CN116243165A - Method and device for determining consistency of batteries, computing equipment and vehicle - Google Patents
Method and device for determining consistency of batteries, computing equipment and vehicle Download PDFInfo
- Publication number
- CN116243165A CN116243165A CN202111482070.1A CN202111482070A CN116243165A CN 116243165 A CN116243165 A CN 116243165A CN 202111482070 A CN202111482070 A CN 202111482070A CN 116243165 A CN116243165 A CN 116243165A
- Authority
- CN
- China
- Prior art keywords
- battery
- data object
- battery cells
- abnormal
- voltage values
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The invention relates to a method, a device, a computing device and a vehicle for determining battery consistency. A method of determining battery consistency, comprising: acquiring voltage values of a plurality of battery units in a battery pack in a first time period, wherein the first time period comprises a plurality of moments; calculating a corresponding voltage information entropy at each of the plurality of times based on the voltage values of the plurality of battery cells at each of the plurality of times; based on a plurality of voltage information entropies corresponding to a plurality of moments, it is determined whether or not a plurality of battery cells have inconsistencies. According to the invention, whether the battery unit has consistency is determined by combining the voltage information entropy, so that the consistency of the battery unit can be dynamically determined, and the accuracy of consistency evaluation of the battery unit is improved, so that a user can take processing measures in advance, and the running safety of equipment is improved.
Description
Technical Field
The present disclosure relates to the field of battery technology, and in particular, to a method, an apparatus, a computing device, and a vehicle for determining battery consistency.
Background
With the increasing prominence of energy and environmental concerns, battery packs or stacks (e.g., lithium ion battery packs, etc.) are increasingly being employed in a variety of devices (e.g., wind power generation devices, vehicles such as hybrid and electric vehicles, etc.). Battery packs are typically composed of a series connection of different numbers of battery cells, each of which, after being manufactured, is subject to certain differences in initial performance (e.g., voltage, capacity, internal resistance, lifetime, temperature effects, self-discharge rate, etc.) itself, due to factors such as ambient temperature, humidity, etc. These performance differences accumulate as the batteries are used, and also result in gradual enlargement of the non-uniformity of the battery cells (e.g., causing different battery degradation) due to the non-uniform use environments of the battery cells within the battery pack. This inconsistency of the battery cells causes the battery pack to be seriously damaged by overcharge or discharge of the battery cells, greatly reducing the life span and the safety of use of the battery pack.
In general, the consistency of battery units is that a battery manufacturing enterprise tests an ac internal resistance value, a capacity and an open-circuit voltage of each battery unit before leaving a factory, evaluates the consistency of the battery units according to the ac internal resistance value, the capacity and the open-circuit voltage obtained by the test, and then selects the battery units with similar ac internal resistance value, capacity and open-circuit voltage to form a battery system. However, this method can only evaluate the consistency of the battery cells before assembly into a battery pack.
Batteries are a core component of devices such as electric vehicles, and a battery system is also one of the main factors of the failure of the devices. Therefore, how to accurately determine the battery uniformity after the battery cells are assembled into a battery system is a technical problem to be solved.
Disclosure of Invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and to provide a method, apparatus, computing device and vehicle for determining battery consistency.
According to a first aspect of the present invention there is provided a method of determining battery consistency comprising: the method comprises the steps of obtaining voltage values of a plurality of battery units in a battery pack in a first time period, wherein the first time period comprises a plurality of moments; a calculation step of calculating a corresponding voltage information entropy at each of the plurality of times based on voltage values of the plurality of battery cells at each of the plurality of times; and a determining step of determining whether or not there is inconsistency in the plurality of battery cells based on a plurality of voltage information entropies corresponding to the plurality of times.
According to a preferred embodiment of the present invention, determining whether there is an inconsistency in the plurality of battery cells based on a plurality of voltage information entropies corresponding to the plurality of time instants comprises: and if the voltage information entropies are smaller than a preset information entropy threshold, determining that the battery units are not inconsistent, or if at least two voltage information entropies in the voltage information entropies are larger than or equal to the preset information entropy threshold, determining that the battery units are inconsistent.
According to a preferred embodiment of the invention, the method further comprises: analyzing, based on a predetermined criterion, voltage values of the plurality of battery cells at each of at least two times corresponding to the at least two voltage information entropies to determine a respective abnormal set of battery cells having abnormal voltage values; and predicting a faulty set of battery cells based on the respective abnormal set of battery cells determined at the at least two moments in time.
