, the accuracy of the data is improved, and then accurate basis is provided for the subsequent calculation of the fitting error of the power battery, and the accuracy of the subsequent analysis result of the safety state of the power battery is ensured.

[0039] Secondly, the pre-processed data is analyzed by the analysis unit 5 of the processor module 3 to extract the safety elements of the data. The product of the current and the voltage is calculated first, and then the mean value of the voltage, the current, and the product of the voltage and the current on the time sequence is calculated in sequence to realize mean filtering. Then, the least square fitting error is calculated according to the equivalent circuit of the power battery. The specific process is as follows:

[0040] It is known that the equivalent circuit model of the power battery is formula (1): V=E-IR

[0041] In the formula, V is the open circuit voltage, E is the electromotive force, I is the current, and R is the equivalent internal resistance.

[0042] The above formula is transformed to obtain formula (2): V-E+IR=0

[0043] Again, transformation can be obtained: That is, formula (3) is obtained:

[0044] Similarly, the following transformation of formula (2) can be obtained:

[0045] In the formula, e is a unit vector, that is,

[0046] Then formula (4) is obtained:

[0047] Substituting formula (4) into formula (3), the following formula is obtained:

[0048] Then formula (5) is obtained:

[0049] The least square fitting is used to solve the equation hyperparameters:

[0050]

[0051] In the formula, LOSS is the least square fitting error;

[0052] Let

[0053] Equation (6) can be obtained: Loss=||ΔV+RΔI|| 2

[0054] Will Δ V and Δ Substituting I into equation (5), we get:

[0055] Substituting into equation (6), we get:

[0056]

[0057] For the same battery pack, the variables related to the current of each cell are the same, so the impedance fitting error can be converted into the mean voltage difference. Δ V and the difference between the mean current Δ The equation for I:

[0058] Loss=ΔV 2 ΔI 2 -(ΔV T ΔI) 2

[0059] In the formula, ΔV represents the median voltage difference. Δ I represents the average current difference.

[0060] Then, a nonlinear mapping and signal amplification are performed on the fitting error to obtain the safety element vector:

[0061] Sf = e αLoss

[0062] In the formula, α is the loss amplification factor constant.

[0063] Third, use variance entropy to quantify the obtained safety element vector, and substitute the safety element vector Sf into the variance entropy calculation formula λ=E. 2 (Sf) / E(Sf 2 The safety quantification feature p is obtained from ), where 0≤λ≤1, and the closer λ is to 1, the smaller the fluctuation of the consistency feature on the time scale, and the safer the battery state.

[0064] Fourth, the safety status of the power battery is identified by the state recognition module 6. First, let p = 1 - λ be the risk quantification feature, and then perform discrete integration on the risk quantification feature over the time scale to obtain the risk curve Sp = ∑p. The risk curve is a monotonically increasing curve. Then, the slope value z of the risk curve is used for safety status identification. The amplitude of the slope value z obtained by identification is used as the absolute risk probability corresponding to the consistent safety feature.

[0065] Specifically, the state recognition module 6 is also used to obtain safety element images based on safety quantification features, and to trace the source of safety risks of the power battery in combination with the safety element images. The moment when the absolute risk probability exceeds the probability threshold is defined as a high-risk point, and safety element images are drawn in the local time before and after the high-risk point. The power battery safety status is determined and the source of safety risks is traced based on the local features of the changes in safety elements in the images.

[0066] Specifically, the probability threshold for the absolute risk probability mentioned above is 0.5. By limiting this probability threshold, high-risk points can be identified more accurately, thereby improving the accuracy of the safety element image and, consequently, the accuracy of the results of power battery safety status and risk tracing.

[0067] As attached Figure 3 As shown, Risk Mode 1: During the discharge process of the abnormal cell shown in the figure, there is an abnormal phenomenon of rapid voltage drop at low SOC. It can be seen that the battery has a problem of poor consistency caused by abnormal cells.

[0068] As attached Figure 4 As shown, Risk Mode 2: During the discharge process, the abnormal cell exhibits an abnormal drop in SOC, indicating a poor battery consistency issue caused by this cell. Furthermore, the abnormal cell charges more slowly and fails to fully charge. If, under these circumstances, a trend of increasing voltage difference occurs during subsequent operation, it can be determined that the vehicle has a self-discharge abnormality fault.

[0069] As attached Figure 5 As shown, Risk Mode 3: Compared to other cells, the abnormal cell in the diagram fluctuates more during driving. During charging while parked, its voltage is significantly higher than other cells, indicating a clear phenomenon of high charging and low discharging. This is because the DC internal resistance of the abnormal cell increases, causing the current oscillation amplitude to increase during the state switching between discharging and braking power feedback while driving. When charging while parked, the ohmic internal resistance of the abnormal cell increases, resulting in a larger voltage drop under constant current.

[0070] The specific implementation process of this embodiment is as follows:

[0071] The first step is to use data acquisition module 1 to collect data packets of the power battery of the new energy vehicle and parse the historical operating data of the power battery. Then, the obtained data is cleaned and preprocessed, and the boundary values ​​of the data signal are limited, interference pulses are identified and marked, and time discontinuities are identified and marked. Finally, the data is subjected to mean filtering.

[0072] The second step involves using the analysis unit 5 of processor module 3 to extract the safety elements of the preprocessed data. First, the product of current and voltage is calculated. Then, the mean values ​​of voltage, current, and the product of voltage and current in the time series are calculated sequentially over the local time window to achieve mean filtering. Next, the least squares fitting error is calculated based on the equivalent circuit of the power battery. Finally, the safety element vector is substituted into the variance entropy formula to quantify the safety element vector and obtain the safety quantification features.

