CN115629323A - Battery pack fault detection method and device and vehicle - Google Patents
Battery pack fault detection method and device and vehicle Download PDFInfo
- Publication number
- CN115629323A CN115629323A CN202211159286.9A CN202211159286A CN115629323A CN 115629323 A CN115629323 A CN 115629323A CN 202211159286 A CN202211159286 A CN 202211159286A CN 115629323 A CN115629323 A CN 115629323A
- Authority
- CN
- China
- Prior art keywords
- data
- characteristic
- feature
- battery
- value
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 71
- 239000000178 monomer Substances 0.000 claims abstract description 75
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 18
- 230000002159 abnormal effect Effects 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims description 68
- 230000005856 abnormality Effects 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 10
- 238000000605 extraction Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 description 61
- 230000008569 process Effects 0.000 description 5
- 230000002349 favourable effect Effects 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000009825 accumulation Methods 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
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]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
-
- 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]
- G01R31/3644—Constructional arrangements
- G01R31/3648—Constructional arrangements comprising digital calculation means, e.g. for performing an algorithm
-
- 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]
- G01R31/389—Measuring internal impedance, internal conductance or related variables
-
- 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]
- G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
- H01M10/4285—Testing apparatus
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Chemical & Material Sciences (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Electrochemistry (AREA)
- General Chemical & Material Sciences (AREA)
- Secondary Cells (AREA)
Abstract
The invention provides a battery pack fault detection method, a battery pack fault detection device and a vehicle, wherein the method comprises the following steps: acquiring a characteristic data combination of each battery monomer in the battery pack, wherein the characteristic data combination comprises a plurality of data characteristics and a characteristic value time sequence corresponding to each data characteristic; for each data feature, comparing the feature value time sequence corresponding to the data feature in each feature data combination, and extracting the difference feature data of each battery cell on the data feature and other battery cells; for each battery monomer, combining all the difference characteristic data corresponding to the battery monomer to obtain difference characteristic combination data of the battery monomer; and based on an abnormal detection algorithm, judging whether the corresponding single battery has a fault according to the difference characteristic combination data. According to the technical scheme, the accuracy of battery pack fault detection is improved, and the single batteries with fault risks can be accurately detected.
Description
Technical Field
The invention relates to the technical field of battery fault detection, in particular to a battery pack fault detection method, a battery pack fault detection device and a vehicle.
Background
The battery pack is formed by connecting a plurality of battery monomers in series and/or in parallel, and is widely applied to the fields of new energy automobiles, consumer electronics and the like, for example, the new energy automobiles adopt the power battery pack to provide power for the automobiles. During the use of the battery pack, the performance of the battery pack is affected by the change of the use environment, the load and the like, and even the battery pack is in failure. Therefore, it is necessary to perform fault detection on the battery pack to secure the safety of the battery pack.
At present, a method for detecting a fault of a battery pack mainly includes: (1) a detection method based on an equivalent circuit model; (2) knowledge and statistics based detection methods; and (3) artificial intelligence algorithm based on data driving. The methods (1) and (2) are poor in instantaneity, mostly can only utilize a single data characteristic to analyze, and are poor in accuracy. The method (3) can utilize multidimensional data characteristics to diagnose faults, but cannot effectively determine the single battery with fault risks, and is not beneficial to subsequent accurate early warning and maintenance.
Disclosure of Invention
The invention solves the problems of how to improve the accuracy of fault detection of the battery pack and realize the fault detection of the single battery level.
In order to solve the problems, the invention provides a battery pack fault detection method, a battery pack fault detection device and a vehicle.
In a first aspect, the present invention provides a battery pack fault detection method, including:
acquiring a characteristic data combination of each battery in a battery pack, wherein the characteristic data combination comprises a plurality of data characteristics and a characteristic value time sequence corresponding to each data characteristic;
for each data feature, comparing the feature value time series corresponding to the data feature in each feature data combination, and extracting the difference feature data between each battery cell and other battery cells on the data feature;
for each battery monomer, combining all the difference characteristic data corresponding to the battery monomer to obtain difference characteristic combination data of the battery monomer;
and judging whether the corresponding battery monomer has a fault or not according to the difference characteristic combination data based on an abnormal detection algorithm.
Optionally, the time series of feature values includes feature values of the data feature at a plurality of different acquisition times;
for each data feature, comparing the time series of feature values corresponding to the data feature in each feature data combination, and extracting difference feature data between each battery cell and other battery cells on the data feature comprises:
for each data feature, determining a first mean value and a first standard deviation of the feature value at the same acquisition time in the corresponding feature value time sequence;
sequentially carrying out centralization processing and absolute value taking processing on the characteristic values in each characteristic value time sequence according to the corresponding first average value to obtain preprocessed characteristic values;
determining a threshold value according to the first standard deviation and a preset hyper-parameter, and performing binarization processing on the characteristic values in each characteristic value time sequence according to the corresponding threshold value to obtain a feature value after binarization processing;
for each single battery, adding all the preprocessed characteristic values corresponding to the single battery to obtain first characteristic data of the single battery; adding all the feature values after binarization processing corresponding to the battery monomer to obtain second feature data of the battery monomer;
and combining the first characteristic data with the corresponding second characteristic data to obtain the difference characteristic data of each battery cell.