According to a preferred embodiment of the present invention, analyzing the voltage values of the plurality of battery cells at each of the at least two moments in time based on a predetermined criterion to determine a respective abnormal set of battery cells having abnormal voltage values comprises: taking a plurality of voltage values corresponding to the plurality of battery units at each moment as a data object cluster; calculating the average value and standard deviation of the data object cluster; determining a respective absolute difference between each data object in the cluster of data objects and the average value; and identifying the data object with the corresponding absolute difference value meeting the preset deviation condition as an abnormal data object based on the standard deviation.
According to a preferred embodiment of the present invention, analyzing the voltage values of the plurality of battery cells at each of the at least two moments in time based on a predetermined criterion to determine a respective abnormal set of battery cells having abnormal voltage values comprises: taking a plurality of voltage values corresponding to the plurality of battery units at each moment as a data object cluster; determining a neighborhood of each data object in the data object cluster, wherein the neighborhood comprises a neighbor object corresponding to each data object; calculating local reachable densities of the data objects and the neighbor objects in the neighbor domain; calculating local outlier factors of each data object in the data object cluster based on the local reachable densities of each data object and the neighbor points, wherein the local outlier factors are the ratio of the average value of the local reachable densities of the neighbor objects of each data object in the neighborhood to the local reachable density of each data object; and identifying the data object with the local outlier meeting the preset outlier condition as an abnormal data object.
According to a preferred embodiment of the present invention, analyzing the voltage values of the plurality of battery cells at each of the at least two moments in time based on a predetermined criterion to determine a respective abnormal set of battery cells having abnormal voltage values comprises: taking a plurality of voltage values corresponding to the plurality of battery units at each moment as a data object cluster; constructing an isolated forest model for the data object cluster; calculating the corresponding average path length of each data object in the data object cluster in the constructed isolated forest model; determining an anomaly score value for each data object based on the respective average path length for said each data object; based on the abnormal score value of each data object in the data cluster, abnormal data objects in the data object cluster are identified.
According to a preferred embodiment of the present invention, predicting a faulty battery cell set based on the respective abnormal battery cell sets determined at the at least two moments in time comprises: determining the abnormal times or abnormal frequencies of each abnormal battery cell; the set of failed battery cells is predicted based on a comparison of the number of anomalies or the frequency of anomalies for each abnormal battery cell with a preset anomaly threshold.
According to a preferred embodiment of the invention, the predetermined criterion comprises a plurality of different predetermined criteria, and the method further comprises: a union operation, an intersection operation, or a voting operation is performed on a plurality of faulty battery cell sets predicted based on the plurality of different predetermined criteria to determine an updated faulty battery cell set.
According to a preferred embodiment of the invention, the method further comprises: selecting one or more battery cells from the predicted set of failed battery cells; obtaining voltage values of remaining battery cells in the battery pack by removing the voltage value of the selected one or more battery cells from the voltage values of the plurality of battery cells in the battery pack; the calculating step and the determining step are performed to determine whether there is an inconsistency in the remaining battery cells in the battery pack based on the voltage values of the remaining battery cells in the battery pack.
According to a preferred embodiment of the present invention, the preset information entropy threshold is determined according to at least one of a current, a temperature or an operation state of each battery cell in the battery pack.
According to a preferred embodiment of the invention, the first period of time comprises one or more of an open circuit phase, a charging phase, or a discharging phase.
According to a second aspect of the present invention there is provided an apparatus for determining battery consistency comprising: the device comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for acquiring voltage values of a plurality of battery units in a battery pack in a first time period, and the first time period comprises a plurality of moments; a calculation module for calculating a corresponding voltage information entropy at each of the plurality of times based on voltage values of the plurality of battery cells at each of the plurality of times; and a determining module for determining whether there is inconsistency in the plurality of battery cells based on a plurality of voltage information entropies corresponding to the plurality of moments.
According to a third aspect of the present invention there is provided a computing device comprising: at least one processor; and a memory for storing machine readable instructions that when executed cause the at least one processor to perform the method of determining battery consistency of the first aspect described above.
According to a fourth aspect of the present invention, there is provided a vehicle comprising: a battery pack; and an apparatus for determining battery consistency according to the second aspect or a computing device according to the third aspect.
According to the invention, the voltage information entropy is calculated by combining the voltage of the battery unit, and whether the battery unit has consistency is determined by utilizing the voltage information entropy, so that the consistency of the battery unit can be dynamically determined, the accuracy of evaluating the consistency of the battery unit is improved, a user can take processing measures in advance (for example, timely find and replace the battery unit with inconsistent state), the performance of the battery pack is improved, the service life of the battery pack is prolonged, the probability of potential safety hazards to equipment such as electric vehicles and the like caused by inconsistent state of the battery can be reduced, and the operation safety of the equipment is improved.