[0073] The third step is to use the state recognition module 6 to identify the safety status of the power battery, obtain the risk quantification feature based on the safety quantification feature, and then perform discrete integration on the risk quantification feature on the time scale to obtain the risk curve. The slope value z of the risk curve is used for safety status identification, and the amplitude of the identified slope value z is used as the absolute risk probability corresponding to the consistent safety feature.

[0074] The fourth step is to determine the safety status and trace the source of risks of the power battery. The moment when the absolute risk probability exceeds the probability threshold is defined as a high-risk point, and safety element images are drawn before and after the high-risk point in a local time period. The safety status of the power battery and the source of safety risks are determined based on the local characteristics of the changes in safety elements in the images.

[0075] In recent years, with the depletion of non-renewable resources such as oil, the research and development of new energy technologies has become increasingly important, serving as a primary choice for solving the energy crisis. For the automotive industry, the development of new energy vehicle technology has ushered in a shift from the era of gasoline-powered vehicles to the era of new energy vehicles. Especially with breakthroughs in power battery technology and the economic and comfort advantages of new energy vehicles, consumer acceptance is rising, and the number of new energy vehicles on the market is increasing, with usage rates gradually rising. However, with the widespread adoption of new energy vehicles, some problems have gradually emerged, particularly the numerous fires and accidents involving new energy vehicles in recent years, causing considerable inconvenience to their use. There are many reasons for fires in new energy vehicles, but the most significant is the power battery. When the power battery is heated, internal chemical reactions can occur, potentially leading to an internal fire. Since the power battery is the power source of new energy vehicles, it generates a large amount of heat during operation, posing a significant threat to its safety.

[0076] Therefore, to ensure the driving safety of new energy vehicles, the most important thing is to accurately analyze and identify the safety status of the power battery. Currently, the methods for identifying the safety status of the power battery are all based on the analysis of the historical operating data of the power battery. However, due to the variable and complex driving environment of new energy vehicles, the operating data that characterizes its safety status has the characteristics of multi-dimensionality, redundancy, heterogeneity and strong coupling. Therefore, if we only analyze based on the working voltage and current of the power battery, we cannot accurately analyze the variable safety status of the power battery.

[0077] This solution is an improvement on the analysis method of the safety status of power batteries based on the above-mentioned problems. It constructs the safety status equation of power battery operation based on the linear relationship between voltage and current in the equivalent circuit model of power battery, and combines GB-32960 data to solve the hyperparameters of the equation by fitting historical vehicle operation data with the least squares method. The obtained least squares fitting error is used as a safety element to quantify the safety characteristics of the consistency state between different cells and output high-risk point image information. Finally, combined with the knowledge of power battery mechanism, it completes the identification of power battery safety status and the source analysis of vehicle risks. It can accurately identify the fault type of power battery and the cause of the fault, thus accurately and objectively characterizing the safety status of power battery, thereby accurately identifying and analyzing the safety status of power battery, improving the operational safety of power battery, and ensuring the operational safety of new energy vehicles.

[0078] The above descriptions are merely embodiments of the present invention. Commonly known structures and characteristics 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. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A power battery safety status identification system based on equivalent circuit least squares error, characterized in that: It includes a processor module, and a data acquisition module, a data processing module, and a status recognition module, which are respectively connected to the processor module; The data acquisition module is used to collect the operating data of the power battery, form a data set, and then send it to the data processing module; The data processing module is used to preprocess the data set to obtain a selected data set; The processor module includes a storage unit and an analysis unit. The storage unit is used to store the processed selected data set. The analysis unit is used to extract security elements from the selected data set and quantify the security features to obtain security quantification features. When extracting safety elements, the equivalent circuit of the power battery is transformed to obtain the least squares fitting error. The least squares fitting error is then nonlinearly mapped and amplified to obtain the safety element vector. The least squares fitting error is: ,in, , V is the open-circuit voltage; I is the average open-circuit voltage; I is the current. The average current; the safety element vector is ,in, The loss amplification factor is a constant; The status recognition module is used to identify the safety status of the power battery based on safety quantification characteristics.

2. The power battery safety status identification system based on equivalent circuit least squares error according to claim 1, characterized in that: The state recognition module is also used to obtain safety element images based on safety quantification features, and to trace the source of safety risks of the power battery in combination with the safety element images.

3. The power battery safety status identification system based on equivalent circuit least squares error according to claim 1, characterized in that: The preprocessing of the data set includes limiting data signal boundary values, identifying and marking interference pulses, identifying and marking time discontinuities, and mean filtering.

4. The power battery safety status identification system based on equivalent circuit least squares error according to claim 2, characterized in that: The process of quantifying security features involves using variance entropy to quantify the obtained security element vector, and then substituting the security element vector into the variance entropy calculation formula to obtain the security quantified features.

5. The power battery safety status identification system based on equivalent circuit least squares error according to claim 4, characterized in that: The formula for calculating the variance entropy is as follows: ,in, .

6. The power battery safety status identification system based on equivalent circuit least squares error according to claim 5, characterized in that: The security status identification includes the following steps: make The risk is quantified by defining risk characteristics, and the risk curve is obtained by discrete integration of these characteristics over a time scale. ; The slope value of the risk curve is used to identify the safety status, and the amplitude of the identified slope value is taken as the absolute risk probability corresponding to the consistent safety feature.

7. The power battery safety status identification system based on equivalent circuit least squares error according to claim 6, characterized in that: The method of tracing the source of safety risks of power batteries involves defining the moment when the absolute risk probability exceeds the probability threshold as a high-risk point, drawing safety element images within a local time before and after the high-risk point, and tracing the source of safety risks of power batteries based on the local characteristics of changes in safety elements in the images.

Images