Optionally, the sequentially performing centering processing and absolute value processing on the feature values in each feature value time sequence according to the corresponding first mean value to obtain the preprocessed feature values includes:
for each acquisition time, subtracting the first average value corresponding to the acquisition time from the characteristic value corresponding to the acquisition time in each characteristic value time sequence to obtain a characteristic value after centralization processing;
and taking an absolute value of each feature value after the centralization treatment to obtain the feature value after the pretreatment.
Optionally, the binarizing the feature values in each feature value time series according to the corresponding threshold, and obtaining the binarized feature values includes:
for each acquisition time, comparing the characteristic value corresponding to the acquisition time in each characteristic value time sequence with the threshold value corresponding to the acquisition time;
setting the characteristic value greater than or equal to the threshold value to 1 and the characteristic value less than the threshold value to 0 according to the comparison result.
Optionally, for each battery cell, after the combining all the difference characteristic data corresponding to the battery cell to obtain the difference characteristic combined data of the battery cell, the method further includes:
and for each single battery, carrying out standardization processing on the difference characteristic combination data of the single battery to obtain the difference characteristic combination data after the standardization processing.
Optionally, for each battery cell, the normalizing the difference feature combination data of the battery cell, and obtaining the normalized difference feature combination data includes:
for each battery cell, determining a second mean value and a second standard deviation of all the difference characteristic data in the difference characteristic combination data of the battery cell;
subtracting the corresponding second average value from each difference characteristic data to obtain processed difference characteristic data;
dividing each processed difference feature data by the corresponding second standard deviation to obtain the normalized difference feature combined data.
Optionally, the data characteristics include a voltage and a temperature of the battery cell, and the anomaly detection algorithm includes at least one of an OPTICS clustering algorithm, an isolated forest algorithm, and a DBSCAN algorithm.
In a second aspect, the present invention provides a battery pack failure detection apparatus, comprising:
the battery pack management system comprises an acquisition module, a storage module and a management module, wherein the acquisition module is used for acquiring a characteristic data combination of each battery monomer in a battery pack, and the characteristic data combination comprises a plurality of data characteristics and a characteristic value time sequence corresponding to each data characteristic;
the extraction module is used for comparing the characteristic value time sequence corresponding to the data characteristic in each characteristic data combination for each data characteristic, and extracting the difference characteristic data of each battery monomer between the data characteristic and other battery monomers; for each battery monomer, combining all the difference characteristic data corresponding to the battery monomer to obtain difference characteristic combination data of the battery monomer;
and the detection module is used for judging whether the corresponding battery monomer has a fault or not according to the difference characteristic combination data based on an abnormal detection algorithm.
In a third aspect, the present invention provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the battery pack failure detection method according to any one of the first aspect.
In a fourth aspect, the invention provides a vehicle comprising a memory and a processor;
the memory for storing a computer program;
the processor is configured to implement the battery pack fault detection method according to any one of the first aspect when the computer program is executed.
The battery pack fault detection method, the battery pack fault detection device and the vehicle have the advantages that: for each battery cell, a characteristic value time series of the battery cell under different data characteristics may be obtained, and the characteristic value time series may include a plurality of characteristic values at different acquisition times. For a battery cell, comparing the characteristic value time sequence with the characteristic time sequence of other battery cells under the same data characteristic, for example: for each data characteristic, comparing the characteristic time sequence of one battery monomer with the characteristic time sequences of other battery monomers, and extracting the difference characteristic data between the battery monomer and the other battery monomers through comparison, wherein the difference characteristic data reflects the difference between the battery monomer and the other battery monomers, and is favorable for realizing the fault detection of the battery monomer level. And combining all the difference characteristic data corresponding to the battery monomer, namely combining the difference characteristic data under each data characteristic corresponding to the battery monomer to obtain the difference characteristic combined data of the battery monomer, wherein the difference characteristic combined data reflects the difference between the battery monomer and other battery monomers from the dimensionality of a plurality of data characteristics, and the comprehensiveness of the extracted difference characteristic data is improved. And processing the difference characteristic combination data by adopting an anomaly detection algorithm to judge whether the corresponding battery monomer has a fault or not, so that the fault detection of the battery monomer level is realized, the difference characteristic combination data is composed of the difference characteristic data of the dimensionalities of a plurality of data characteristics, the accuracy of the fault detection is improved, and the battery monomer with the fault risk in the battery pack can be accurately detected.
Drawings
Fig. 1 is a schematic flow chart of a battery pack fault detection method according to an embodiment of the present invention;
fig. 2 is a detailed flowchart of step S200 of the battery pack fault detection method according to the embodiment of the present invention;
fig. 3 is a schematic structural diagram of a battery pack fault detection apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below. While certain embodiments of the present invention are shown in the drawings, it should be understood that the present invention may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but rather are provided for a more complete and thorough understanding of the present invention. It should be understood that the drawings and the embodiments of the present invention are illustrative only and are not intended to limit the scope of the present invention.