Drawings
Other features and advantages of the present invention will be better understood from the following detailed description of the preferred embodiment taken in conjunction with the accompanying drawings, in which like reference numerals identify the same or similar elements.
Fig. 1 shows a schematic diagram of an exemplary apparatus or system in which embodiments of the present disclosure may be applied.
Fig. 2 illustrates a flowchart of an exemplary method for determining battery consistency according to an embodiment of the present disclosure.
Fig. 3 illustrates a flowchart of an exemplary method for determining an abnormal set of battery cells, according to an embodiment of the present disclosure.
Fig. 4 illustrates a flowchart of another exemplary method for determining an abnormal set of battery cells, according to an embodiment of the present disclosure.
Fig. 5 illustrates a flowchart of another exemplary method for determining an abnormal set of battery cells, according to an embodiment of the present disclosure.
Fig. 6 illustrates a block diagram of an exemplary apparatus for determining battery consistency in accordance with an embodiment of the present disclosure.
FIG. 7 illustrates a block diagram of an exemplary computing device, according to an embodiment of the present disclosure.
Detailed Description
As described below, some example embodiments of the present disclosure provide methods, apparatuses, computing devices, and vehicles for determining battery uniformity, and more particularly, methods, apparatuses, computing devices, and vehicles including the same for evaluating uniformity of cells of a battery pack.
In the following description, a battery pack (battery pack) may include, but is not limited to, a lithium ion battery, a lead acid battery, and the like. A battery pack may be composed of a plurality of battery cells (cells) connected in series, each of which has a certain difference in initial performance (e.g., parameters of voltage, capacity, internal resistance, lifetime, temperature influence, self-discharge rate, etc.) itself, due to factors such as ambient temperature, humidity, etc., after being manufactured. As previously mentioned, this difference in performance may accumulate over the use of the battery. This inconsistency of the battery cells causes the battery pack to be seriously damaged by overcharge or discharge of the battery cells, greatly reducing the life span and the safety of use of the battery pack.
Referring to fig. 1, a schematic diagram of an exemplary apparatus or system 100 in which embodiments of the present disclosure may be applied is shown. The apparatus or system 100 may be a vehicle or a battery energy storage device (e.g., wind power plant, etc.), etc., and includes a battery pack 101. The battery pack 101 may include a plurality of battery cells.
In examples where the device or system 100 is a vehicle, exemplary vehicles may include electric vehicles (e.g., pure electric vehicles, hybrid vehicles, etc.), marine vehicles, aeronautical vehicles, or other vehicles that use batteries. The vehicle 100 may also include a component 102 coupled to the battery pack 101, such as a load powered by the battery pack 101. For example, the load 102 may be a motor or hybrid motor or the like that consumes energy provided by the battery pack 101 to propel the vehicle 100, so the battery pack 101 is a power or traction battery.
In examples where the apparatus or system 100 is a battery energy storage device, the example battery energy storage device may also include a component 102, such as an energy storage inverter, connected to the battery pack 101 as an interface of the battery pack 101 with the power grid.
As previously mentioned, each battery cell is typically tested by the battery manufacturer to evaluate the consistency of the battery cells before shipping, whether it is a vehicle or a grid energy storage device, for example, however, this method is not applicable to situations where the battery cells are assembled into a battery system. An improved solution for determining battery uniformity is described below in conjunction with fig. 2-7 to address the deficiencies of the prior art.
Fig. 2 illustrates an exemplary method 200 for determining battery consistency according to an embodiment of the present disclosure. Method 200 may include steps 210-230, and optionally steps 240-260.
In step 210, voltage values of a plurality of battery cells in a battery pack within a first time period are obtained, wherein the first time period includes a plurality of moments. For example, the voltage values of the plurality of battery cells in the battery pack collected by the vehicle may be obtained from a data acquisition terminal (e.g., TBox) of the vehicle over a first period of time, or from a database or other data storage device (e.g., a data lake or cloud in which data is transmitted and stored by the TBox), or obtained using a voltage sensor disposed at each battery cell.
In step 220, based on the voltage values of the plurality of battery cells at each of the plurality of times, a corresponding voltage information entropy at each time is calculated.
For example, assume that the battery pack includes m battery cells C1, C2, …, cm, and the first period includes n times T1, T2, …, tn. For each time Ti, the voltage information entropy is calculated from the voltages { V (Ti, cj), j=1, …, m } of the m battery cells using the following equation (1):
H(Ti)＝-∑Pk*log2(Pk) (1)
where Pk is the probability of occurrence of the voltage value. The value of H (Ti) approaches 0 if the uniformity of the battery cells is good, and the greater the value of H (Ti) if the uniformity of the battery cells is poor.
