A method for predicting cell failure in a lithium battery system and a lithium battery system

Abstract

This invention provides a cell failure prediction method and a lithium battery system, belonging to the field of lithium battery failure prediction technology. The lithium battery system of this invention includes a lithium battery device and a lithium battery monitoring device. The lithium battery monitoring device includes a primary BMS, a secondary BMS, and a prediction device. The primary BMS is used to achieve full charge cutoff to prevent overcharging. The secondary BMS is used to record the full charge data of multiple lithium batteries. The prediction device is used to calculate the remaining charging capacity and predict whether a micro-short circuit failure has occurred in the lithium battery cell based on the remaining charging capacity, and to identify the cell in the lithium battery that has experienced a micro-short circuit failure. This invention utilizes a local BMS for full charge cutoff, a secondary BMS to directly record full charge data, and a powerful prediction device to calculate the remaining charging capacity of each lithium battery cell to predict whether cell failure has occurred. This satisfies both data timeliness requirements, avoiding calculation errors caused by data delays, and the computational requirements of the powerful prediction device.

Description

Technical Field

[0001] This invention relates to the field of lithium battery failure technology, and in particular to a cell failure prediction method and a lithium battery system. Background Technology

[0002] Lithium-ion batteries typically consist of a group of multiple cells. During the manufacturing process, contamination from impurities, or overcharging or over-discharging during use that triggers metal dendrites and causes separator rupture, can easily lead to micro-short circuits. These micro-short circuits often have a long evolutionary cycle. In the mid-to-late stages, the lithium-ion battery will generate a large amount of heat in a short period of time, and there are no effective countermeasures to prevent thermal runaway at this stage. Therefore, early diagnosis of micro-short circuit faults in lithium-ion batteries is crucial to preventing further damage caused by thermal runaway.

[0003] However, in the initial stage of a micro-short circuit, the failure phenomenon is not obvious. At this time, the short-circuit resistance is relatively large, so the short-circuit current is relatively small, and the resulting thermal effect is not significant, making it difficult to determine based on the external temperature of the lithium battery. In addition, the impact on the battery terminal voltage is also small in the initial stage of a micro-short circuit, making it difficult to determine based on the battery terminal voltage. Therefore, existing lithium battery failure prediction methods based on battery temperature or terminal voltage face significant challenges in predicting micro-short circuit failure. Summary of the Invention

[0004] This invention provides a cell failure prediction method and a lithium battery system to solve the problem of high difficulty and inaccurate prediction of micro-short circuit failure in lithium batteries.

[0005] In a first aspect, the present invention provides a cell failure prediction method for a lithium battery system. The lithium battery system includes a lithium battery device having multiple lithium batteries and a lithium battery monitoring device. The lithium batteries include multiple cells connected in series. The lithium battery monitoring device is connected to each of the lithium batteries and is used to predict whether a cell micro-short circuit failure occurs in each lithium battery. The lithium battery monitoring device includes: a primary BMS disposed within the lithium battery, a secondary BMS corresponding to the multiple lithium batteries and disposed within the lithium battery device and directly communicatively connected to the primary BMS, and a prediction device communicating with each secondary BMS. The method includes: when the primary BMS detects that the charging voltage of any cell exceeds the charging cutoff voltage, it determines that the current charging process is a full charge and controls all cells of the corresponding lithium battery to stop charging; the secondary BMS... MS records the full-charge data of each lithium battery during each full charge. This full-charge data includes the charging voltage and current of each cell at several moments during the full charge process, as well as the full-charge voltage at the moment of full charge. The prediction device acquires the full-charge data of each lithium battery to calculate the remaining charging capacity of each cell during this full charge. When the number of full charges during this full charge is greater than 1, it calculates the difference between the remaining charging capacity of each cell during this full charge and the remaining charging capacity during the previous full charge for each cell of the lithium battery. The prediction device also identifies abnormal differences and corresponding cells among the differences in the remaining charging capacity of each cell, and based on the abnormal differences and corresponding cells, determines that a micro-short circuit failure has occurred in the lithium battery and identifies the cell in the lithium battery that has experienced the micro-short circuit failure.

[0006] Secondly, the present invention provides a lithium battery system, comprising: a lithium battery device having multiple lithium batteries and a lithium battery monitoring device; the lithium batteries include multiple cells connected in series; the lithium battery monitoring device is connected to each of the lithium batteries and includes: a primary BMS disposed within the lithium batteries, a secondary BMS corresponding to the multiple lithium batteries and disposed within the lithium battery device for direct communication connection to the primary BMS, and a prediction device communicating with each secondary BMS; the lithium battery monitoring device uses the method described in the foregoing technical solution to predict whether each lithium battery has experienced a cell micro-short circuit failure.

[0007] The lithium battery system of this invention includes a lithium battery device and a lithium battery monitoring device. The lithium battery monitoring device includes a primary BMS, a secondary BMS, and a prediction device. The primary BMS is used to achieve full charge cutoff to prevent overcharging. The secondary BMS is used to record full charge data of multiple lithium batteries. The prediction device is used to calculate the remaining charging capacity and predict whether a micro-short circuit failure has occurred in the lithium battery based on the remaining charging capacity, and to identify the cell in the lithium battery that has experienced a micro-short circuit failure. Specifically, the prediction device calculates the difference in remaining charging capacity between two consecutive effective full charges for the same cell in the lithium battery, and identifies abnormal differences among the differences of each cell. Since the difference in remaining charging capacity of a micro-short-circuited cell is greater than that of a normal cell, the micro-short-circuited cell can be identified based on the abnormal difference. The remaining charging capacity of a micro-short-circuited cell increases continuously during the charging and discharging process. This invention identifies micro-short-circuited cells based on the difference in remaining charging capacity between two consecutive effective full charges, which relatively reduces the computational load and lowers the implementation difficulty.