It should be understood that the various steps recited in the method embodiments of the present invention may be performed in a different order and/or performed in parallel. Moreover, method embodiments may include additional steps and/or omit performing the illustrated steps. The scope of the invention is not limited in this respect.
The term "include" and variations thereof as used herein are open-ended, i.e., "including but not limited to". The term "based on" is "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; the term "optionally" means "alternative embodiments". Relevant definitions for other terms will be given in the following description. It should be noted that the terms "first", "second", and the like in the present invention are only used for distinguishing different devices, modules or units, and are not used for limiting the order or interdependence relationship of the functions performed by the devices, modules or units.
It is noted that references to "a" or "an" or "the" modification(s) in the present invention are intended to be illustrative rather than limiting and that those skilled in the art will understand that reference to "one or more" unless the context clearly indicates otherwise.
The names of messages or information exchanged between devices in the embodiments of the present invention are for illustrative purposes only, and are not intended to limit the scope of the messages or information.
As shown in fig. 1, the method for detecting a failure of a battery pack provided by the present invention includes:
step S100, acquiring a characteristic data combination of each battery in the battery pack, wherein the characteristic data combination comprises a plurality of data characteristics and a characteristic value time sequence corresponding to each data characteristic.
Specifically, the data characteristics include the temperature and the voltage of the battery cells, the characteristic value time sequence includes characteristic values of the data characteristics at a plurality of different acquisition moments, and numerical values of the data characteristics such as the temperature and the voltage of each battery cell can be acquired at fixed time steps.
For any single battery, arranging all collected voltage values of the single battery according to the time sequence of the collection time to form the voltage value time sequence of the single battery; arranging all the acquired temperature values of the single battery according to the time sequence of the acquisition time, and forming the temperature value time sequence of the single battery.
Step S200, for each data characteristic, comparing the characteristic value time series corresponding to the data characteristic in each characteristic data combination, and extracting the difference characteristic data between each battery cell and other battery cells on the data characteristic.
For example, for any battery cell, the voltage value time series of the battery cell is compared with the voltage value time series of other battery cells, so that the difference characteristic data between the battery cell and other battery cells in terms of voltage can be extracted. Similarly, the difference characteristic data of the battery cell and other data characteristics such as temperature and the like can also be extracted.
Step S300, for each battery cell, combining all the difference feature data corresponding to the battery cell to obtain difference feature combined data of the battery cell.
Specifically, for any battery cell, combining the difference characteristic data of the battery cell with the difference characteristic data of other battery cells in terms of voltage, the difference characteristic data of the battery cell with the difference characteristic data of other battery cells in terms of temperature and the like to obtain the difference characteristic combination data of the battery cell, wherein the difference characteristic combination data reflects the difference between the battery cell and other battery cells in multiple dimensions.
And S400, judging whether the corresponding battery monomer has a fault or not according to the difference characteristic combination data based on an abnormality detection algorithm.
Optionally, the anomaly detection algorithm comprises at least one of an OPTICS clustering algorithm, an isolated forest algorithm, and a DBSCAN algorithm.
In this embodiment, for each single battery, a characteristic value time series of the single battery under different data characteristics may be obtained, and the characteristic value time series may include a plurality of characteristic values at different acquisition times. For a battery cell, comparing the characteristic value time sequence with the characteristic time sequence of other battery cells under the same data characteristic, for example: for each data characteristic, comparing the characteristic time sequence of one battery monomer with the characteristic time sequences of other battery monomers, and extracting the difference characteristic data between the battery monomer and the other battery monomers through comparison, wherein the difference characteristic data reflects the difference between the battery monomer and the other battery monomers, and is favorable for realizing the fault detection of the battery monomer level. All the difference feature data corresponding to the battery single body are combined, namely the difference feature data under each data feature corresponding to the battery single body are combined to obtain the difference feature combined data of the battery single body, the difference feature combined data reflects the difference between the battery single body and other battery single bodies from the dimensionality of a plurality of data features, and the comprehensiveness of the extracted difference feature data is improved. And processing the difference characteristic combination data by adopting an anomaly detection algorithm to judge whether the corresponding battery monomer has a fault or not, so that the fault detection of the battery monomer level is realized, the difference characteristic combination data is composed of the difference characteristic data of the dimensionalities of a plurality of data characteristics, the accuracy of the fault detection is improved, and the battery monomer with the fault risk in the battery pack can be accurately detected.
Optionally, as shown in fig. 2, for each data feature, the comparing, in each feature data combination, the feature value time series corresponding to the data feature, and the extracting the difference feature data between each battery cell and the other battery cells in the data feature includes:
step S210, for each data feature, determining a first mean value and a first standard deviation of the feature value at the same acquisition time in the corresponding feature value time series.
Specifically, taking the data characteristic as the voltage as an example, a first mean value and a first standard deviation of the voltage at the same acquisition time in the characteristic value time sequence corresponding to each battery cell are determined.
In an optional embodiment, for each data feature, a data feature matrix may be constructed according to a feature value time sequence corresponding to the data feature in each feature data combination, where rows of the data feature matrix represent feature values of each battery cell at the same acquisition time, and columns represent feature values of each battery cell at different acquisition times, or columns of the data feature matrix represent feature values of each battery cell at the same acquisition time, and rows represent feature values of each battery cell at different acquisition times.