In step 230, it is determined whether there is an inconsistency in the plurality of battery cells based on the plurality of voltage information entropies corresponding to the plurality of time instants.
The voltage information entropy is calculated by combining the voltage of the battery unit, and whether the battery unit has consistency is determined by utilizing the voltage information entropy, so that the consistency of the battery unit can be dynamically determined, the accuracy of consistency evaluation of the battery unit is improved, a user can take processing measures in advance (for example, the inconsistency can be found out 24-72 hours in advance, and the battery unit with the inconsistency can be timely checked and found out), and the running safety of equipment is improved.
In some embodiments, prior to step 220, method 200 may further comprise: and performing data cleaning on the obtained voltage data to clean incomplete data, error data and repeated data, so that voltage information entropy can be calculated based on the cleaned voltage data. Through data cleaning, data consistency, processing missing values, invalid values, repeated values and the like can be checked, so that the reliability of subsequent calculation is improved.
In some embodiments, step 230 may include: and if the plurality of voltage information entropies are smaller than a preset information entropy threshold value, determining that the plurality of battery units are not inconsistent, or if at least two voltage information entropies in the plurality of voltage information entropies are larger than or equal to the preset information entropy threshold value, determining that the plurality of battery units are inconsistent. In the step, the inconsistency of the battery units is determined based on the fact that the voltage information entropy at a plurality of moments exceeds the preset information entropy threshold, and accuracy of consistency evaluation can be further improved. Further, the preset information entropy threshold may be determined according to at least one of a current, a temperature, or an operating state of each battery cell in the battery pack.
In some embodiments, after determining that there is an inconsistency in the battery cells based on step 230, the method 200 may further include step 240. In step 240, the voltage values of the plurality of battery cells at each of at least two times corresponding to the at least two voltage information entropies are analyzed to determine a respective set of abnormal battery cells having abnormal voltage values based on a predetermined criterion, and the set of failed battery cells is predicted based on the respective set of abnormal battery cells determined at the at least two times. In this step, the abnormal battery cells are identified by identifying the abnormal voltage data in the voltage value at each moment, so that the faulty battery cells are further predicted, so that the user can take processing measures in advance (for example, find and replace the battery cells predicted to be faulty in time), thereby improving the performance of the battery pack and prolonging the service life of the battery pack, and the probability of potential safety hazards caused by inconsistency of the battery cells can be reduced, thereby improving the operation safety of the device.
In some embodiments, step 240 may further comprise: determining the abnormal times or abnormal frequencies of each abnormal battery cell; the set of failed battery cells is predicted based on a comparison of the number of anomalies or the frequency of anomalies for each abnormal battery cell with a preset anomaly threshold. In this step, by counting the abnormal battery cell sets obtained at the time when there is inconsistency in the battery cells, the accuracy of prediction can be improved so as not to erroneously predict a failed battery cell sporadically.
In some embodiments, the predetermined criteria includes a plurality of different predetermined criteria, and the method 200 may further include step 250. At step 250, a union operation, an intersection operation, or a voting operation is performed on the plurality of failed battery cell sets predicted based on the plurality of different predetermined criteria to determine an updated failed battery cell set. For example, the predetermined criteria may include data distribution-based criteria (e.g., 3 sigma criteria based on probability distribution, etc.), data density-based criteria (e.g., LOF (Local Outlier Factor) outlier factor algorithm, etc.), ensemble-based criteria (e.g., isolation Forest (iForest) algorithm), etc. The plurality of sets of failed battery cells predicted based on the plurality of different predetermined criteria may be the same or different by, for example, taking intersections (i.e., taking the smallest possible common set), union (i.e., taking the largest possible set), or voting (e.g., counting the number of times a battery cell is predicted to fail, and counting it into the final set of failed battery cells when the number of times of failure exceeds a threshold). In this step, by performing an operation on the set of failed battery cells predicted under different criteria, the accuracy of prediction can be improved so as not to accidentally mispredict the failed battery cells.
In some embodiments, the method 200 may further include step 260. At step 260, one or more battery cells are selected from the predicted set of failed battery cells; obtaining voltage values of remaining battery cells in the battery pack by removing the voltage value of the selected one or more battery cells from the voltage values of the plurality of battery cells in the battery pack; steps 220 and 230 are performed to determine whether there is an inconsistency in the remaining battery cells in the battery pack based on the voltage values of the remaining battery cells in the battery pack. For example, the most likely to fail cells from the predicted set of failed cells may be screened and the voltage data for the number of cells may be removed from the voltage values for the plurality of cells in the battery pack, and steps 220 and 230, and optionally steps 240-260, may be re-performed to determine more likely to fail cells serially rather than in parallel. In this step, by repeatedly performing steps 220 to 260 in series, the malfunctioning battery cell is predicted stepwise, and the accuracy of the prediction can be improved to further find a battery cell that is likely to malfunction.