[0008] In this process, the acquisition of full-charge data needs to be highly timely, especially the full-charge voltage of each cell at the moment of full charge. If deviations occur due to data delays, they will cause large errors in the calculation of the remaining charging capacity, thus affecting the prediction of micro-short circuit failure. However, the calculation of the remaining charging capacity and the prediction of micro-short circuit failure are relatively complex. Therefore, this invention uses the local primary BMS of the lithium battery device to cut off full charge, uses a secondary BMS that communicates directly with the primary BMS to directly record the full-charge data, and uses a powerful prediction device to calculate the remaining charging capacity of each cell of the lithium battery to predict whether cell failure has occurred. This satisfies the data timeliness requirement, avoids calculation and prediction errors caused by data delays, and utilizes a powerful prediction device to meet the corresponding calculation requirements. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is an application scenario diagram of the cell failure prediction method for lithium battery systems provided in the embodiments of the present invention;

[0011] Figure 2 This is a schematic diagram illustrating the capacity difference of series-connected battery cells provided in an embodiment of the present invention;

[0012] Figure 3 This is a schematic diagram of the RCC change of a micro-short-circuited cell during the charging and discharging process provided in an embodiment of the present invention;

[0013] Figure 4 This is a schematic diagram of the cell voltage curve transformation with different parameters provided in the embodiments of the present invention;

[0014] Figure 5 This is a schematic diagram of the voltage curves of each cell in the lithium battery provided in the embodiments of the present invention;

[0015] Figure 6 This is a schematic diagram of a box-shaped structure provided in an embodiment of the present invention. Detailed Implementation

[0016] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0017] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0018] Figure 1 This diagram illustrates an application scenario of the lithium battery cell failure prediction method provided in this invention. This invention can be applied to various application scenarios involving the prediction of lithium battery failure; specific application scenarios are not limited here. For example, in one application scenario, such as… Figure 1 As shown, a lithium battery consists of multiple cells connected in series.

[0019] Lithium-ion battery failures are mainly classified into two categories: performance failures and safety failures. Performance failures refer to the lithium-ion battery's performance failing to meet usage requirements and relevant indicators, primarily including capacity decay, short cycle life, and performance degradation at high and low temperatures. Safety failures refer to lithium-ion batteries exhibiting certain safety risks due to improper use or abuse, primarily including thermal runaway, short circuits, gas expansion, leakage, lithium plating, and expansion deformation.

[0020] Lithium battery failure analysis aims to predict and prevent its occurrence, enabling battery monitoring devices to take effective measures before serious consequences occur, reduce losses, and provide data support for subsequent accident handling. This invention only analyzes the prediction of lithium battery failure; failure prevention is not addressed.

[0021] The occurrence of lithium battery failure is accompanied by failure phenomena, and a particular cause of failure may manifest differently over time. Generally, lithium battery failure phenomena may escalate over time, and failure to address them promptly can lead to incalculable losses. Therefore, predicting lithium battery failure should begin with early failure phenomena, analyzing abnormal battery data, and ultimately aiming to predict failure as early as possible. The following section explains the micro-short-circuit failure phenomenon of lithium batteries and the principles of failure prediction.

[0022] Impurities during lithium battery production or overcharging / over-discharging during use can trigger metal dendrite formation, causing separator rupture and potentially triggering micro-short circuits. These micro-short circuits often have a long evolutionary cycle; in the mid-to-late stages, the lithium battery generates a large amount of heat in a short period, and there are no effective countermeasures to prevent thermal runaway at this stage. Therefore, early diagnosis of lithium battery micro-short circuit faults is crucial to preventing further damage caused by thermal runaway.

[0023] However, in the initial stage of a micro-short circuit, the failure phenomenon is not obvious. At this time, the short-circuit resistance is relatively large, therefore the short-circuit current is relatively small, and the resulting thermal effect is not significant, making it difficult to determine based on the external temperature of the lithium battery. Furthermore, the impact on the battery terminal voltage is also small in the initial stage of a micro-short circuit, making it difficult to determine based on the battery terminal voltage alone. Therefore, existing lithium battery cell failure prediction methods based on battery temperature or terminal voltage have significant difficulty predicting micro-short circuit failures. Therefore, it is necessary to use models and algorithms to reflect the abnormal electrothermal characteristics of the faulty battery, thereby diagnosing battery micro-short circuit faults. This invention mainly predicts micro-short circuit failures based on the remaining charging capacity of the lithium battery.

[0024] The following section first explains the remaining charging capacity of lithium batteries. Due to inconsistencies in manufacturing processes and uneven operating environments, inconsistencies in the cells within a lithium battery pack are unavoidable.

[0025] Figure 2 This is a schematic diagram illustrating the capacity difference of series-connected battery cells according to an embodiment of the present invention. (Refer to...) Figure 2 A lithium battery pack consists of eight cells connected in series. The black squares represent the initial capacity, and the diagonal lines represent the charging capacity. Differences in the production and use of lithium battery packs result in the total capacity of each cell differing from the initial capacity. Because the cells are connected in series, the charge and discharge amounts of each cell are the same during charging and discharging (i.e., the diagonal lines on each cell are equal). When one cell is fully charged, for example... Figure 2 When cell #7 is fully charged, the BMS will stop charging to prevent overcharging. At this time, the difference between the total capacity of the other cells and the current capacity (initial capacity + charging capacity) is the battery's remaining charging capacity (RCC). Figure 2The blank space between the diagonal line and the dashed line.

[0026] The following further explains the relationship between micro-short circuits and remaining charging capacity in lithium batteries. Ignoring battery aging, if no micro-short circuit occurs, the RCC value of the same cell should be the same each time the battery is fully charged. If a cell experiences a micro-short circuit, it will continuously consume the battery's stored capacity, causing the RCC value of that cell to increase continuously when fully charged. The following uses a small lithium battery pack model to illustrate this situation.

[0027] Figure 3 This is a schematic diagram illustrating the RCC change of a micro-short-circuited cell during charging and discharging, provided as an embodiment of the present invention. (Refer to...) Figure 3 The lithium battery pack consists of two cells, denoted as X# and Y#. X# and Y# have the same total capacity, but cell X# has experienced a micro-short circuit. Based on the charging and discharging principles of lithium battery packs, charging and discharging will stop as soon as one cell reaches its charging or discharging cutoff voltage. The following will use... Figure 3 Explain the various states ① to ⑤ during the charging and discharging process.

[0028] State ①: Initially, both X# and Y# are fully charged, as shown in State ①. At this time, the remaining charging capacity of cell X# is denoted as RCC0, and RCC0 = 0.