For example, assuming that the data characteristic is a voltage, the data characteristic matrix is a voltage matrix V, rows of the voltage matrix V represent voltage values of the battery cells at the same acquisition time, and columns represent voltage values of the battery cells at different acquisition times, for example, a first mean value mean and a first standard deviation std of the voltage values of each row of the voltage matrix V can be directly calculated.
Step S220, according to the corresponding first mean value, sequentially performing centralization processing and absolute value taking processing on the characteristic values in the characteristic value time sequences to obtain preprocessed characteristic values.
Specifically, each feature value is subjected to centering processing according to a first average value corresponding to the acquisition time to obtain a feature value after centering processing, and then an absolute value is taken from the feature value after centering processing to obtain a feature value after preprocessing. Through the centralization processing and the absolute value taking processing, the deviation between the average values of each battery cell and all the battery cells at the same acquisition time can be extracted, and the larger the characteristic value after the preprocessing is, the more the deviation of the characteristic value of the battery cell on the data characteristic is.
In the following example, each voltage value in the voltage matrix V is subjected to centering processing according to the first average mean of the row in which the voltage value is located, so as to obtain a centered voltage matrix V _ mean, and then each numerical value in the centered voltage matrix V _ mean is subjected to absolute value calculation, so as to obtain a voltage absolute deviation matrix V _ abs, where each numerical value in the voltage absolute deviation matrix V _ abs is a preprocessed characteristic value.
Step S230, determining a threshold according to the first standard deviation and a preset hyper-parameter, and performing binarization processing on the feature values in each feature value time sequence according to the corresponding threshold to obtain binarized feature values.
Specifically, the product of the first standard deviation and the preset hyper-parameter may be used as a threshold, and each feature value is binarized according to the threshold calculated by the first standard deviation corresponding to the acquisition time, so as to obtain a feature value after binarization processing.
As an example, the product of the first standard deviation std of the voltage value of each row of the voltage matrix V and the preset super parameter Z _ score may be used as the threshold M of the voltage value of the row, where the preset super parameter Z _ score may be set to 3, and each voltage value in the voltage matrix V is binarized according to the threshold M of the row, so as to obtain the voltage super-threshold matrix V _ margin, where each value in the voltage super-threshold matrix V _ margin is the feature value after the binarization processing.
Step S240, adding all the preprocessed characteristic values corresponding to each single battery cell to obtain first characteristic data of the single battery cell; and adding all the feature values after the binarization processing corresponding to the single battery to obtain second feature data of the single battery.
Specifically, for any single battery, the preprocessed characteristic values of the single battery at each acquisition time are added to obtain first characteristic data of the single battery, wherein the first characteristic data is an accumulated value of all deviations of the single battery exceeding the corresponding data mean value at each acquisition time; and adding all the feature values of the single battery after the binarization processing at each acquisition time to obtain second feature data of the single battery, wherein the second feature data is an accumulated value of the out-of-tolerance times of the single battery at each acquisition time.
In the above example, the data in the voltage absolute deviation matrix V _ abs are accumulated in rows to obtain a voltage deviation accumulated value matrix V _ acc, and each data in the voltage deviation accumulated value matrix V _ acc is the first feature data. And accumulating the data in the voltage exceeding threshold matrix V _ margin according to columns to obtain a voltage exceeding difference number accumulation matrix V _ num, wherein each data in the voltage exceeding difference number accumulation matrix V _ num is second characteristic data. Assuming that the voltage matrix V is an m × n matrix, the voltage deviation accumulated value matrix V _ acc and the voltage out-of-tolerance number accumulated matrix V _ num are 1 × n matrices.
Step S250, combining the first characteristic data with the corresponding second characteristic data to obtain the difference characteristic data of each battery cell.
Specifically, for any battery cell, the difference feature matrix of the battery cell can be obtained by combining the corresponding first feature data and the second feature data.
Continuing to the previous example, the voltage deviation cumulative value matrix V _ acc and the voltage out-of-tolerance times cumulative matrix V _ num are combined in rows to obtain a voltage characteristic matrix V _ fea, the voltage characteristic matrix V _ fea is a 2 × n matrix, and a row of data in the voltage characteristic matrix V _ fea is voltage difference characteristic data of a single battery.
The difference characteristic matrix of each battery cell, in which the data characteristics are temperature and the like, is calculated according to the same method as in the above example, and the same process is performed on the temperature matrix T, for example, a temperature characteristic matrix T _ fea can be obtained, where one column of data of the temperature characteristic matrix T _ fea represents temperature difference characteristic data of one battery cell.