In some embodiments, the first period of time may include one or more of an open circuit phase, a charge phase, or a discharge phase. For example, the first period of time may occur in one or more of a variety of operating conditions of the device, such as an open circuit phase, a charge phase, or a discharge phase, as the battery cells may exhibit different characteristics in different states, such that the consistency of the battery cells may be dynamically determined and the accuracy of the consistency assessment may be improved by evaluating the various operating conditions.
Various methods 300-500 for determining abnormal cell sets under different criteria are described below in connection with fig. 3-5, and the methods 300-500 may be sub-steps of step 240 of method 200.
Fig. 3 illustrates a flowchart of an exemplary method 300 for determining an abnormal set of battery cells, according to an embodiment of the present disclosure. The method 300 is based on criteria of data distribution (e.g., 3 sigma criteria based on probability distribution, etc.), and includes steps 310-340.
In step 310, a plurality of voltage values corresponding to the plurality of battery cells at each time when the voltage information entropy is greater than or equal to the preset information entropy threshold are used as the data object cluster.
At step 320, the mean and standard deviation of the data object clusters are calculated.
At step 330, a respective absolute difference between each data object in the cluster of data objects and the average value is determined.
In step 340, the data object whose corresponding absolute difference satisfies the preset deviation condition is identified as an abnormal data object based on the standard deviation.
For example, for each time Tq, the average value E and standard deviation σ of the voltage values of a plurality of battery cells (e.g., m) may be calculated according to the following formulas (2) and (3):
for example, cells having voltage values outside E-3σ to E+3σ are identified as abnormal cells. Further, a failed battery cell set may be predicted by counting the abnormal battery cell sets obtained at a plurality of times when there is inconsistency in the battery cells, as described above with respect to step 240.
Fig. 4 illustrates a flowchart of another exemplary method 400 for determining an abnormal set of battery cells, according to an embodiment of the present disclosure. The method 400 is based on criteria of data density (e.g., based on LOF, etc.), and includes steps 410-450.
In step 410, a plurality of voltage values corresponding to the plurality of battery cells at each time when the voltage information entropy is greater than or equal to the preset information entropy threshold are used as the data object cluster.
In step 420, a neighborhood of each data object in the data object cluster is determined, the neighborhood including a neighbor object corresponding to the each data object.
At step 430, the local reachable densities of each data object and neighbor object within the neighborhood are calculated.
At step 440, a local outlier factor is calculated for each data object within the data object cluster based on the local reachable densities of each data object and the neighbor point, wherein the local outlier factor is a ratio of an average of the local reachable densities of the neighbor objects of each data object in the neighborhood to the local reachable density of each data object.
At step 450, data objects for which the local outlier satisfies the preset outlier condition are identified as outlier data objects.
For example, the neighborhood of a data object is a K-distance neighborhood, and all objects that are not more than the K-distance from the data object are neighbor objects (KNN) of the data object.
The local reachable density (local reachablity density, lrd) of each data object p can be calculated according to the following formula (4):
representing the inverse of the average reachable distance of the neighbor object to p within the kth neighborhood of data object p.
The local outlier factor (local outlier factor, LOF) of the data object p may be calculated according to the following equation (5):
neighborhood object N representing data object p k An average of the ratio of the locally reachable densities of (p) to the locally reachable densities of the data object p.
The closer the value of LOF is to 1, the more likely it is a normal sample, while the greater the value of LOF is to 1, the more likely it is an abnormal sample. Thus, the preset outlier factor condition may be: if LOF >1, the data object is an outlier and if LOF is close to 1, the data object is a normal data point, identifying an outlier data object.
Fig. 5 illustrates a flowchart of another exemplary method 500 for determining an abnormal set of battery cells, according to an embodiment of the present disclosure. The method 500 is based on criteria of data density (e.g., based on LOF, etc.), and includes steps 510-550.
In step 510, a plurality of voltage values corresponding to the plurality of battery cells at each time when the voltage information entropy is greater than or equal to the preset information entropy threshold value are used as the data object cluster.
In step 520, an isolated forest model is built for the data object cluster.
At step 530, a respective average path length of each data object in the cluster of data objects in the constructed isolated forest model is calculated.
At step 540, an anomaly score value for each data object is determined based on the respective average path length for each data object.
At step 550, the abnormal data objects in the data object cluster are identified based on the abnormal score value for each data object in the data cluster.