[0029] State ②: The lithium battery is being discharged. Due to the energy loss from the micro-short circuit, cell X# will reach full discharge sooner. At the end of the discharge, as in State ②, the remaining discharging capacity (RDC) of cell Y# is equal to the energy loss from the micro-short circuit during the discharge of cell X#. The remaining discharging capacity of cell Y# corresponds to... Figure 3 The black square portion in state ②.

[0030] State ③: The lithium battery pack is being charged, and cell Y# will reach full charge sooner. At the end of charging, as in State ③, the remaining charge capacity of cell X# is measured as RCC1.

[0031] State 4: Discharging the lithium battery pack. At the end of the discharge, it returns to state 4.

[0032] Status 5: Charging the lithium battery pack. When charging is complete, as in Status 5, the remaining charge capacity of cell X# is measured as RCC2.

[0033] During the above charge and discharge process, it is obvious that RCC0 < RCC1 < RCC2. This is because the micro-short circuit will continuously consume the power of the X# cell. Therefore, every time the lithium battery pack is fully charged, the remaining charging capacity of the X# cell continuously increases. At the same time, every time the lithium battery pack is fully discharged, the remaining discharge power of the Y# cell continuously increases. This also shows that if the micro-short circuit cell is not processed in time, the capacity difference between the normal cells and the micro-short circuit cell of the lithium battery pack will become larger and larger, and finally it will not be able to charge and discharge normally. If there is no micro-short circuit in X#, the capacity change during the charge and discharge process is exactly the same as that of the Y# cell, and the RCC is always 0. The above process shows that every time the lithium battery pack is fully charged, the RCC value of the normal cell basically does not change, while the RCC value of the micro-short circuit cell will continue to increase. Therefore, it is possible to determine whether a cell has a micro-short circuit by the change of the RCC value of the cell.

[0034] In the embodiment of the present invention, by calculating the difference in the remaining charging capacity between two adjacent full charges based on the same cell, and determining the micro-short circuit failure cell based on the abnormal difference, the problem of the large difficulty in predicting the micro-short circuit failure of the lithium battery is solved.

[0035] The cell failure prediction method provided by the embodiment of the present invention is applied to the lithium battery system of the embodiment of the present invention. The lithium battery system includes a lithium battery device having a plurality of lithium batteries and a lithium battery monitoring device. The lithium battery includes a plurality of serially connected cells. The lithium battery monitoring device is connected to each lithium battery and predicts whether each lithium battery has a cell micro-short circuit failure by executing the following method steps.

[0036] In this embodiment, the lithium battery monitoring device includes: a primary BMS disposed in the lithium battery, a secondary BMS corresponding to a plurality of lithium batteries and disposed in the lithium battery device to directly communicate with the primary BMS, and a prediction device communicating with each secondary BMS.

[0037] The method includes:

[0038] When the primary BMS detects that the charging voltage of any cell exceeds the charging cut-off voltage, it determines that this charging process is a full charge and controls all the cells of the corresponding lithium battery to stop charging.

[0039] Generally speaking, during the charging process of the lithium battery, when any cell reaches the charging cut-off voltage, it is determined that the lithium battery is fully charged. Correspondingly, the primary BMS controls the lithium battery to stop charging. Based on the above description of the remaining charging capacity, after a certain cell in the plurality of cells of the lithium battery is fully charged, the other cells are in an undercharged state. At this time, the lithium battery includes at least one fully charged cell and a plurality of undercharged cells.

[0040] The secondary BMS records the full charge data of each lithium battery during each full charge; wherein, the full charge data includes: the charging voltage and charging current of each cell at several moments during this full charge process, as well as the full charge voltage at the moment of full charge.

[0041] In this invention, the secondary BMS can record full-charge data each time the lithium battery is fully charged, or it can record full-charge data only when other limiting conditions are met during the full-charge process. The latter is defined as a valid full charge in this invention and has high predictive accuracy. Therefore, this embodiment will be described using the example of recording full-charge data only when valid full charge is met. The reason for its high predictive accuracy will be explained in detail below. As for the embodiment that records full-charge data each time the full charge is completed, it is only necessary to remove the above-mentioned limiting conditions. Correspondingly, the valid full charge mentioned in the following embodiments can be replaced with ordinary full charge, which will not be elaborated further.

[0042] The prediction device acquires the full-charge data of each lithium battery to calculate the remaining charging capacity of each cell in the current full charge. When the number of full charges in the current full charge is greater than 1, it calculates the difference between the remaining charging capacity of each cell in the current full charge and the remaining charging capacity at the previous full charge. The prediction device also identifies abnormal differences and corresponding cells among the differences in the remaining charging capacity of each cell, and based on these abnormal differences and corresponding cells, determines that a micro-short circuit failure has occurred in the lithium battery and identifies the cell in the lithium battery that experienced the micro-short circuit failure. In this embodiment, the prediction device is a cloud server located outside the lithium battery device and communicatively connected to each of the secondary BMS.

[0043] This invention, in its embodiments, calculates the difference in remaining charging capacity between two consecutive full charges of the same cell during the charging and discharging process of a lithium battery, and identifies abnormal differences among these differences. Since the difference in remaining charging capacity of a micro-short-circuited cell is greater than that of a normal cell, a micro-short-circuited cell can be identified based on this abnormal difference. The remaining charging capacity of a micro-short-circuited cell increases continuously during the charging and discharging process. This invention, by identifying micro-short-circuited cells based on the difference in remaining charging capacity between two consecutive full charges, reduces the computational complexity and lowers the implementation difficulty.

[0044] Furthermore, the acquisition of full-charge data in the above process requires high timeliness, especially the full-charge voltage of each cell at the moment of full charge. If deviations occur due to data delays, it will produce a large error in the calculation of the remaining charging capacity, thereby affecting the prediction of micro-short-circuit failure. The reasons for this will be explained below. However, the calculation of the remaining charging capacity and the prediction of micro-short-circuit failure are relatively complex. Therefore, this invention utilizes the local primary BMS of the lithium battery device to cut off full charge, uses a secondary BMS that communicates directly with the primary BMS to directly record the full-charge data, and uses a powerful prediction device to calculate the remaining charging capacity of each cell of the lithium battery to predict whether cell failure has occurred. This satisfies the requirements for data timeliness, avoids calculation and prediction errors caused by data delays, and utilizes a powerful prediction device to meet the corresponding calculation requirements.

[0045] Specifically, the secondary BMS can execute the following step 400 to determine whether a valid full charge has been achieved. In step 400, when each lithium battery is fully charged, it is determined whether the charging constitutes a valid full charge, and when a valid full charge is achieved, the full charge data for each lithium battery for each full charge is recorded.

[0046] Furthermore, the term "effective full charge" indicates that the charging voltage curve of the fully charged cell and the standard charging voltage curve of the cell are close to each other in terms of full charge voltage and / or terminal trend. The standard charging voltage curve of the cell is the charging voltage curve after the cell has been fully discharged and then fully charged. The closeness of the full charge voltages of the two charging voltage curves means that the difference in voltage values ​​at the moment of full charge is less than a preset voltage difference threshold. Correspondingly, the closeness of the terminal trends of the two charging voltage curves means that within a preset time period before the moment of full charge, the difference in the overall voltage change rate of the two charging voltage curves is less than a preset change rate difference threshold. It can be understood that, among the conditions constituting an effective full charge, either the full charge voltage or the terminal trend can be satisfied, or both must be satisfied simultaneously.

[0047] The reason for proposing the concept of effective full charge and defining it as above in this invention is mainly because the applicant of this invention has discovered that the concept and calculation of remaining charging capacity are related to the charging process of a fully charged cell. As will be described in step 401 and its sub-steps below, the calculation of remaining charging capacity can be based on a fully charged cell. For example, the following embodiments of this invention provide two methods for calculating remaining charging capacity, both of which are based on the high correlation between the charging processes of fully charged and partially charged cells. The remaining charging time of each cell is obtained based on a fully charged cell, and then the remaining charging capacity is obtained.

[0048] During the research and development process, the applicant discovered that the method of determining the remaining charging capacity based on a fully charged cell resulted in significant fluctuations in the calculation results each time. Specifically, the applicant found that the calculated remaining charging capacity was not accurate for every full charge. Furthermore, the applicant discovered that when a lithium battery is effectively fully charged, the charging curve of a fully charged cell differs little from that of a partially charged cell. Therefore, obtaining the remaining charging capacity result based on a fully charged cell is more accurate, leading to more accurate predictions of micro-short circuit failures.

[0049] In one possible implementation, step 400 may include: determining whether the charging time is greater than a preset time after full charging is completed. If the charging time is greater than the preset time, the lithium battery is determined to be effectively fully charged.

[0050] It should be noted that, during the research and development process, the applicant of this invention discovered that when the charging time is too short, even if the battery is fully charged, the full-charge voltage at the end of the charging process may differ significantly from the full-charge voltage of the complete discharge-to-full-charge curve (i.e., the aforementioned standard charging voltage curve). This leads to a decrease in the accuracy of the remaining charging capacity and an excessively high prediction error rate. Therefore, the effective full-charge judgment condition of this invention includes a charging time longer than a preset time, in order to improve the accuracy of calculating the remaining charging capacity and to improve the accuracy of predicting micro-short circuit anomalies.

[0051] For example, the preset charging time is 30 minutes. That is, the continuous charging time must not be less than 30 minutes.

[0052] In one possible implementation, step 400 may also include: after full charging is completed, determining whether the charging rate is less than a preset rate. If the charging rate is less than the preset rate, the lithium battery is determined to be effectively fully charged.

[0053] For example, the charging rate is a measure of how fast a battery charges, referring to the current required to charge the battery to its rated capacity within a specified time.

[0054] It should be further noted that during the research and development process, the applicant of this invention also discovered that when the charging rate is too high, the voltage change rate at the end of the charging of the first fully charged cell is large. When charging stops, the voltage of other cells may still be in the plateau region, resulting in a large error in calculating the remaining charging time of other cells, which in turn leads to a decrease in the accuracy of the remaining charging capacity and a high prediction error rate. Therefore, the effective full-charge judgment condition of this invention includes a charging rate less than a preset rate, in order to improve the accuracy of calculating the remaining charging capacity and to improve the accuracy of micro-short circuit anomaly prediction.

[0055] For example, the preset value of the charging rate can be determined based on experimental data of this type of battery cell. The preset value of the charging rate may vary for different battery cells.

[0056] It is understood that in some other embodiments, both of the above conditions must be met simultaneously to be considered a valid full charge. Furthermore, charging time and charging rate are two charging process parameters that the applicant of this invention has currently discovered that can affect the charging voltage curve of a fully charged cell. This invention does not exclude the possibility that other charging process parameters may also affect the degree to which the charging voltage curve of a fully charged cell approaches the standard charging voltage curve at the end of charging.

[0057] Therefore, the method and apparatus provided by this invention not only propose the above-mentioned scheme for determining whether a battery cell has experienced a micro-short circuit failure based on the difference in remaining charging capacity after adjacent full charges, but more importantly, it also proposes that when using the above scheme to predict micro-short circuit failure of a battery cell, certain restrictions need to be placed on the conditions for calculating the remaining charging capacity in order to avoid erroneous judgments as much as possible. Specifically, before each calculation of the remaining charging capacity, the method and apparatus of this invention will first determine whether a valid full charge has been achieved. Only when the charging voltage curve formed by a fully charged cell after the completion of this charge is close to the standard charging voltage curve of this type of battery cell as a whole will it be considered as a valid full charge. In particular, the full charge voltages of the curves are close and / or the trends at the ends of the curves are close. This will enable a more accurate value to be obtained when using the charging voltage curve of the fully charged cell to calculate the remaining charging capacity of each incompletely charged cell, thereby obtaining a more accurate calculation result and effectively avoiding erroneous prediction results.

[0058] Next, after obtaining the full charge data of each lithium battery, the prediction device can perform the following steps 401 to 405 to predict whether the lithium battery has experienced a cell micro-short circuit failure.

[0059] In step 401, when the lithium battery is effectively fully charged, the remaining charging capacity of each cell is calculated based on the full charge data.

[0060] Correspondingly, the remaining charging capacity corresponds to the remaining charging capacity of a fully charged cell and a partially charged cell when the lithium battery is fully charged. Therefore, the remaining charging capacity of each cell can only be calculated when the lithium battery is at least fully charged in a conventional manner. Furthermore, when the lithium battery is fully charged, the remaining charging capacity of a fully charged cell is 0, while the remaining charging capacity of a partially charged cell can be calculated based on the charging curve transformation of the fully charged cell. Specifically, as mentioned earlier, based on the high correlation between the charging processes of fully charged and partially charged cells, the remaining charging time of each cell can be obtained using the fully charged cell as a benchmark, and thus the remaining charging capacity can be obtained. This will be discussed in detail below.

[0061] In step 402, if the number of valid full charges in this valid full charge is greater than 1, the remaining charging capacity of each cell at the time of the last valid full charge is obtained.

[0062] In some embodiments, the above-mentioned effective full charge counts refer to the effective full charge counts used to record the remaining charging capacity of each cell.

[0063] It should be noted that the lithium battery can be a battery undergoing its first charge or a battery that has been used for a period of time and has undergone multiple charge-discharge cycles. In this embodiment of the invention, during the charging and discharging process of a lithium battery, if the remaining charging capacity of each cell is obtained for the first effective full charge in step 401 (i.e., the number of effective full charges in step 402 is 1), then step 401 continues to be executed. When the lithium battery is effectively fully charged again, the remaining charging capacity of each cell is calculated for that next effective full charge. Thus, the difference in remaining charging capacity between two adjacent effective full charges can be obtained in subsequent steps.

[0064] For example, the remaining charging capacity of each cell is expressed as C. n,i For example, C n,1 C n,2 C n,M , where n represents the nth charging that meets the conditions, i represents the i-th battery cell, and M is the total number of battery cells.

[0065] In step 403, for each cell, the difference between the remaining charging capacity of the cell during the current effective full charge and the previous effective full charge is calculated.

[0066] For example, the difference in remaining charging capacity can be expressed as ΔC. n,i For example, ΔC n,1 ΔC n,2 , …, ΔC n,M , where n represents the nth charging that meets the conditions, i represents the i-th battery cell, and M is the total number of battery cells.

[0067] It should be noted that in lithium batteries, the RCC value of a micro-short-circuited cell increases continuously during the charging and discharging process. Therefore, the difference ΔC between the RCC values ​​of two consecutive effective full charges of the same cell is calculated. n,i =C n,i -C n-1,i ΔC of micro-short-circuited cells n,i The difference will be greater than that of a normal cell. That is, among the RCC differences between two consecutive effective full charges of each cell, the difference of the micro-short-circuited cell will be greater than that of the normal cell, which can be used as a basis for determining micro-short circuits.

[0068] In step 404, abnormal differences and corresponding cells are identified from the differences in the remaining charging capacity of each cell.

[0069] In some embodiments, the difference in remaining charging capacity of a normal battery cell is smaller than the difference in remaining charging capacity of an abnormal battery cell. For example, the change in remaining charging capacity of a normal battery cell is small, while the abnormal battery cell has a larger difference in remaining charging capacity between two effective full charges due to the consumption of charging capacity by a micro-short circuit.

[0070] In some embodiments, a fixed difference threshold is preset. Among the differences in the remaining charging capacity of each cell, differences exceeding the aforementioned difference threshold are identified as abnormal differences.

[0071] In some embodiments, determining the corresponding cell includes obtaining the cell number corresponding to the abnormal difference.

[0072] In step 405, based on the abnormal difference and the corresponding cell, the cell with micro-short circuit failure in the lithium battery is identified and the lithium battery is determined to have experienced cell micro-short circuit failure.

[0073] In some embodiments, for a given battery cell, an abnormal difference between the remaining charge capacity after two consecutive full charges indicates that the cell has significant charging losses during charging. Since the difference is greater for a micro-short-circuited cell than for a normal cell, this suggests that the cell is more likely to be micro-short-circuited.

[0074] In some embodiments, if the difference in remaining charging capacity of any cell is determined to be abnormal, then the cell is determined to have a micro-short circuit failure.

[0075] This invention is applicable to lithium battery packs with varying health levels. Theoretically, if the health of a single cell in a lithium battery pack remains constant, its remaining charging capacity will always remain unchanged. However, in actual use, the cell's health gradually decreases, potentially affecting its remaining charging capacity. In this invention, the difference in remaining charging capacity between two adjacent charging cycles is used for prediction. Since the cell's health does not change significantly between adjacent charging cycles, the difference is essentially unaffected by the cell's health. Therefore, this solution is applicable to battery packs with varying health levels.

[0076] It should also be noted that the embodiments of the present invention require full charging. If the lithium battery pack is charged at a low frequency, the amount of data will be insufficient, making it impossible to predict micro-short circuits in a timely manner. This issue will not cause false alarms for micro-short circuits, but may only result in missed alarms. Accordingly, during actual battery use, prompts can be made to periodically charge the battery to perform battery maintenance, increasing the data available for micro-short circuit prediction.

[0077] The following explains the calculation of the remaining charging capacity in step 401. The remaining charging capacity is the difference between the total capacity of the cells when the lithium battery pack is fully charged and the current capacity, which is difficult to measure directly. This embodiment of the invention uses the cell charging voltage curve for indirect calculation.

[0078] The main factors affecting the charging voltage curve of a battery cell are total capacity, internal resistance, and initial capacity. When the total capacity, internal resistance, and initial capacity of different cells in a lithium battery pack are the same, their charging voltage curves completely overlap. When the total capacity, internal resistance, and initial capacity are different, their charging voltage curves can overlap after translation and scaling.

[0079] Figure 4 This is a schematic diagram illustrating the transformation of cell voltage curves with different parameters provided in an embodiment of the present invention. (Refer to...) Figure 4 The figure shows the charging voltage curves of two battery cells (cell A and cell B) with different total capacity, internal resistance, and initial capacity. From ① to ②, the cell B curve eliminates the effect of inconsistent initial capacity by horizontally shifting ΔAh. From ② to ③, the cell B curve eliminates the effect of inconsistent internal resistance by vertically shifting ΔU. From ③ to ④, the cell B curve eliminates the effect of inconsistent total capacity by scaling. After transformation, the cell B curve overlaps with the cell A curve.

[0080] Therefore, this embodiment of the invention uses the voltage curve of the first fully charged cell during charging as a benchmark to analyze the charging time required for other cells to reach the charging cutoff voltage under conditions where they can continue charging, and calculates the remaining charging capacity. The following describes the method for calculating the remaining charging capacity of cells in the lithium battery pack that are not fully charged at the end of charging.

[0081] Referring to the above, when the total capacity, internal resistance, and initial capacity of the cells in a lithium battery pack are all consistent with the first fully charged cell, the voltage curve of that cell will be completely consistent with the first fully charged cell, i.e., RCC will be 0. Next, we will analyze the situation where the cell parameters are inconsistent with the first fully charged cell.

[0082] The following is a specific method for calculating the remaining charging capacity of each battery cell:

[0083] In one possible implementation, when the lithium battery is effectively fully charged, the remaining charging capacity of each cell at this effective full charge includes:

[0084] In step 4011, when the lithium battery is effectively fully charged, a charging voltage curve for each cell is formed based on the full charge data; wherein, the lithium battery when effectively fully charged includes a fully charged cell and multiple cells that are not fully charged.

[0085] Figure 5 This is a schematic diagram of the voltage curves of each cell in a lithium battery provided in an embodiment of the present invention. (Refer to...) Figure 5 The lithium battery consists of four cells, Cell1 to Cell4. The solid lines represent the actual voltage curves of each cell. At time t1, cell Cell1 (corresponding to...) Figure 5 The leftmost curve in the diagram reaches the charging cutoff voltage first, at which point the BMS stops charging to prevent overcharging. Cell1 is a fully charged cell, while Cell2, Cell3, and Cell4 are not fully charged cells.

[0086] In step 4012, for any cell that is not fully charged, the charging voltage curve of a fully charged cell is used as a reference curve. The reference curve is transformed so that the transformed reference curve overlaps with the charging voltage curve of the cell that is not fully charged, thus obtaining the full charge curve of the cell that is not fully charged.

[0087] In some embodiments, the initial capacity of a partially charged cell is different from that of a fully charged cell, while the total capacity and internal resistance are the same. For example, the total capacity and internal resistance of cell 3 are the same as those of cell 1, but its initial capacity is lower than that of cell 1. Therefore, when charging ends, cell 3 has not reached the charging cutoff voltage, and the voltage curve corresponds to... Figure 5 The rightmost solid curve. According to the transformation method above, the voltage curve of cell 1 from time t1-Δt3 to time t1 can be shifted to the right to obtain the dashed part of cell 3, and then the complete voltage curve of cell 3 at full charge can be obtained.

[0088] In some embodiments, the total capacity and initial capacity of a partially charged cell are the same as those of a fully charged cell, but their internal resistance is different. For example, the total capacity and initial capacity of cell 2 are the same as those of cell 1, but its internal resistance is lower than that of cell 1. According to the transformation method described above, the cell 1 curve can be transformed downwards to obtain the complete voltage curve of cell 2 when fully charged.

[0089] In some embodiments, the internal resistance and initial capacity of a partially charged cell are the same as those of a fully charged cell, but their total capacity is different. For example, the internal resistance and initial capacity of cell 4 are the same as those of cell 1, but their total capacity is higher than that of cell 1. According to the transformation method described above, the cell 1 curve can be scaled to obtain the complete voltage curve of cell 4 when fully charged.

[0090] In some embodiments, considering that lithium battery packs are screened during practical applications, the difference in cell internal resistance and total capacity within the same battery pack is relatively small. For example, a 10% difference in RCC is within an acceptable range. For simplicity, the impact of the difference in internal resistance and total capacity on the RCC value can be ignored. That is, the total capacity and internal resistance of all cells in the battery pack are considered to be the same, and the voltage curve of the cell that is fully charged first is used as the reference curve to calculate the RCC value of the remaining cells.

[0091] In step 4013, the remaining charging time of the partially charged cell is determined based on the full charge curve of the partially charged cell.

[0092] For example, Δt3 represents the remaining charging time.

[0093] In some embodiments, the first full-charge time of a fully charged cell is obtained based on its full-charge curve. The second full-charge time of a partially charged cell is obtained based on its full-charge curve. The remaining charging time of the partially charged cell is obtained by subtracting the first full-charge time from the second full-charge time.

[0094] In step 4014, the remaining charging capacity of the battery cell is obtained based on the remaining charging time and charging current.

[0095] Considering that the current for each charge may not be the same in actual operation, it is not possible to directly use the change in the remaining charging time as the basis for judging whether a micro short circuit has occurred. Instead, the RCC value is calculated based on this time to make the judgment.

[0096] In one possible implementation, the remaining charging capacity of the battery cell is obtained based on the remaining charging time and charging current, including by using the following formula:

[0097]

[0098] Among them, C i t1 represents the remaining charging capacity of the battery cell, t1 represents the moment when the cell is fully charged, and Δt represents the charging time of the battery cell. i This indicates the remaining charging time of the battery cell, and I represents the charging current.

[0099] The following is another way to calculate the remaining charging capacity of each battery cell:

[0100] In one possible implementation, when the lithium battery is effectively fully charged, based on the full-charge voltage of each cell in the full-charge charging data at the moment of full charge, for any cell that is not fully charged, during the charging process of the fully charged cell, the charging time of the fully charged cell when its full-charge voltage is the same as that of the not-fully-charged cell is determined; the duration between the aforementioned charging time and the moment the lithium battery is fully charged is determined as the remaining charging time of the not-fully-charged cell; based on the remaining charging time and charging current, the remaining charging capacity of the cell is obtained using the above formula. It is worth noting that, in the full-charge voltage of each cell at the moment of full charge, for the not-fully-charged cell, its full-charge voltage means its voltage value at the moment of full charge.

[0101] It can be seen that, regardless of the method, the embodiments of the present invention have high requirements for the timeliness of cell data. For example, when acquiring the full-charge voltage of each cell at the moment of full charging of the lithium battery, the voltage of the cells that are not fully charged drops rapidly at the end of charging. If the voltage of the cells that are not fully charged at that moment cannot be accurately obtained, it will affect the calculation of the remaining charging capacity of the cell, and thus affect the accuracy of the prediction. This is why the embodiments of the present invention use a secondary BMS to directly record the full-charge data and use a prediction device with strong computing power to calculate the remaining charging capacity of each cell of the lithium battery to predict whether cell failure has occurred. This satisfies the data timeliness requirement, avoids calculation errors caused by data delay, and also meets the calculation requirements by utilizing a prediction device with strong computing power.

[0102] Regarding step 404, how to determine abnormal differences, the following is an example based on the box plot method.

[0103] Figure 6This is a schematic diagram of a box-type structure provided in an embodiment of the present invention. (Refer to...) Figure 6 :

[0104] In one possible implementation, identifying abnormal differences and corresponding cells among the differences in remaining charging capacity of each cell includes: determining whether abnormal differences exist among the differences in remaining charging capacity of each cell using a box plot analysis method. If abnormal differences exist, the cell corresponding to the abnormal difference is then identified.

[0105] A box plot is a statistical chart used to display the dispersion of a set of data. Its advantage is that it is not affected by outliers and can accurately and stably depict the discrete distribution of data.

[0106] This invention uses a box plot to perform a consistency analysis on the remaining charging capacity difference of all battery cells, determining whether there is any abnormal data, and recording the corresponding cell number if so. The calculation of the remaining charging capacity difference and box plot analysis only needs to be performed after each effective full charge, thus reducing the computational load and resource consumption.

[0107] It should be noted that during charge-discharge cycles, the remaining charge capacity of each cell continuously changes, and the difference in remaining charge capacity between cells also continuously changes. For lithium batteries of different models and used in different environments, it is difficult to determine a fixed threshold to judge whether the aforementioned differences in remaining charge capacity are abnormal. Therefore, this embodiment of the invention uses a box plot to determine abnormal values ​​based on the comparison of the RCC difference data of each cell. As the remaining charge capacity of each cell continuously changes during charge-discharge cycles, the difference between normal and abnormal cells gradually increases, and the corresponding upper and lower limits for abnormal judgment also continuously change, thus avoiding the need to set a fixed judgment threshold and preventing misjudgments.

[0108] The following examples illustrate the specific process of determining outlier values ​​using box plot analysis.

[0109] In one possible implementation, determining whether there are abnormal differences in the remaining charging capacity of each cell based on box plot analysis includes:

[0110] In step 501, based on the difference in the remaining charging capacity of each cell, the upper quartile, median, and lower quartile of each difference are determined.

[0111] In some embodiments, the difference in remaining charging capacity of each cell, i.e., ΔC of all cells, is used. n,i After arranging all values ​​from smallest to largest, the 75th percentile is the upper quartile Q3, the 50th percentile is the median Q2, and the 25th percentile is the lower quartile Q1.

[0112] In step 502, the lower quartile is subtracted from the upper quartile to obtain the interquartile range.

[0113] In some embodiments, the interquartile range (IQR) is calculated based on the following formula: IQR = Q3 - Q1.

[0114] In step 503, the upper limit and lower limit are obtained based on the upper quartile, lower quartile, and interquartile range.

[0115] In some embodiments, the upper limit is calculated based on the following formula: Q3 + 1.5IQR.

[0116] In some embodiments, the lower limit is calculated based on the following formula: Q1 - 1.5IQR.

[0117] In step 504, if any difference is greater than the upper limit or less than the lower limit, then the difference is determined to be an abnormal difference.

[0118] In some embodiments, the ΔC of all battery cells is... n,i Values ​​exceeding the upper limit or falling below the lower limit are identified as abnormal differences.

[0119] In one possible implementation, after determining the abnormal difference and the corresponding cell, the method further includes: calculating the remaining charging capacity and difference of each cell during the next effective full charge of the lithium battery, and determining the abnormal difference and the corresponding cell. This process is repeated multiple times to determine the abnormal difference and the corresponding cell for each effective full charge.

[0120] Accordingly, based on the abnormal differences and the corresponding cells, the cells in the lithium battery that have micro-short-circuit failures are identified as follows: based on the abnormal differences of each effective full charge and the corresponding cells, cells in the lithium battery with an abnormal number greater than or equal to a preset number are identified. These cells with an abnormal number greater than or equal to the preset number are then identified as cells in the lithium battery that have micro-short-circuit failures.

[0121] For example, the preset number of times is greater than or equal to 2. More exemplarily, the preset number of times is 20. That is, a cell is determined to have a micro-short circuit only when the number of times a certain cell is recorded as having an abnormal difference is greater than or equal to 20.

[0122] It should be noted that during the research and development process, the applicant of this invention discovered that when a lithium battery experiences a micro-short circuit, it typically does not show a significant difference from a normal battery within a single operating cycle. Therefore, in step 405, if a cell is judged as having a micro-short circuit abnormality simply because it experiences only one RCC difference anomaly, there is a possibility of misjudgment.

[0123] Therefore, in this embodiment of the invention, after multiple full-charge cycles, cells with abnormal RCC differences greater than or equal to a preset number of cycles are identified as micro-short-circuit failure cells, reducing the probability of misjudgment and improving the accuracy of micro-short-circuit failure.

[0124] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0125] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for predicting cell failure in a lithium battery system, the lithium battery system comprising a lithium battery device having multiple lithium batteries and a lithium battery monitoring device, the lithium battery comprising multiple cells connected in series, the lithium battery monitoring device being connected to each of the lithium batteries and used to predict whether each lithium battery will experience a cell micro-short circuit failure. Its features are: The lithium battery monitoring device includes: a primary BMS located within the lithium battery, a secondary BMS corresponding to multiple lithium batteries and located within the lithium battery device to be directly connected to the primary BMS, and a prediction device communicating with each secondary BMS. The method includes: When the first-level BMS detects that the charging voltage of any cell exceeds the charging cutoff voltage, it determines that the charging process is a full charge and controls all cells of the corresponding lithium battery to stop charging. The secondary BMS records the full charge data of each lithium battery during each full charge; wherein, the full charge data includes: the charging voltage and charging current of each cell at several moments during this full charge process, as well as the full charge voltage at the moment of full charge. The prediction device acquires the full-charge data of each lithium battery to calculate the remaining charging capacity of each cell of the lithium battery during the current full charge. When the number of full charges during the current full charge is greater than 1, the prediction device calculates the difference between the remaining charging capacity of each cell during the current full charge and the remaining charging capacity during the previous full charge for each cell of the lithium battery. The prediction device also identifies abnormal differences and corresponding cells in the differences of the remaining charging capacity of each cell, and based on the abnormal differences and corresponding cells, determines that the lithium battery has experienced a cell micro-short circuit failure and identifies the cell in the lithium battery that has experienced a micro-short circuit failure.

2. The cell failure prediction method for a lithium battery system according to claim 1, characterized in that, The prediction device is a cloud server, which is located outside the lithium battery device and is communicatively connected to each of the secondary BMS.

3. The cell failure prediction method for a lithium battery system according to claim 2, characterized in that, Each time the lithium battery is fully charged, the lithium battery includes at least one fully charged cell and multiple partially charged cells; The prediction device calculates the remaining charging capacity of each cell of the lithium battery during each full charge by the following steps: The charging voltage curve for each cell is generated based on the full charge data. For any partially charged cell, the charging voltage curve of the fully charged cell is used as a reference curve. The reference curve is transformed so that the transformed reference curve overlaps with the charging voltage curve of the partially charged cell to obtain the full charge curve of the partially charged cell. Based on the full charge curve of the partially charged cell, the remaining charging time of the partially charged cell is determined. Based on the remaining charging time and charging current, the remaining charging capacity of the battery cell is obtained.

4. The cell failure prediction method for a lithium battery system according to claim 2, characterized in that, Each time the lithium battery is fully charged, the lithium battery includes at least one fully charged cell and multiple partially charged cells; The prediction device calculates the remaining charging capacity of each cell of the lithium battery during each full charge by the following steps: Based on the full charge voltage of each cell at the full charge time in the full charge data, for any cell that is not fully charged, the charging time of the fully charged cell when it is the same as the full charge voltage of the not fully charged cell is determined during the charging process of the fully charged cell. The time between the charging time and the full charge time of the lithium battery is determined as the remaining charging time of the uncharged cell. Based on the remaining charging time and charging current, the remaining charging capacity of the battery cell is obtained.

5. The cell failure prediction method for a lithium battery system according to claim 3 or 4, characterized in that, The process of obtaining the remaining charging capacity of the battery cell based on the remaining charging time and charging current includes: The remaining charging capacity of the battery cell can be obtained based on the following formula: Among them, C i t1 represents the remaining charging capacity of the battery cell, t1 represents the moment when the cell is fully charged, and Δt represents the charging time of the battery cell. i This indicates the remaining charging time of the battery cell, and I represents the charging current.

6. The cell failure prediction method for a lithium battery system according to claim 2, characterized in that, The prediction device identifies abnormal differences and corresponding battery cells from the differences in the remaining charging capacity of each battery cell, including: Based on box plot analysis, determine whether there are any abnormal differences in the remaining charging capacity of each cell; If an abnormal difference exists, the corresponding battery cell is determined based on the abnormal difference. The determination of whether there are abnormal differences in the remaining charging capacity of each cell based on the box plot analysis method includes: Based on the difference in the remaining charging capacity of each cell, the upper quartile, median, and lower quartile of each difference are determined. Subtract the lower quartile from the upper quartile to obtain the interquartile range; Based on the upper quartile, lower quartile, and interquartile range, the upper and lower limits are obtained; If any difference is greater than the upper limit or less than the lower limit, then the difference is identified as an abnormal difference; The prediction device, based on the abnormal difference and the corresponding battery cell, determines that the lithium battery has experienced a micro-short circuit failure and identifies the battery cells within the lithium battery that have experienced the micro-short circuit failure, including... After multiple full charges, the abnormal differences in each full charge and the corresponding battery cells were determined. Based on the abnormal difference of each full charge and the corresponding cell, identify the cells in the lithium battery that have an abnormal number greater than or equal to a preset number. Cells with an abnormal number greater than or equal to a preset number are identified as cells with micro-short circuit failure in lithium batteries, and it is determined that the lithium battery has experienced cell micro-short circuit failure.

7. The cell failure prediction method for a lithium battery system according to claim 1 or 2, characterized in that: When each lithium battery is fully charged, the secondary BMS determines whether the charging constitutes a valid full charge, and records the full charge data of each lithium battery for each full charge when a valid full charge is achieved. Each time a lithium battery is fully charged, the lithium battery includes at least one fully charged cell and multiple partially charged cells; the effective full charge is used to characterize the charging voltage curve of the fully charged cell as being close to the full charge voltage and / or the end trend of the standard charging voltage curve of the cell.

8. The cell failure prediction method for a lithium battery system according to claim 7, characterized in that: The standard charging voltage curve of the battery cell is the charging voltage curve of the battery cell after it is fully discharged and then fully charged. The trend at the end of the two charging voltage curves is close to the following: within a preset time before the full charge time, the difference in the overall voltage change rate of the two charging voltage curves is less than a preset change rate difference threshold.

9. The cell failure prediction method for a lithium battery system according to claim 8, characterized in that, The determination of whether this charging constitutes a valid full charge includes: After full charging is complete, check if the charging time is greater than the preset time. If the charging time exceeds the preset time, the lithium battery is determined to be effectively fully charged; and / or After full charging is complete, check if the charging rate is less than the preset rate. If the charging rate is less than the preset rate, the lithium battery is determined to be effectively fully charged.

10. A lithium battery system, comprising: A lithium battery device and a lithium battery monitoring device having multiple lithium batteries; the lithium batteries include multiple cells connected in series; characterized in that: the lithium battery monitoring device is connected to each of the lithium batteries and includes: a primary BMS disposed within the lithium batteries, a secondary BMS corresponding to the multiple lithium batteries and disposed within the lithium battery device to be directly communicatively connected to the primary BMS, and a prediction device communicating with each secondary BMS; the lithium battery monitoring device uses the method as described in any one of claims 1-9 to predict whether a cell micro-short circuit failure has occurred in each lithium battery.

Images