In this optional embodiment, first the first mean value and the first standard deviation of all feature values of the same data feature acquired by each battery cell at the same time are calculated, and the feature values are subjected to centering processing and absolute value processing according to the first mean value corresponding to the acquisition time, so that the deviation between each feature value and the first mean value corresponding to the acquisition time can be determined, and the difference features between each battery cell and the mean values of all battery cells at different times are extracted. The first standard deviation may be multiplied by a preset hyperparameter to determine the threshold. And carrying out binarization processing on each characteristic value according to a threshold value corresponding to the acquisition time, wherein the characteristic value after binarization processing can judge whether each characteristic value is out of tolerance or not, and can be used for judging whether the corresponding monomer is abnormal or not. The preprocessed characteristic values at all the collection moments are added, and the binarized characteristic values at all the collection moments are added, so that the influence of data errors at a single collection moment on the accuracy of the whole abnormal characteristic can be reduced. And combining the first characteristic data and the second characteristic data, namely combining the difference characteristics respectively extracted from the plurality of data characteristics to obtain difference characteristic data reflecting the difference between the battery monomer and other battery monomers. By combining an anomaly detection algorithm and difference characteristic data of each battery cell, the fault detection of the battery cell level can be realized, and the difference characteristics among the battery cells are extracted from the dimensionality of a plurality of data characteristics, so that the accuracy of the extracted difference characteristics is improved, and the accuracy of the detection of the battery cells with fault risks can be improved.
Optionally, the sequentially performing centering processing and absolute value processing on the feature values in each feature value time sequence according to the corresponding first mean value to obtain the preprocessed feature values includes:
and for each acquisition time, respectively subtracting the first average value corresponding to the acquisition time from the characteristic value corresponding to the acquisition time in each characteristic value time sequence to obtain the characteristic value after the centralization processing.
And taking an absolute value of each feature value after the centralization treatment to obtain the feature value after the pretreatment.
In the following example, the first mean value mean of the row is subtracted from each voltage value in the voltage matrix V to obtain a centralized matrix, and the data in the centralized matrix is the feature value after the centralized processing. And taking an absolute value of each feature value after the centralization processing to obtain a voltage absolute deviation matrix V _ abs, wherein each data in the voltage absolute deviation matrix V _ abs is the preprocessed feature value.
In this optional embodiment, the preprocessed feature values reflect deviations between each single battery cell and the average values of all the corresponding single batteries at different acquisition times, wherein the larger the preprocessed feature values are, the larger the degree of deviation of the feature values from the average values is, the more possible there is an abnormality; the smaller the feature value after the preprocessing, the smaller the degree of deviation of the feature value from the average value, and the lower the possibility of abnormality. Through centralization processing and absolute value processing, a preprocessed characteristic value reflecting deviation from an average value is extracted, and the fault detection of the battery monomer level is facilitated.
And accumulating the preprocessed characteristic values at each acquisition time, wherein the obtained first characteristic data can reflect the condition that the single battery deviates from the corresponding average value at a plurality of acquisition times, and if the first characteristic data is larger, the first characteristic data indicates that the single battery possibly deviates from the corresponding average value for a long time, and the possibility of abnormality is higher. Therefore, the preprocessed characteristic values at each acquisition time are added, so that the influence caused by possible errors of the preprocessed characteristic values at a single acquisition time can be reduced, and the accuracy of the extracted characteristic is improved.
Optionally, the binarizing the feature values in each feature value time sequence according to the corresponding threshold, and obtaining the binarized feature values includes:
for each acquisition time, comparing the characteristic value corresponding to the acquisition time in each characteristic value time sequence with the threshold value corresponding to the acquisition time;
setting the characteristic value greater than or equal to the threshold value to 1 and the characteristic value less than the threshold value to 0 according to the comparison result.
In the following example, the voltage values in the voltage matrix V are compared with the threshold values M of the rows, respectively, the voltage value greater than or equal to the threshold value M is set to 1, and the voltage value smaller than the threshold value M is set to 0, thereby forming the voltage super-threshold matrix V _ margin.
In this optional embodiment, the distance of the characteristic value from the average can be determined by comparing the characteristic value with the multiple of the corresponding standard deviation, and the larger the difference between the characteristic value and the threshold value is, the farther the characteristic value is from the average value is, the lower the possibility of the characteristic value is, that is, the higher the possibility of the battery cell corresponding to the characteristic value being abnormal is. Setting the characteristic value greater than or equal to the threshold value to 1 through binarization processing, which indicates that the corresponding battery cell has high possibility of abnormality at the moment; setting the characteristic value smaller than the threshold value to 0 indicates that the corresponding battery cell is less likely to have an abnormality at that time. The extracted feature value after binarization can visually reflect the possibility of abnormity of the single battery, and is favorable for realizing the subsequent fault detection of the single battery level.
And subsequently accumulating the feature values after the binarization processing at each acquisition time, wherein the obtained second feature data reflects the times that the feature value is 1, and also reflects the times that the single battery cell is judged to have high abnormal possibility. Therefore, the greater the second characteristic data is, the greater the number of times the battery cell is determined to have a high possibility of abnormality, the greater the probability that the battery cell has abnormality. The feature values after the binarization processing at each acquisition time are added, so that the influence caused by possible errors of the feature values after the binarization processing at a single acquisition time can be reduced, and the accuracy of the extracted features is improved.
Optionally, after the combining all the difference characteristic data corresponding to the battery cell to obtain the difference characteristic combined data of the battery cell for each battery cell, the method further includes:
and for each battery cell, carrying out standardization processing on the difference characteristic combination data of the battery cell to obtain the difference characteristic combination data after standardization processing.