And (3) sampling the data object cluster t times, and establishing t isolated trees iTree through the randomly selected splitting values, so as to construct an isolated forest iForest model comprising the t isolated trees. Calculating the path length of each data object s in each isolated tree in the isolated forest iflastmodel, thereby obtaining the average path length h(s) of the data objects in the isolated forest iflastmodel, and calculating the abnormal score value of the data object s according to the average path length and the following formula (6):
wherein, C (n) is the average path length of a binary search tree constructed by n points, and the calculation formula is as follows:
in the above formula (7), H (k) =ln (k) +γ, and γ is euler constant 0.577215665.
The value of the outlier Score(s) is 0< Score(s) <1, and the larger the Score(s), the more outlier the Score(s) becomes. For example, a threshold value (e.g., 0.5) of the outlier score may be set, and a data object satisfying the outlier score greater than the threshold value may be determined as an outlier data object.
Fig. 6 illustrates a block diagram of an exemplary apparatus 600 for determining battery consistency in accordance with an embodiment of the present disclosure. The example apparatus 600 may be or be included in a Battery Management System (BMS) of the apparatus 100 (e.g., a vehicle) of fig. 1. The apparatus 600 includes an acquisition module 610, a calculation module 610, and a determination module 630.
The acquisition unit 610 is configured to: voltage values of a plurality of battery cells in the battery pack 101 are acquired within a first period of time, wherein the first period of time includes a plurality of times.
The computing module 620 is configured to: based on the voltage values of the plurality of battery cells at each of the plurality of times, a corresponding voltage information entropy at each time is calculated.
The determination module 630 is configured to: based on a plurality of voltage information entropies corresponding to a plurality of moments, it is determined whether or not a plurality of battery cells have inconsistencies.
In some embodiments, the acquisition unit 610 may be configured to perform various operations or processes as described above with respect to step 210.
In some embodiments, the computing unit 620 may be configured to perform various operations or processes as described above with respect to step 220.
In some embodiments, the determination unit 630 may be configured to perform various operations or processes as described above with respect to step 220.
In some embodiments, the apparatus 600 may further include a data cleansing module configured to: and performing data cleaning on the obtained voltage data to clean incomplete data, error data and repeated data, so that voltage information entropy can be calculated based on the cleaned voltage data.
In some embodiments, the apparatus 600 may further comprise a prediction module 640. The prediction module 640 is configured to: the method includes analyzing voltage values of the plurality of battery cells at each of at least two times corresponding to the at least two voltage information entropies based on a predetermined criterion to determine a respective set of abnormal battery cells having abnormal voltage values, and predicting a set of failed battery cells based on the respective set of abnormal battery cells determined at the at least two times.
In some embodiments, the prediction module 640 may be configured to perform various operations or processes as described above with respect to step 240.
In some embodiments, the apparatus 600 may further include an update module 650. The update module 650 is configured to: a union operation, an intersection operation, or a voting operation is performed on a plurality of faulty battery cell sets predicted based on a plurality of different predetermined criteria to determine an updated faulty battery cell set.
In some embodiments, the update module 650 may be configured to perform various operations or processes as described above with respect to step 250.
In some embodiments, the apparatus 600 may further include a selection module 660. The selection module 660 is configured to: selecting one or more battery cells from the predicted set of failed battery cells; obtaining voltage values of remaining battery cells in the battery pack by removing the voltage value of the selected one or more battery cells from the voltage values of the plurality of battery cells in the battery pack; based on the voltage values of the remaining battery cells in the battery pack, a calculation module 620 and a determination module 630 are executed to determine whether there is an inconsistency in the remaining battery cells in the battery pack.
In some embodiments, the selection module 660 may be configured to perform various operations or processes as described above with respect to step 260.
Fig. 7 illustrates a block diagram of an exemplary computing device 700, according to an embodiment of the disclosure. The computing device 700 includes at least one processor 701 and a memory 702 coupled with the at least one processor 701. The memory 702 is used to store machine-readable instructions that, when executed by the at least one processor 701, cause the processor 701 to perform the methods of the above embodiments (e.g., any one or more of the steps of the methods 200-500 described previously).
Referring back to fig. 1, the apparatus 100 may be a battery-powered vehicle that includes a battery pack 101, and the apparatus 100 may further include an apparatus 600 or computing device 700 for determining battery consistency as described with reference to fig. 6 and 7.
As discussed above, although methods of determining battery inconsistencies in vehicle applications are discussed herein, the methods, apparatus, and computing devices of the present disclosure are applicable to other arrangements in which it is desirable to determine battery inconsistencies.
The foregoing discussion is merely a disclosure and description of exemplary embodiments of the invention. From this discussion, and the accompanying drawings and claims, those skilled in the art will readily recognize that various changes, modifications and variations can be made therein without departing from the spirit and scope of the invention as defined in the following claims.
