Battery pack self-discharge diagnosis method and storage medium

Abstract

The invention discloses a battery pack self-discharge diagnosis method and a storage medium. According to the battery pack self-discharge diagnosis method, a confidence coefficient mechanism is introduced, the influence of diagnosis scene confidence coefficient, battery pack service life confidence coefficient, battery pack system intrinsic discreteness and the like on self-discharge diagnosis confidence coefficient is considered, so that a comprehensive weighted risk coefficient is obtained, and the self-discharge diagnosis accuracy is improved by comparing the comprehensive weighted risk coefficient with a risk threshold value. And determining whether the battery pack generates self-discharge or not. According to the invention, the self-discharge of the battery pack is accurately and reliably monitored, and meanwhile, the severity of the self-discharge can be quantitatively analyzed.

Description

Technical Field

[0001] This invention relates to the field of battery technology, specifically to a method for diagnosing self-discharge of a battery pack and a storage medium. Background Technology

[0002] As the electrification of automobiles continues to deepen, safety issues related to electric vehicles are becoming increasingly prominent. Therefore, the diagnosis of battery faults is of paramount importance.

[0003] Defects in the manufacturing process can easily lead to micro-short circuits inside lithium-ion batteries, causing premature battery failure and safety hazards. Online self-discharge diagnosis of electric vehicle battery packs is a core battery management technology. Under real-world vehicle conditions, due to issues such as voltage plateaus, battery aging, and varying operating scenarios, self-discharge algorithms based on short-term voltage consistency have low accuracy, are prone to false alarms and missed alarms, and may lead to serious battery safety accidents. Summary of the Invention

[0004] This application provides a method for diagnosing battery pack self-discharge, in order to solve or at least improve the problems in the above-mentioned background art, aiming to achieve accurate and reliable monitoring of battery pack self-discharge, while also being able to quantitatively analyze the severity of self-discharge and diagnose and warn of self-discharge in its early stages.

[0005] This application provides a method for diagnosing battery pack self-discharge, comprising the following steps: (a) establishing a first rule and a second rule for judging battery pack consistency; (b) monitoring the actual voltage of the cells in the battery pack, and determining warning information based on the actual voltage, the first rule, and the second rule, wherein the warning information includes good battery pack consistency and poor battery pack consistency; and recording the operating information of the battery pack corresponding to each warning information, wherein the operating information includes at least one of temperature, SOC, intrinsic characteristics of the cell system, battery pack calendar life, battery pack cumulative discharge capacity, and battery pack health status; (c) determining a confidence level based on the operating information; (d) determining a battery pack risk coefficient corresponding to each warning information based on the warning information and the confidence level; (e) selecting a predetermined time window, wherein the time window contains multiple battery pack risk coefficients, and determining a comprehensive weighted risk coefficient of the battery pack within the predetermined time window based on the multiple battery pack risk coefficients; (f) setting a risk threshold, and determining whether the battery pack has self-discharged based on the comprehensive weighted risk coefficient and the risk threshold.

[0006] This application also provides a computer-readable storage medium that stores computer instructions, which, when executed, perform the above-described battery pack self-discharge diagnosis method.

[0007] By introducing a confidence mechanism, the present invention takes into account the impacts of diagnostic scenario confidence, battery pack life confidence, intrinsic discreteness of the battery pack system, etc. on the confidence of self-discharge diagnosis. That is, it considers the impacts of temperature, SOC, intrinsic characteristics of the battery cell system, calendar life of the battery pack, cumulative discharge capacity of the battery pack, and health state of the battery pack on the confidence of self-discharge diagnosis, thereby being able to improve the accuracy of judging the self-discharge of the battery pack. At the same time, it can quantitatively analyze the severity of self-discharge and diagnose and give early warnings in the early stage of self-discharge. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 is a schematic flowchart of the battery pack self-discharge diagnosis method according to an embodiment of the present application.

[0009] Figure 2 is a schematic diagram of a typical SOC-OCV curve according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0010] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0011] The present application provides a battery pack self-discharge diagnosis method. Referring also to Figure 1 , it includes: Step (a), establishing the first rule and the second rule for judging the consistency of the battery pack.

[0012] In a specific embodiment, a cloud battery pack consistency feature table is established. First, basic algorithms for good and poor battery pack consistency are deployed in the cloud, and battery cell data is uploaded to the cloud consistency feature table.

[0013] The first rule is: If (Vmean - Vmin) > Th1 and (Vmax - Vmin) < N * (Vmean - Vmin), then it is determined that the battery pack has poor consistency, where the value of N is [1.1 - 1.3], for example, N is 1.1, 1.2 or 1.3, etc.; the second rule is: If (Vmax - Vmin) < Th2, then it is determined that the battery pack has good consistency.

[0014] In the above first rule and second rule, Vmax is the maximum voltage of the battery cells in the battery pack, Vmin is the minimum voltage of the battery cells in the battery pack, Vmean is the average voltage of the battery cells in the battery pack; Th1 is the first threshold, and Th2 is the second threshold.

[0015] In this embodiment, the first threshold and the second threshold are fixed values. For example, the first threshold is derived from a calibration test at room temperature (e.g., 25℃±3℃), and is the minimum pressure difference corresponding to a 3% SOC difference on the SOC-OCV curve. The second threshold is also derived from a calibration test at room temperature (e.g., 25℃±3℃), and is the maximum pressure difference corresponding to a 1% SOC difference on the SOC-OCV curve. Of course, in other embodiments, the first and second thresholds may be derived from empirical values. For example, the first threshold is 20mV, and the second threshold is 10mV.

[0016] For example, the fields recorded in the cloud-based battery pack consistency characteristic table include: time, Vmax, Vmin, SOC, temperature, Vmax-Vmin, and Vmean-Vmin. Additionally, current, Vmax-Vmean, etc., may optionally be included.

[0017] Step (b) involves real-time monitoring of the actual voltage of the cells within the battery pack, and determining warning information based on the actual voltage, the first rule, and the second rule. The warning information includes good battery pack consistency and poor battery pack consistency. The operating information of the battery pack corresponding to each warning message is recorded. This operating information includes at least one of the following: temperature, SOC (Status of Charge), intrinsic characteristics of the cell system, battery pack calendar life, battery pack cumulative discharge capacity, and battery pack health status.

[0018] In one specific implementation, for a particular battery pack, the voltage of each cell within the battery pack is measured at predetermined time intervals, such as 5 seconds. The maximum voltage Vmax, minimum voltage Vmin, and average voltage Vmean of the cells within the battery pack are obtained and recorded. Vmax-Vmin and Vmean-Vmin are then obtained and recorded, thereby determining whether the battery pack has good or poor consistency based on the first and second rules in step (a). Simultaneously, the battery pack's operating information is recorded, including temperature, SOC, intrinsic characteristics of the cell system (e.g., whether it is lithium iron phosphate or ternary lithium), battery pack calendar life, cumulative discharge capacity, and battery pack health status.

[0019] It is understandable that the battery pack's operational information includes diagnostic scenario information, system information, and lifespan information. For ease of description, the diagnostic scenario information is defined as temperature and SOC; the system information is defined as the intrinsic characteristics of the cell system; and the lifespan information is defined as the battery pack's calendar life, cumulative discharge capacity, and health status.

[0020] For example, in order to record the above information, a basic characteristic table of the battery pack also needs to be established, including fields such as: battery pack manufacturing date (i.e., battery pack calendar life), battery pack cumulative discharge capacity, and battery pack health status.

[0021] In related technologies, algorithms based on short-term voltage inconsistency are prone to false self-discharge alarms. Therefore, in this embodiment, to reduce the false alarm rate, a confidence level mechanism is introduced, considering the impact of diagnostic scenario confidence level, battery pack lifespan confidence level, and intrinsic dispersion of the battery pack system on the self-discharge diagnosis confidence level. Specifically, it considers the influence of temperature, SOC, intrinsic characteristics of the cell system, battery pack calendar lifespan, battery pack cumulative discharge capacity, and battery pack health status on the self-discharge diagnosis confidence level. All six factors mentioned above affect the differential voltage monitoring of the cells; therefore, introducing the confidence levels of each factor improves the accuracy of battery pack self-discharge judgment.

[0022] It is understood that in other embodiments, only one or more of the above six influencing factors may be introduced, which can also improve the defect of high false alarm rate of battery pack self-discharge, and does not depart from the essence of this application.

[0023] Step (c) determines the confidence level based on the battery pack's operational information. In other words, the confidence level is determined based on the recorded battery pack temperature, SOC, intrinsic characteristics of the cell system, battery pack calendar life, battery pack cumulative discharge capacity, and battery pack health status.

[0024] Specifically, it includes: step (c1), determining the slope weighted correction coefficient Wk of the cell SOC-OCV curve, where Wk is a first function of SOC.

[0025] For example, the first function is:

[0026] Wk=f1(soc_k)=m2-(soc_k-soc_kmax)*(m2-m1) / (soc_kmin-soc_kmax),

[0027] Where soc_k is the slope of the SOC-OCV curve, and f1(soc_k) is a linear mapping function that maps the maximum slope soc_kmax to m1 and the minimum slope soc_kmin to m2. The value of m1 is [0.1, 0.4] and the value of m2 is [0.6, 0.9]. For example, m1 is 0.1 and m2 is 0.9.

[0028] Step (c2) determines the discrete weighted correction coefficient Wd of the cell SOC-OCV curve, where Wd is a second function of SOC.

[0029] For example, the second function is:

[0030] Wd=f2(soc_dV)=m4-(soc_dV-soc_dVmax)*(m4-m3) / (soc_dVmin-soc_dVmax),

[0031] Among them, the SOC-OCV curves of M battery cells are measured. The voltage difference soc_dV corresponding to each SOC is used as the dispersion. f2(soc_dV) is a linear mapping function that maps the maximum dispersion soc_dVmax to m3 and the minimum dispersion soc_dVmin to m4. The value of m3 is [0.1, 0.4] and the value of m4 is [0.6, 0.9]. For example, m3 is 0.1 and m4 is 0.9. M is an integer greater than 5, for example, M is 10.

[0032] Step (c3) determines the temperature-dependent weighted correction coefficient WT of the cell's SOC-OCV curve. WT is a third function of temperature T and SOC.

[0033] For example, the third function is:

[0034] WT=f3(soc_dVdT)=m6-(soc_dVdT-soc_dVdTmax)*(m6-m5) / (soc_dVdTmin-soc_dVdTmax),

[0035] Among them, the SOC-OCV curves of the battery cell at different temperatures are measured, and the soc_dVdT corresponding to each SOC is calculated as the dispersion. f3(soc_dVdT) is a linear mapping function, which maps the maximum dispersion soc_dVdTmax to m5 and the minimum dispersion soc_dVdTmin to m6. The value of m5 is [0.8, 1] and the value of m6 is [1, 1.2]. For example, m5 is 0.9 and m6 is 1.1.

[0036] Step (c4) determines the battery pack calendar life weighted correction factor Wcal, which is a fourth function of the battery pack calendar life.

[0037] For example, the fourth function is:

[0038]

[0039] Where lifecal is the battery pack calendar lifespan, and the value of n1 is [0.3, 0.5], for example, n1 is 0.3.

[0040] Step (c5) determines the weighted correction coefficient WQ for the cumulative discharge capacity of the battery pack. WQ is the fifth function of the cumulative discharge capacity of the battery pack.

[0041] For example, the fifth function is:

[0042]

[0043] Where Q is the cumulative discharge capacity of the battery pack, Qmax is the maximum discharge capacity of the battery pack, and is the number of warranty cycles * nominal capacity; where the value of n2 is [0.3, 0.5], for example, n2 is 0.3.

[0044] Step (c6) determines the weighted correction coefficient WSOH for battery pack health status, which is the sixth function relating to battery pack health status.

[0045] For example, the sixth function is:

[0046]

[0047] Wherein, SOH represents the battery pack health status, and the value of n3 is [0.3, 0.4], for example, n3 is 0.3.

[0048] It is understood that the above functional relationship is an exemplary representation, and this application does not specifically limit the form of the function. Other mapping functions that can satisfy the effect of this embodiment do not depart from the essence of the present invention.

[0049] Step (c7) determines the confidence level, which is the product of the above six coefficients Wk, Wd, WT, Wcal, WQ, and WSOH.

[0050] Of course, it is understood that in other embodiments, based on the recorded battery pack operating information, the confidence level may also be one or the product of several of the six coefficients Wk, Wd, WT, Wcal, WQ, and WSOH, without departing from the essence of this application. For example, only the impact of battery pack lifespan confidence level on self-discharge diagnosis confidence level is considered, i.e., only the impact of battery pack calendar lifespan, battery pack cumulative discharge capacity, and battery pack health status is considered to obtain the coefficients Wcal, WQ, and WSOH. Therefore, in this embodiment, the confidence level is the product of Wcal, WQ, and WSOH.

[0051] Step (d): Based on the warning information and the confidence level, determine the battery pack risk coefficient corresponding to each warning information.

[0052] It is understandable that if the warning information of the battery pack is consistent, it will reduce the overall risk coefficient of the battery pack; conversely, if the warning information of the battery pack is inconsistent, it will increase the overall risk coefficient of the battery pack.

[0053] Therefore, in step (d), the battery pack risk coefficient is the product of the consistency warning information coefficient and the confidence level; wherein, the determination of the consistency warning information coefficient includes: when the warning information in step (b) is good cell consistency, the warning information coefficient is -1; when the warning information in step (b) is poor cell consistency, the warning information coefficient is 1.

[0054] Step (e) involves selecting a predetermined time window, which includes multiple battery pack risk coefficients. Based on these multiple battery pack risk coefficients, a comprehensive weighted risk coefficient for each battery pack within the predetermined time window is determined. In step (e), the comprehensive weighted risk coefficient is the sum of the risk coefficients for each battery pack.

[0055] In other words, Risk = ∑A*Wk*Wd*WT*Wcal*WQ*WSOH, where A is the consistency warning information coefficient, with a value of 1 or -1, and Risk is the comprehensive weighted risk coefficient.

[0056] For example, the scheduled time window can be selected as 3 months, 6 months, 12 months, etc.

[0057] (f) Set a risk threshold and determine whether the battery pack will self-discharge based on the comprehensive weighted risk coefficient and the risk threshold.

[0058] For example, when the comprehensive weighted risk coefficient is greater than or equal to the risk threshold, it is determined that the battery pack has self-discharged; otherwise, it is determined that the battery pack has not self-discharged.

[0059] Therefore, by comparing the comprehensive weighted risk coefficient with the risk threshold, it was finally determined whether the battery pack had self-discharged.

[0060] Furthermore, it can be understood that the greater the weighted risk coefficient exceeds the risk threshold, the more severe the self-discharge of the battery pack.

[0061] Risk_thresh represents the risk threshold, which is set from a calibration value. In one embodiment, Risk_thresh is 0.8.

[0062] In this embodiment, to further determine the specific cells within the battery pack that experienced self-discharge, step (b) further includes: recording the percentage of each cell within the battery pack that is the lowest voltage cell in the battery pack; after step (f), the method further includes: after determining that the battery pack has experienced self-discharge, determining the cells that experienced self-discharge based on the percentage. In other words, when a certain cell is the lowest voltage cell the most times, its percentage is the highest, and therefore it is determined that the cell has experienced self-discharge. This allows for accurate location of the faulty cells, facilitating subsequent maintenance.

[0063] When a self-discharge fault occurs in the battery pack, an alarm or other means can be used to alert the user and output self-discharge information.

[0064] The table below shows data from a specific embodiment, which includes warning information on good and poor battery pack consistency obtained from monitoring a lithium iron phosphate battery pack over a historical 6-month period. The risk threshold is set at 0.8.

[0065]

[0066] The table shows six warning messages related to battery pack consistency detected over the past six months. For each warning message, the corresponding SOC, temperature, battery pack calendar life, cumulative discharge capacity, and state of health (SOH) were recorded. Based on the mapping relationship, the corresponding coefficients Wk, Wd, WT, Wcal, WQ, and WSOH were obtained. These coefficients were then multiplied to obtain the confidence level. Simultaneously, depending on whether the warning message indicated good or poor consistency, the corresponding consistency warning message coefficient was -1 or 1. The consistency warning message coefficient was then multiplied by the confidence level to obtain the risk coefficient for each warning message. Finally, the six risk coefficients were summed to obtain the comprehensive weighted risk coefficient.

[0067] As can be seen, the overall weighted risk coefficient is greater than the risk threshold, therefore it is determined that the battery pack has self-discharged.

[0068] Furthermore, based on the recorded voltage of each cell, if cell number 5, being the lowest voltage cell in the battery pack, has the highest percentage, then it is determined that cell number 5 has experienced self-discharge.

[0069] The battery pack self-discharge diagnostic method of this application can be deployed not only in the cloud, but also in the vehicle-side battery management system.

[0070] Finally, this application also provides a computer-readable storage medium that stores computer instructions, which, when executed, perform the above-described battery pack self-discharge diagnosis method.

[0071] This application addresses the diagnosis of battery pack self-discharge by introducing a confidence level mechanism. It considers the impact of factors such as the confidence level of the diagnostic scenario, the confidence level of the battery pack lifespan, and the intrinsic dispersion of the battery pack system on the confidence level of self-discharge diagnosis. This achieves accurate and reliable monitoring of battery pack self-discharge, allows for quantitative analysis of the severity of self-discharge, and enables diagnosis and early warning in the early stages of self-discharge. Furthermore, by recording the percentage of cells with the lowest voltage, the self-discharging cells can be accurately located, achieving precise diagnosis.

[0072] It should be noted that although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope defined in the claims of the present invention.

Claims

1. A method for diagnosing self-discharge in a battery pack, characterized in that, It includes the following steps: (a) Establish the first rule and the second rule for judging the consistency of the battery pack; (b) Monitor the actual voltage of the battery cells in the battery pack, and determine the warning information based on the actual voltage, the first rule and the second rule. The warning information includes that the battery pack has good consistency and the battery pack has poor consistency; And record the working information of the battery pack corresponding to each piece of the warning information. The working information includes at least one of temperature, SOC, intrinsic characteristics of the battery cell system, calendar life of the battery pack, cumulative discharge capacity of the battery pack, and health state of the battery pack; (c) Determine the confidence level based on the working information; (d) Determine the risk coefficient of the battery pack corresponding to each piece of the warning information based on the warning information and the confidence level; (e) Select a predetermined time window. There are multiple battery pack risk coefficients within the time window. Based on the multiple battery pack risk coefficients, determine the comprehensive weighted risk coefficient of the battery pack within the predetermined time window; (f) Set a risk threshold. Based on the comprehensive weighted risk coefficient and the risk threshold, determine whether the battery pack has self-discharged.

2. The battery pack self-discharge diagnosis method according to claim 1, characterized in that, In step (a), the first rule is: If (Vmean - Vmin) > Th1 and (Vmax - Vmin) < N * (Vmean - Vmin), then it is determined that the battery pack has poor consistency, where the value of N is [1.1, 1.3]; The second rule is: If (Vmax - Vmin) < Th2, then it is determined that the battery pack has good consistency; Where, Vmax is the maximum voltage of the battery cells in the battery pack, Vmin is the minimum voltage of the battery cells in the battery pack, Vmean is the average voltage of the battery cells in the battery pack, Th1 is the first threshold, and Th2 is the second threshold.

3. The battery pack self-discharge diagnosis method according to claim 2, characterized in that, The first threshold comes from a calibration test at room temperature, and the first threshold is the minimum voltage difference corresponding to a 3% SOC difference on the SOC-OCV curve; The second threshold comes from a calibration test at room temperature, and the second threshold is the maximum voltage difference corresponding to a 1% SOC difference on the SOC-OCV curve.

4. The battery pack self-discharge diagnosis method according to any one of claims 1-3, characterized in that, In step (b), the real-time monitoring of the actual voltage of the battery cells in the battery pack includes: measuring the voltage of each battery cell in the battery pack at a predetermined time interval, and obtaining and recording the maximum voltage, minimum voltage and average voltage of the battery cells in the battery pack.

5. The battery pack self-discharge diagnosis method according to claim 4, characterized in that, In step (c), the confidence level is the product of one or more of the coefficients Wk, Wd, WT, Wcal, WQ, and WSOH; Wherein, Wk is the slope-weighted correction coefficient of the cell SOC-OCV curve, which is the first function of SOC; Wd is the discrete weighted correction coefficient of the cell SOC-OCV curve, which is the second function of SOC; WT is the temperature-dependent weighted correction coefficient of the cell SOC-OCV curve, which is the third function of temperature and SOC; Wcal is the battery pack calendar life weighted correction coefficient, which is the fourth function of battery pack calendar life; WQ is the battery pack cumulative discharge capacity weighted correction coefficient, which is the fifth function of battery pack cumulative discharge capacity; and WSOH is the battery pack health status weighted correction coefficient, which is the sixth function of battery pack health status.

6. The battery pack self-discharge diagnosis method according to claim 5, characterized in that, The first function is: W k =f1(soc_k)=m2-(soc_k-soc_kmax)*(m2-m1) / (soc_kmin-soc_kmax), Where soc_k is the slope of the SOC-OCV curve, and f1(soc_k) is a linear mapping function that maps the maximum slope soc_kmax to m1 and the minimum slope soc_kmin to m2, where the value of m1 is [0.1, 0.4] and the value of m2 is [0.6, 0.9]. And / or, the second function is: Wd=f2(soc_dV)=m4-(soc_dV-soc_dVmax)*(m4-m3) / (soc_dVmin-soc_dVmax), Among them, the SOC-OCV curve of M battery cells is measured. The voltage difference soc_dV corresponding to each SOC is used as the dispersion. f2(soc_dV) is a linear mapping function. The maximum dispersion soc_dVmax is mapped to m3, and the minimum dispersion soc_dVmin is mapped to m4. The value of m3 is [0.1, 0.4], the value of m4 is [0.6, 0.9], and M is an integer greater than 5. And / or, the third function is: WT=f3(soc_dVdT)=m6-(soc_dVdT-soc_dVdTmax)*(m6-m5) / (soc_dVdTmin-soc_dVdTmax), Among them, the SOC-OCV curves of the battery cell at different temperatures are measured, and the soc_dVdT corresponding to each SOC is calculated as the dispersion. f3(soc_dVdT) is a linear mapping function, which maps the maximum dispersion soc_dVdTmax to m5 and the minimum dispersion soc_dVdTmin to m6, where the value of m5 is [0.8, 1] and the value of m6 is [1, 1.2]. And / or, the fourth function is: Where lifecal is the battery pack calendar lifespan, and n1 has a value of [0.3, 0.5]. And / or, the fifth function is: Where Q is the cumulative discharge capacity of the battery pack, Qmax is the maximum discharge capacity of the battery pack, and n is the number of warranty cycles multiplied by the nominal capacity; the value of n2 is [0.3, 0.5]. And / or, the sixth function is: Where SOH represents the battery pack health status, and the value of n3 is [0.3, 0.4].

7. The battery pack self-discharge diagnosis method according to claim 1, characterized in that, In step (d), the battery pack risk coefficient is the product of the consistency warning information coefficient and the confidence level; The determination of the consistency warning information coefficient includes: when the warning information in step (b) is good cell consistency, the warning information coefficient is -1; when the warning information in step (b) is poor cell consistency, the warning information coefficient is 1.

8. The battery pack self-discharge diagnosis method according to claim 1, characterized in that, In step (e), the comprehensive weighted risk coefficient is the sum of the risk coefficients of the multiple battery packs.

9. The battery pack self-discharge diagnosis method according to claim 1, characterized in that, In step (f), determining whether the battery pack has self-discharged includes: if the comprehensive weighted risk coefficient is greater than or equal to the risk threshold, it is determined that the battery pack has self-discharged; otherwise, it is determined that the battery pack has not self-discharged.

10. The battery pack self-discharge diagnosis method according to claim 9, characterized in that, Step (b) further includes: recording the percentage of each cell in the battery pack as the lowest voltage cell in the battery pack; The process further includes, after step (f): when it is determined that the battery pack has self-discharged, determining the cells that have self-discharged based on the percentage.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, which, when executed, perform the battery pack self-discharge diagnosis method as described in any one of claims 1-10.

Images