In the following example, the voltage characteristic matrix V _ fea and the temperature characteristic matrix T _ fea are combined in columns to form a 4 × n matrix, so as to obtain a characteristic matrix F, where each datum in the characteristic matrix F represents a difference characteristic datum, and each column datum represents a difference characteristic combination datum of one battery cell.
Specifically, the normalization process is to subtract the mean value of each difference feature data in the difference feature combination data and then divide the difference feature data by the standard deviation.
Optionally, for each battery cell, the normalizing the difference feature combination data of the battery cell, and obtaining the normalized difference feature combination data includes:
for each battery cell, determining a second mean value and a second standard deviation of all the difference characteristic data in the difference characteristic combination data of the battery cell.
Continuing with the above example, a second mean and a second standard deviation of each column of data of the feature matrix F are determined.
And subtracting the corresponding second average value from each difference characteristic data to obtain the processed difference characteristic data.
In the following example, the second mean value of the column is subtracted from each data in the feature matrix F to obtain a processed feature matrix, and each data in the processed feature matrix is the processed difference feature matrix.
Dividing each processed difference feature data by the corresponding second standard deviation to obtain the normalized difference feature combined data.
In the following example, each data in the processed feature matrix is divided by the second standard deviation of the column to obtain the final feature matrix F _ final, and each column of data in the final feature matrix F _ final represents the difference feature combination data after the normalization processing of one battery cell.
And based on an OPTIC clustering algorithm, carrying out clustering analysis according to the final feature matrix F _ final, and if a certain point in the final feature matrix F _ final is identified as a noise point, judging that the corresponding single battery has a fault.
The OPTIC clustering algorithm is a density-based clustering algorithm, is an improved version of DBSCAN, and has better generalization. The algorithm builds a reachability graph that assigns each sample a reachability distance and a point in a cluster ordering attribute; these two attributes are assigned when fitting the model for determining the membership of the cluster.
In this optional embodiment, multiple data features can be unified to the same evaluation dimension through standardization processing, so that subsequent data processing is facilitated, and fault detection is realized.
It can be understood that, in the above example, the rows of the data feature matrix represent the feature values of the battery cells at the same acquisition time, the columns represent the feature values of the battery cells at different acquisition times, the columns of the data feature matrix represent the feature values of the battery cells at the same acquisition time, the cases where the rows represent the feature values of the battery cells at different acquisition times are similar to the above example, and the rows and the columns of the processing process may be adjusted correspondingly, which is not described herein again.
In a second aspect, as shown in fig. 3, the present invention provides a battery pack failure detection apparatus, comprising:
the battery pack management system comprises an acquisition module, a storage module and a management module, wherein the acquisition module is used for acquiring a characteristic data combination of each battery monomer in a battery pack, and the characteristic data combination comprises a plurality of data characteristics and a characteristic value time sequence corresponding to each data characteristic;
the extraction module is used for comparing the characteristic value time sequence corresponding to the data characteristic in each characteristic data combination for each data characteristic, and extracting the difference characteristic data of each battery monomer on the data characteristic and other battery monomers; for each battery monomer, combining all the difference characteristic data corresponding to the battery monomer to obtain difference characteristic combination data of the battery monomer;
and the detection module is used for judging whether the corresponding battery monomer has a fault or not according to the difference characteristic combination data based on an abnormal detection algorithm.
The battery pack fault detection device can be deployed on a vehicle and also can be deployed at the cloud end and is used for carrying out online detection and early warning on the characteristic data combination of the battery monomers acquired in real time.
The battery pack fault detection device of the present invention is used for implementing the battery pack fault detection method, and the advantages thereof over the prior art are the same as the advantages of the battery pack fault detection method over the prior art, and are not described herein again.
Optionally, the time series of feature values includes feature values of the data feature at a plurality of different acquisition times; the data characteristics include a voltage and a temperature of the battery cell, and the abnormality detection algorithm includes at least one of an OPTICS clustering algorithm, an isolated forest algorithm, and a DBSCAN algorithm.
Optionally, the extracting module is specifically configured to:
for each data feature, determining a first mean value and a first standard deviation of the feature value at the same acquisition time in the corresponding feature value time sequence;
sequentially carrying out centralization processing and absolute value taking processing on the characteristic values in each characteristic value time sequence according to the corresponding first average value to obtain preprocessed characteristic values;
determining a threshold value according to the first standard deviation and a preset hyper-parameter, and performing binarization processing on the characteristic values in each characteristic value time sequence according to the corresponding threshold value to obtain a feature value after binarization processing;
for each battery monomer, adding all the preprocessed characteristic values corresponding to the battery monomer to obtain first characteristic data of the battery monomer; adding all the feature values after binarization processing corresponding to the battery monomer to obtain second feature data of the battery monomer;
and combining the first characteristic data with the corresponding second characteristic data to obtain the difference characteristic data of each battery cell.
Optionally, the extracting module is specifically configured to: for each acquisition time, respectively subtracting the first average value corresponding to the acquisition time from the characteristic value corresponding to the acquisition time in each characteristic value time sequence to obtain a characteristic value after centralized processing; and taking an absolute value of each feature value after the centralization treatment to obtain the feature value after the pretreatment.