Claims (15)
1. A method of determining battery consistency, comprising the steps of:
the method comprises the steps of obtaining voltage values of a plurality of battery units in a battery pack in a first time period, wherein the first time period comprises a plurality of moments;
a calculation step of calculating a corresponding voltage information entropy at each of the plurality of times based on voltage values of the plurality of battery cells at each of the plurality of times; and
and a determining step of determining whether or not the plurality of battery cells are inconsistent based on a plurality of voltage information entropies corresponding to the plurality of times.
2. The method of claim 1, wherein determining whether there is an inconsistency for the plurality of battery cells based on a plurality of voltage information entropies corresponding to the plurality of time instants comprises:
if the voltage information entropy is smaller than the preset information entropy threshold, determining that the plurality of battery units are not inconsistent or
And if at least two voltage information entropies in the plurality of voltage information entropies are larger than or equal to the preset information entropy threshold value, determining that the plurality of battery units are inconsistent.
3. The method of claim 2, further comprising:
analyzing, based on a predetermined criterion, voltage values of the plurality of battery cells at each of at least two times corresponding to the at least two voltage information entropies to determine a respective abnormal set of battery cells having abnormal voltage values; and
based on the respective abnormal battery cell sets determined at the at least two moments, a failed battery cell set is predicted.
4. The method of claim 3, wherein analyzing the voltage values of the plurality of battery cells at each of the at least two times to determine a respective abnormal set of battery cells having abnormal voltage values based on a predetermined criterion comprises:
taking a plurality of voltage values corresponding to the plurality of battery units at each moment as a data object cluster;
calculating the average value and standard deviation of the data object cluster;
determining a respective absolute difference between each data object in the cluster of data objects and the average value;
and identifying the data object with the corresponding absolute difference value meeting the preset deviation condition as an abnormal data object based on the standard deviation.
5. The method of claim 3, wherein analyzing the voltage values of the plurality of battery cells at each of the at least two times to determine a respective abnormal set of battery cells having abnormal voltage values based on a predetermined criterion comprises:
taking a plurality of voltage values corresponding to the plurality of battery units at each moment as a data object cluster;
determining a neighborhood of each data object in the data object cluster, wherein the neighborhood comprises a neighbor object corresponding to each data object;
calculating local reachable densities of the data objects and the neighbor objects in the neighbor domain;
calculating local outlier factors of each data object in the data object cluster based on the local reachable densities of each data object and the neighbor objects, wherein the local outlier factors are the ratio of the average value of the local reachable densities of the neighbor objects of each data object in the neighborhood to the local reachable density of each data object;
and identifying the data object with the local outlier meeting the preset outlier condition as an abnormal data object.
6. The method of claim 3, wherein analyzing the voltage values of the plurality of battery cells at each of the at least two times to determine a respective abnormal set of battery cells having abnormal voltage values based on a predetermined criterion comprises:
taking a plurality of voltage values corresponding to the plurality of battery units at each moment as a data object cluster;
constructing an isolated forest model for the data object cluster;
calculating the corresponding average path length of each data object in the data object cluster in the constructed isolated forest model;
determining an anomaly score value for each data object based on the respective average path length for said each data object;
based on the abnormal score value of each data object in the data cluster, abnormal data objects in the data object cluster are identified.
7. The method of claim 3, wherein predicting a failed battery cell set based on the respective abnormal battery cell sets determined at the at least two moments in time comprises:
determining the abnormal times or abnormal frequencies of each abnormal battery cell;
the set of failed battery cells is predicted based on a comparison of the number of anomalies or the frequency of anomalies for each abnormal battery cell with a preset anomaly threshold.
8. The method of any of claims 3 to 7, wherein the predetermined criteria comprises a plurality of different predetermined criteria, and the method further comprises:
a union operation, an intersection operation, or a voting operation is performed on a plurality of faulty battery cell sets predicted based on the plurality of different predetermined criteria to determine an updated faulty battery cell set.
9. A method according to claim 3, further comprising:
selecting one or more battery cells from the predicted set of failed battery cells;
obtaining voltage values of remaining battery cells in the battery pack by removing the voltage value of the selected one or more battery cells from the voltage values of the plurality of battery cells in the battery pack;
the calculating step and the determining step are performed to determine whether there is an inconsistency in the remaining battery cells in the battery pack based on the voltage values of the remaining battery cells in the battery pack.
10. The method of claim 2, wherein the preset information entropy threshold is determined according to at least one of a current, a temperature, or an operating state of each battery cell in the battery pack.
11. The method of claim 1, wherein the first period of time comprises one or more of an open circuit phase, a charge phase, or a discharge phase.