Optionally, the extracting module is specifically configured to: for each acquisition time, comparing the characteristic value corresponding to the acquisition time in each characteristic value time sequence with the threshold value corresponding to the acquisition time; setting the characteristic value greater than or equal to the threshold value to 1 and the characteristic value less than the threshold value to 0 according to the comparison result.
Optionally, the battery pack fault detection apparatus further includes a standardization processing module, and the standardization processing module is configured to: and for each battery cell, carrying out standardization processing on the difference characteristic combination data of the battery cell to obtain the difference characteristic combination data after standardization processing.
Optionally, the normalization processing module is specifically configured to: for each battery cell, determining a second mean value and a second standard deviation of all the difference characteristic data in the difference characteristic combination data of the battery cell; subtracting the corresponding second average value from each difference characteristic data to obtain processed difference characteristic data; dividing each processed difference feature data by the corresponding second standard deviation to obtain the normalized difference feature combined data.
In a third aspect, the present invention provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the battery pack failure detection method according to any one of the first aspect.
In a fourth aspect, the invention provides a vehicle comprising a memory and a processor;
the memory for storing a computer program;
the processor is configured to implement the battery pack fault detection method according to any one of the first aspect when the computer program is executed.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like. In this application, the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment of the present invention. In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
Although the present disclosure has been described above, the scope of the present disclosure is not limited thereto. Various changes and modifications may be effected therein by one of ordinary skill in the pertinent art without departing from the spirit and scope of the present disclosure, and these changes and modifications are intended to be within the scope of the present disclosure.
Claims (10)
1. A battery pack fault detection method, comprising:
acquiring a characteristic data combination of each battery in a battery pack, wherein the characteristic data combination comprises a plurality of data characteristics and a characteristic value time sequence corresponding to each data characteristic;
for each data feature, comparing the feature value time sequence corresponding to the data feature in each feature data combination, and extracting the difference feature data between each battery cell and other battery cells on the data feature;
for each battery monomer, combining all the difference characteristic data corresponding to the battery monomer to obtain difference characteristic combination data of the battery monomer;
and judging whether the corresponding battery monomer has a fault or not according to the difference characteristic combination data based on an abnormal detection algorithm.
2. The battery pack fault detection method according to claim 1, wherein the characteristic value time series includes characteristic values of the data characteristic at a plurality of different acquisition times;
for each data feature, comparing the feature value time series corresponding to the data feature in each feature data combination, and extracting the difference feature data between each battery cell and other battery cells on the data feature comprises:
for each data feature, determining a first mean value and a first standard deviation of the feature value at the same acquisition time in the corresponding feature value time sequence;
sequentially carrying out centralization processing and absolute value taking processing on the characteristic values in each characteristic value time sequence according to the corresponding first average value to obtain preprocessed characteristic values;
determining a threshold value according to the first standard deviation and a preset hyper-parameter, and performing binarization processing on the characteristic values in each characteristic value time sequence according to the corresponding threshold value to obtain a feature value after binarization processing;
for each battery monomer, adding all the preprocessed characteristic values corresponding to the battery monomer to obtain first characteristic data of the battery monomer; adding all the feature values after the binarization processing corresponding to the single battery to obtain second feature data of the single battery;
and combining the first characteristic data with the corresponding second characteristic data to obtain the difference characteristic data of each battery cell.
3. The battery pack fault detection method according to claim 2, wherein the sequentially performing centering processing and absolute value taking processing on the characteristic values in each characteristic value time sequence according to the corresponding first mean value to obtain a preprocessed characteristic value comprises:
for each acquisition time, respectively subtracting the first average value corresponding to the acquisition time from the characteristic value corresponding to the acquisition time in each characteristic value time sequence to obtain a characteristic value after centralized processing;
and taking an absolute value of each feature value after the centralization treatment to obtain the feature value after the pretreatment.
4. The battery pack fault detection method according to claim 2, wherein the binarizing the feature values in each feature value time series according to the corresponding threshold value, and obtaining binarized feature values comprises:
for each acquisition time, comparing the characteristic value corresponding to the acquisition time in each characteristic value time sequence with the threshold value corresponding to the acquisition time;
setting the characteristic value greater than or equal to the threshold value to 1 and the characteristic value less than the threshold value to 0 according to the comparison result.
5. The battery pack fault detection method according to any one of claims 1 to 4, wherein, for each battery cell, after all the difference characteristic data corresponding to the battery cell are combined to obtain the difference characteristic combined data of the battery cell, the method further comprises:
and for each battery cell, carrying out standardization processing on the difference characteristic combination data of the battery cell to obtain the difference characteristic combination data after standardization processing.
6. The battery pack fault detection method according to claim 5, wherein the step of normalizing the difference feature combination data of each battery cell to obtain normalized difference feature combination data comprises the steps of:
for each battery cell, determining a second mean value and a second standard deviation of all the difference characteristic data in the difference characteristic combination data of the battery cell;
subtracting the corresponding second average value from each difference characteristic data to obtain processed difference characteristic data;
dividing each processed difference feature data by the corresponding second standard deviation to obtain the normalized difference feature combined data.
7. The battery pack fault detection method according to any one of claims 1 to 4, wherein the data characteristics include a voltage and a temperature of the battery cell, and the abnormality detection algorithm includes at least one of an OPTIC clustering algorithm, an outlier forest algorithm, and a DBSCAN algorithm.
8. A battery pack failure detection device, comprising:
the battery pack management system comprises an acquisition module, a storage module and a management module, wherein the acquisition module is used for acquiring a characteristic data combination of each battery monomer in a battery pack, and the characteristic data combination comprises a plurality of data characteristics and a characteristic value time sequence corresponding to each data characteristic;
the extraction module is used for comparing the characteristic value time sequence corresponding to the data characteristic in each characteristic data combination for each data characteristic, and extracting the difference characteristic data of each battery monomer on the data characteristic and other battery monomers; for each battery monomer, combining all the difference characteristic data corresponding to the battery monomer to obtain difference characteristic combination data of the battery monomer;
and the detection module is used for judging whether the corresponding battery monomer has a fault or not according to the difference characteristic combination data based on an abnormal detection algorithm.
9. A computer-readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when being executed by a processor, carries out a battery pack failure detection method according to any one of claims 1 to 7.
10. A vehicle comprising a memory and a processor;
the memory for storing a computer program;
the processor, when executing the computer program, for implementing the battery pack failure detection method according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211159286.9A CN115629323A (en) | 2022-09-22 | 2022-09-22 | Battery pack fault detection method and device and vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211159286.9A CN115629323A (en) | 2022-09-22 | 2022-09-22 | Battery pack fault detection method and device and vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115629323A true CN115629323A (en) | 2023-01-20 |
Family
ID=84903611
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211159286.9A Pending CN115629323A (en) | 2022-09-22 | 2022-09-22 | Battery pack fault detection method and device and vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115629323A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115877222A (en) * | 2023-02-14 | 2023-03-31 | 国网浙江省电力有限公司宁波供电公司 | Energy storage power station fault detection method and device, medium and energy storage power station |
-
2022
- 2022-09-22 CN CN202211159286.9A patent/CN115629323A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115877222A (en) * | 2023-02-14 | 2023-03-31 | 国网浙江省电力有限公司宁波供电公司 | Energy storage power station fault detection method and device, medium and energy storage power station |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112092675B (en) | Battery thermal runaway early warning method, system and server | |
| CN102245437B (en) | Vehicle failure diagnostic device | |
| CN111208445A (en) | Power battery abnormal monomer identification method and system | |
| CN113760670A (en) | Cable joint abnormity early warning method and device, electronic equipment and storage medium | |
| CN107942994A (en) | A kind of satellite temperature control system method for diagnosing faults based on temperature curve feature | |
| CN115865649B (en) | Intelligent operation and maintenance management control method, system and storage medium | |
| CN115629323A (en) | Battery pack fault detection method and device and vehicle | |
| CN116167010B (en) | Rapid identification method for abnormal events of power system with intelligent transfer learning capability | |
| CN116678552B (en) | Abnormality monitoring method for optical fiber stress sensor in variable temperature environment | |
| CN110715678A (en) | Sensor abnormity detection method and device | |
| CN115877222A (en) | Energy storage power station fault detection method and device, medium and energy storage power station | |
| CN116593897A (en) | Power battery fault diagnosis method, system, vehicle and storage medium | |
| CN115327417A (en) | Early warning method and system for abnormity of power battery monomer and electronic equipment | |
| CN117048524A (en) | Method and device for detecting vehicle faults, vehicle and storage medium | |
| CN110703013B (en) | Online identification method and device for low-frequency oscillation mode of power system and electronic equipment | |
| CN117031294A (en) | Battery multi-fault detection method, device and storage medium | |
| CN111159251A (en) | Method and device for determining abnormal data | |
| CN115980585A (en) | Battery fault detection method and device, computer equipment and storage medium | |
| CN115345241A (en) | Battery abnormality detection method, battery abnormality detection device, electronic apparatus, and computer storage medium | |
| CN112782589A (en) | Vehicle-mounted fuel cell remote fault classification diagnosis method and device and storage medium | |
| CN117648612B (en) | Parallel battery pack fault detection method, device, electronic equipment and storage medium | |
| CN110751747A (en) | Data processing method and device | |
| CN116736173B (en) | Energy storage battery model construction and energy storage battery state judgment method and device | |
| CN117572837B (en) | Intelligent power plant AI active operation and maintenance method and system | |
| CN111632384B (en) | Game online number detection method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20231108 Address after: 310051 No. 1760, Jiangling Road, Hangzhou, Zhejiang, Binjiang District Applicant after: ZHEJIANG GEELY HOLDING GROUP Co.,Ltd. Applicant after: NINGBO GEELY ROYAL ENGINE COMPONENTS Co.,Ltd. Address before: 310051 No. 1760, Jiangling Road, Hangzhou, Zhejiang, Binjiang District Applicant before: ZHEJIANG GEELY HOLDING GROUP Co.,Ltd. Applicant before: ZHEJIANG GEELY POWER ASSEMBLY Co.,Ltd. |
|
| TA01 | Transfer of patent application right |