12. The method of claim 1, wherein prior to the calculating step, the method further comprises:
and cleaning the obtained voltage values to remove incomplete data, error data and repeated data.
13. An apparatus for determining battery consistency, comprising:
the device comprises an acquisition module, a control module and a control module, wherein the acquisition module is configured to acquire voltage values of a plurality of battery units in a battery pack in a first time period, and the first time period comprises a plurality of moments;
a calculation module configured to calculate a corresponding voltage information entropy at each of the plurality of times based on voltage values of the plurality of battery cells at each of the plurality of times; and
a determination module configured to determine whether there is an inconsistency in the plurality of battery cells based on a plurality of voltage information entropies corresponding to the plurality of moments.
14. A computing device, comprising:
at least one processor; and
a memory storing machine-readable instructions that, when executed by the at least one processor, cause the at least one processor to perform the method of any of claims 1-12.
15. A vehicle, comprising:
a battery pack; and
the apparatus for determining battery consistency of claim 13 or the computing device of claim 14.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111482070.1A CN116243165A (en) | 2021-12-07 | 2021-12-07 | Method and device for determining consistency of batteries, computing equipment and vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111482070.1A CN116243165A (en) | 2021-12-07 | 2021-12-07 | Method and device for determining consistency of batteries, computing equipment and vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116243165A true CN116243165A (en) | 2023-06-09 |
Family
ID=86624668
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111482070.1A Pending CN116243165A (en) | 2021-12-07 | 2021-12-07 | Method and device for determining consistency of batteries, computing equipment and vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116243165A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116626505A (en) * | 2023-07-21 | 2023-08-22 | 江苏海平面数据科技有限公司 | Battery pack consistency anomaly detection method based on Internet of vehicles big data |
-
2021
- 2021-12-07 CN CN202111482070.1A patent/CN116243165A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116626505A (en) * | 2023-07-21 | 2023-08-22 | 江苏海平面数据科技有限公司 | Battery pack consistency anomaly detection method based on Internet of vehicles big data |
| CN116626505B (en) * | 2023-07-21 | 2023-10-13 | 江苏海平面数据科技有限公司 | Battery pack consistency anomaly detection method based on Internet of vehicles big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gan et al. | Data-driven fault diagnosis of lithium-ion battery overdischarge in electric vehicles | |
| Yang et al. | On-board diagnosis of soft short circuit fault in lithium-ion battery packs for electric vehicles using an extended Kalman filter | |
| Shahriari et al. | Online state-of-health estimation of VRLA batteries using state of charge | |
| Kim et al. | State-of-charge estimation and state-of-health prediction of a Li-ion degraded battery based on an EKF combined with a per-unit system | |
| Barcellona et al. | Analysis of ageing effect on Li‐polymer batteries | |
| US8340934B2 (en) | Method of performance analysis for VRLA battery | |
| Baumann et al. | Cloud-connected battery management for decision making on second-life of electric vehicle batteries | |
| CN113391213B (en) | Battery micro-fault diagnosis method based on battery pack consistency evolution | |
| Banaei et al. | Real time condition monitoring in Li-Ion batteries via battery impulse response | |
| Juang | Online battery monitoring for state-of-charge and power capability prediction | |
| Sangwan et al. | An empirical capacity degradation modeling and prognostics of remaining useful life of Li-ion battery using unscented Kalman filter | |
| CN111257753A (en) | Battery system fault diagnosis method | |
| Liu et al. | A comprehensive review of health indicators of li-ion battery for online state of health estimation | |
| Kim et al. | Outlier mining-based fault diagnosis for multiceli lithium-ion batteries using a low-priced microcontroller | |
| CN114503392A (en) | Determination device, power storage system, determination method, and determination program for a plurality of batteries | |
| Zhu et al. | Battery voltage prediction using neural networks | |
| CN116243165A (en) | Method and device for determining consistency of batteries, computing equipment and vehicle | |
| Ananto et al. | The state of health of Li-Po batteries based on the battery's parameters and a fuzzy logic system | |
| CN113391214A (en) | Battery micro-fault diagnosis method based on battery charging voltage ranking change | |
| Song et al. | Series-connected lithium-ion battery pack health modeling with cell inconsistency evaluation | |
| Fasih | Modeling and fault diagnosis of automotive lead-acid batteries | |
| CN116819364A (en) | Battery health determination method, application method, electronic device and storage medium | |
| CN113420494B (en) | Super-capacitor Bayes probability fusion modeling method | |
| US20230184838A1 (en) | Device and method for diagnosing battery | |
| Lin et al. | Research on Inconsistency Identification of Lithium-ion Battery Pack Based on Operational Data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |