CN109765490A - A kind of power battery fault detection method and system based on high dimensional data diagnosis - Google Patents
A kind of power battery fault detection method and system based on high dimensional data diagnosis Download PDFInfo
- Publication number
- CN109765490A CN109765490A CN201811336819.XA CN201811336819A CN109765490A CN 109765490 A CN109765490 A CN 109765490A CN 201811336819 A CN201811336819 A CN 201811336819A CN 109765490 A CN109765490 A CN 109765490A
- Authority
- CN
- China
- Prior art keywords
- point
- voltage data
- matrix
- dot matrix
- threshold
- 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.)
- Granted
Links
Abstract
The present invention discloses a kind of power battery fault detection method and system based on high dimensional data diagnosis, it is pre-processed by the voltage data to battery cell, obtain voltage data matrix, then clustering processing is carried out to voltage data matrix, obtain preliminary abnormal dot matrix and core dot matrix, the application is only needed to preliminary abnormal dot matrix using the rejecting outliers method based on angle, the failure of power battery can be diagnosed, compared to the workload that all raw data sets of detection greatly reduce data processing, reduce the complexity of data processing, and then improve detection efficiency.
Description
Technical field
The present invention relates to battery detecting fields, examine more particularly to a kind of power battery failure based on high dimensional data diagnosis
Survey method and system.
Background technique
Electrokinetic cell system is the most key component of electric vehicle, and multiple battery cells are usually serially connected in battery
Group uses.Although battery cell is grouped after tested and preferably, but still there are performance differences, these differences are in the long-term of battery
New difference can be generated in operational process to some extent, Individual cells monomer performance is decreased obviously, and seriously affects battery pack
Can, or even cause the accident, it is therefore desirable to the decline of battery in battery pack monomer performance and initial failure are detected, so as to
Disposal Measures are taken in time to it.
Traditional method detected based on angle variance to high dimensional data abnormality detection point, is needed to all initial data
Collection is calculated, and complexity is high, and it is long to calculate the time.
Summary of the invention
The object of the present invention is to provide a kind of power battery fault detection methods and system based on high dimensional data diagnosis, reach
To the effect for improving power battery fault detection efficiency.
To achieve the above object, the present invention provides following schemes:
A kind of power battery fault detection method based on high dimensional data diagnosis, which comprises
The voltage data of battery cell is pre-processed, voltage data matrix is obtained;
Clustering processing is carried out to the voltage data matrix using clustering algorithm, obtains cluster result, the cluster result
Including preliminary abnormal the dot matrix T, the core dot matrix Q being made of core point being made of boundary point and abnormal point;
According to the cluster result, exceptional value score is calculated using rejecting outliers algorithm;
Compare the size of the exceptional value score and threshold value, if the exceptional value score be less than the threshold value, determine described in
The corresponding battery cell of exceptional value score breaks down, and exports the number of the battery cell to break down.
Optionally, the voltage data matrix to battery cell pre-processes, and obtains voltage data matrix, specific to wrap
It includes:
It deletes and repeats the data at record moment and the data of misregistration in the voltage data of the battery cell, obtain clear
Wash rear battery cell voltage data;
Utilize formula Δ Ui=Ui+1-UiFirst-order difference processing is carried out to battery cell voltage data after the cleaning, is obtained
The voltage data matrix;
Wherein, UiIndicate the voltage data of the battery cell of i-th acquisition, Δ UiIndicate the institute acquired twice in succession
The difference of the voltage data of battery cell is stated, i indicates the number of the voltage data acquisition of the battery cell.
Optionally, described that clustering processing is carried out to the voltage data matrix using clustering algorithm, cluster result is obtained, is had
Body includes:
Initialization area threshold value and quantity threshold;
Determine that the number of the point in the voltage data matrix described in the region threshold is greater than the area of the quantity threshold
Domain is nucleus, and the core dot matrix Q is made of the point of the nucleus;
Determine that the number of the point in voltage data matrix described in the region threshold and outside the nucleus is less than
The region of the quantity threshold is borderline region, and the point of the borderline region is the boundary point;
The point for determining the voltage data matrix described other than the region threshold is the abnormal point;The boundary point and institute
It states abnormal point and constitutes the preliminary abnormal dot matrix T;Determine that the core dot matrix Q and the preliminary exception dot matrix T are poly-
Class result;
Judge whether the cluster result reaches cluster target, obtains judging result;
When judging result expression is, the cluster result is exported;
When the judging result indicates no, the region threshold and the quantity threshold are adjusted, and is back to step " really
It is core space that the number for being scheduled on the point in the region threshold in the voltage data matrix, which is greater than the region of the quantity threshold,
Domain is made of the core dot matrix Q " the point of the nucleus.
Optionally, the cluster target is that the number of the preliminary abnormal point accounts for the number of the point in the voltage data matrix
Purpose 10%~20%.
Optionally, described according to the cluster result, exceptional value score is calculated using rejecting outliers algorithm, it is specific to wrap
It includes:
The preliminary exception dot matrix T and core dot matrix Q is mapped into higher dimensional space;
Calculate the angle that any two point is formed in the point and the core dot matrix Q of the preliminary abnormal dot matrix T;
According to the angle calcu-lation included angle cosine value, included angle cosine value set is obtained;
The variance for calculating the cosine value set determines that the variance is the exceptional value score of the point of the sample data P.
A kind of power battery fault detection system based on high dimensional data diagnosis, the system comprises
Pretreatment unit is pre-processed for the voltage data to battery cell, obtains voltage data matrix;
Clustering processing unit is clustered for carrying out clustering processing to the voltage data matrix using clustering algorithm
As a result, the cluster result includes the preliminary abnormal dot matrix T being made of boundary point and abnormal point, the core being made of core point
Dot matrix Q;
Exceptional value score calculating unit, for calculating exceptional value using rejecting outliers algorithm according to the cluster result
Score;
Fail battery determination unit, for the size of the exceptional value score and threshold value, if the exceptional value score
It less than the threshold value, determines that the corresponding battery cell of the exceptional value score breaks down, and exports the battery list to break down
The number of body.
Optionally, the pretreatment unit specifically includes:
Data cleansing subelement repeats the data and note that record the moment in the voltage data for deleting the battery cell
Record the data of mistake, battery cell voltage data after being cleaned;
First-order difference subelement, for utilizing formula Δ Ui=Ui+1-UiTo battery cell voltage data after the cleaning into
The processing of row first-order difference, obtains the voltage data matrix;
Wherein, UiIndicate the voltage data of the battery cell of i-th acquisition, Δ UiIndicate the institute acquired twice in succession
The difference of the voltage data of battery cell is stated, i indicates the number of the voltage data acquisition of the battery cell.
Optionally, the clustering processing unit specifically includes:
Threshold value initializes subelement, is used for initialization area threshold value and quantity threshold;
Nucleus determines subelement, for determining the number of the point in the voltage data matrix described in the region threshold
The region that mesh is greater than the quantity threshold is nucleus, and the core dot matrix Q is made of the point of the nucleus;
Boundary Region area determines subelement, for determining voltage number described in the region threshold and outside the nucleus
It is Boundary Region area according to the region that the number of the point in matrix is less than the quantity threshold, the point of the borderline region is the boundary
Point;
Abnormal point determines subelement, for determining that the point of the voltage data matrix described other than the region threshold is described
Abnormal point;The boundary point and the abnormal point constitute the preliminary abnormal dot matrix T;Determine the core dot matrix Q and institute
Stating preliminary exception dot matrix T is cluster result;
Cluster result judgment sub-unit obtains judging result for judging whether the cluster result reaches cluster target;
Cluster result exports subelement, for exporting the cluster result when judging result expression is;
Adjusting thresholds subelement, for adjusting the region threshold and the number when the judging result indicates no
Threshold value, and be back to step and " determine that the number of the point in the voltage data matrix described in the region threshold is greater than the number
The region of mesh threshold value is nucleus, and the core dot matrix Q " is made of the point of the nucleus.
Optionally, the cluster target is that the number of the preliminary abnormal point accounts for the number of the point in the voltage data matrix
Purpose 10%~20%.
Optionally, the exceptional value score calculating unit specifically includes:
Subelement is mapped, for the preliminary exception dot matrix T and core dot matrix Q to be mapped to higher dimensional space;
Angle calcu-lation subelement, it is any in the point and the core dot matrix Q for calculating the preliminary abnormal dot matrix T
The angle that two points are formed;
Included angle cosine value computation subunit, for obtaining included angle cosine value collection according to the angle calcu-lation included angle cosine value
It closes;
Exceptional value score determines subelement, for calculating the variance of the cosine value set, determines that the variance is described
The exceptional value score of the point of sample data P.
The specific embodiment provided according to the present invention, the invention discloses following technical effects: the present invention passes through to battery
The voltage data of monomer is pre-processed, and voltage data matrix is obtained, and is then carried out clustering processing to voltage data matrix, is obtained
Preliminary exception dot matrix and core dot matrix, the application are only needed to preliminary abnormal data matrix using the exceptional value based on angle
Detection method, it will be able to which the failure for diagnosing power battery greatly reduces data processing compared to all raw data sets are detected
Workload, reduce the complexity of data processing, and then improve detection efficiency.
Detailed description of the invention
It in order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, below will be to institute in embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention
Example, for those of ordinary skill in the art, without any creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is the process for the power battery fault detection method based on high dimensional data diagnosis that the embodiment of the present invention provides
Figure;
Fig. 2 is the structure for the power battery fault detection system based on high dimensional data diagnosis that the embodiment of the present invention provides
Block diagram;
Fig. 3 is the schematic diagram after the cluster result that the embodiment of the present invention provides is visualized.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The object of the present invention is to provide a kind of power battery fault detection method and system based on high dimensional data diagnosis, drops
The complexity of low data processing, and then improve detection efficiency.
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, with reference to the accompanying drawing and specific real
Applying mode, the present invention is described in further detail.
As shown in Figure 1, the power battery fault detection method provided in this embodiment based on high dimensional data diagnosis includes:
Step 101: the voltage data of battery cell being pre-processed, voltage data matrix is obtained.
The step 101 specifically includes:
It deletes and repeats the data at record moment and the data of misregistration in the voltage data of the battery cell, obtain clear
Wash rear battery cell voltage data;
Utilize formula Δ Ui=Ui+1-UiFirst-order difference processing is carried out to battery cell voltage data after the cleaning, is obtained
The voltage data matrix;
Wherein, UiIndicate the voltage data of the battery cell of i-th acquisition, Δ UiIndicate the institute acquired twice in succession
The difference of the voltage data of battery cell is stated, i indicates the number of the voltage data acquisition of the battery cell.
It is extracted in battery cell voltage data from platform, deletes the data and prison electric car BMS for repeating the record moment
System operates normally, and is deleted according to monomer discharge voltage lower threshold and charging upper limit threshold value error logging value, not only
The treating capacity of voltage monomer data can be reduced, additionally it is possible to improve the accuracy of detection;And first-order difference side is used to data
Method processing, eliminates linear trend factor, obtains stationary sequence.
Step 102: clustering processing being carried out to the voltage data matrix using clustering algorithm, obtains cluster result;It is described
Cluster result includes the preliminary abnormal dot matrix T, the core dot matrix Q for having core point to constitute being made of boundary point and abnormal point.
The step 102 specifically includes:
Initialization area threshold value and quantity threshold;
Determine that the number of the point in the voltage data matrix described in the region threshold is greater than the area of the quantity threshold
Domain is nucleus, and the core dot matrix Q is made of the point of the nucleus;
Determine that the number of the point in voltage data matrix described in the region threshold and outside the nucleus is less than
The region of the quantity threshold is borderline region, and the point of the borderline region is the boundary point;
The point for determining the voltage data matrix described other than the region threshold is the abnormal point;The boundary point and institute
It states abnormal point and constitutes the preliminary abnormal dot matrix T;Determine that the core dot matrix Q and the preliminary exception dot matrix T are poly-
Class result;
Judge whether the cluster result reaches cluster target, obtains judging result;The cluster target is described preliminary
The number of abnormal point accounts for the 10%~20% of the number of the point in the voltage data matrix;
When judging result expression is, the cluster result is exported;
When the judging result indicates no, the region threshold and the quantity threshold are adjusted, and is back to step " really
It is core space that the number for being scheduled on the point in the region threshold in the voltage data matrix, which is greater than the region of the quantity threshold,
Domain is made of the core dot matrix Q " the point of the nucleus.
The preliminary abnormal data set number obtained after clustered processing is reduced compared to raw data set data sample sum
60%-70%, greatly reduces data processing amount, improves detection efficiency.
Can also include: after step 102
Dimension-reduction treatment is carried out to the cluster result;
Visualize the cluster result.
Dimension-reduction treatment is carried out to the cluster result and is visualized, it is pre- can intuitively to find out whether the result of cluster meets
Phase requires, convenient for being adjusted to threshold value.
As shown in figure 3, the large circle point in figure indicates that the core point, dot indicate the abnormal point and the boundary
Point.
Step 103: according to the cluster result, calculating exceptional value score using rejecting outliers algorithm.
The step 103 specifically includes:
The preliminary exception dot matrix T and core dot matrix Q is mapped into higher dimensional space;
Calculate the angle that any two point is formed in the point and the core dot matrix Q of the preliminary abnormal dot matrix T;
According to the angle calcu-lation included angle cosine value, included angle cosine value set is obtained;
The variance for calculating the cosine value set determines that the variance is the exceptional value score of the point of the sample data P.
Exceptional value score of the preliminary abnormal point relative to core data point is calculated, for measuring the degree that peels off of data point,
The value the low, shows that the degree that peels off of the point is bigger, can primarily determine abnormal point.
Step 104: the size of the exceptional value score and threshold value, if the exceptional value score is less than the threshold value,
It determines that the corresponding battery cell of the exceptional value score breaks down, and exports the number of the battery cell to break down.
According to exceptional value score compared with the size of threshold value, it can determine that the battery cell to break down is exceptional value score
Battery cell corresponding less than threshold value, compared to by all data point calculation scores and threshold value comparison, computational complexity and fortune
Evaluation time is decreased obviously, and improves the efficiency of fault detection.
As shown in Fig. 2, the power battery fault detection system provided in this embodiment based on high dimensional data diagnosis includes: pre-
Processing unit 201, clustering processing unit 202, exceptional value score calculating unit 203 and fail battery determination unit 204.
Pretreatment unit 201 is pre-processed for the voltage data to battery cell, obtains voltage data matrix.
The pretreatment unit 201 specifically includes:
Data cleansing subelement repeats the data and note that record the moment in the voltage data for deleting the battery cell
Record the data of mistake, battery cell voltage data after being cleaned;
First-order difference subelement, for utilizing formula Δ Ui=Ui+1-UiTo battery cell voltage data after the cleaning into
The processing of row first-order difference, obtains the voltage data matrix;
Wherein, UiIndicate the voltage data of the battery cell of i-th acquisition, Δ UiIndicate the institute acquired twice in succession
The difference of the voltage data of battery cell is stated, i indicates the number of the voltage data acquisition of the battery cell.
It is extracted in battery cell voltage data from platform, deletes the data and prison electric car BMS for repeating the record moment
System operates normally, and is deleted according to monomer discharge voltage lower threshold and charging upper limit threshold value error logging value, not only
The treating capacity of voltage monomer data can be reduced, additionally it is possible to improve the accuracy of detection;And first-order difference side is used to data
Method processing, eliminates linear trend factor, obtains stationary sequence.
Clustering processing unit 202 is gathered for carrying out clustering processing to the voltage data matrix using clustering algorithm
Class result.
The clustering processing unit 202 specifically includes:
Threshold value initializes subelement, is used for initialization area threshold value and quantity threshold;
Nucleus determines subelement, for determining the number of the point in the voltage data matrix described in the region threshold
The region that mesh is greater than the quantity threshold is nucleus, and the core dot matrix Q is made of the point of the nucleus;
Boundary Region area determines subelement, for determining voltage number described in the region threshold and outside the nucleus
It is borderline region according to the region that the number of the point in matrix is less than the quantity threshold, the point of the borderline region is the boundary
Point;
Abnormal point determines subelement, for determining that the point of the voltage data matrix described other than the region threshold is described
Abnormal point;The boundary point and the abnormal point constitute the preliminary abnormal dot matrix T;Determine the core dot matrix Q and institute
Stating preliminary exception dot matrix T is cluster result;
Cluster result judgment sub-unit obtains judging result for judging whether the cluster result reaches cluster target;
The cluster target is that the number of the preliminary abnormal point accounts for the 10%~20% of the number of the point in the voltage data matrix;
Cluster result exports subelement, for exporting the cluster result when judging result expression is;
Adjusting thresholds subelement, for adjusting the region threshold and the number when the judging result indicates no
Threshold value, and be back to step and " determine that the number of the point in the voltage data matrix described in the region threshold is greater than the number
The region of mesh threshold value is nucleus, and the core dot matrix Q " is made of the point of the nucleus.
The preliminary abnormal data set number obtained after clustered processing is reduced compared to raw data set data sample sum
60%-70%, greatly reduces data processing amount, improves detection efficiency.
It is described that clustering processing is carried out to the voltage data matrix using clustering algorithm, it obtains also wrapping after cluster result
It includes:
Dimension-reduction treatment unit, for carrying out dimension-reduction treatment to the cluster result;
Visualization, for visualizing the cluster result.
Dimension-reduction treatment is carried out to the cluster result and is visualized, it is pre- can intuitively to find out whether the result of cluster meets
Phase requires, convenient for being adjusted to threshold value.
As shown in figure 3, the large circle point in figure indicates that the core point, dot indicate the abnormal point and the boundary
Point.
Exceptional value score calculating unit 203, for being calculated using rejecting outliers algorithm abnormal according to the cluster result
It is worth score.
The exceptional value score calculating unit 203 specifically includes:
Subelement is mapped, for the preliminary exception dot matrix T and core dot matrix Q to be mapped to higher dimensional space;
Angle calcu-lation subelement, it is any in the point and the core dot matrix Q for calculating the preliminary abnormal dot matrix T
The angle that two points are formed;
Included angle cosine value computation subunit, for obtaining included angle cosine value collection according to the angle calcu-lation included angle cosine value
It closes;
Exceptional value score determines subelement, for calculating the variance of the cosine value set, determines that the variance is described
The exceptional value score of the point of sample data P.
Exceptional value score of the preliminary abnormal point relative to core data point is calculated, for measuring the degree that peels off of data point,
The value the low, shows that the degree that peels off of the point is bigger, can primarily determine abnormal point.
Fail battery determination unit 204, for the size of the exceptional value score and threshold value, if the exceptional value point
Number is less than the threshold value, determines that the corresponding battery cell of the exceptional value score breaks down, and exports the battery to break down
The number of monomer.
According to exceptional value score compared with the size of threshold value, it can determine that the battery cell to break down is exceptional value score
Battery cell corresponding less than threshold value, compared to by all data point calculation scores and threshold value comparison, computational complexity and fortune
Evaluation time is decreased obviously, and improves the efficiency of fault detection.
For the system disclosed in the embodiment, since it is corresponded to the methods disclosed in the examples, so the ratio of description
Relatively simple, reference may be made to the description of the method.
Used herein a specific example illustrates the principle and implementation of the invention, and above embodiments are said
It is bright to be merely used to help understand method and its core concept of the invention;At the same time, for those skilled in the art, foundation
Thought of the invention, there will be changes in the specific implementation manner and application range.In conclusion the content of the present specification is not
It is interpreted as limitation of the present invention.
Claims (10)
1. a kind of power battery fault detection method based on high dimensional data diagnosis, which is characterized in that the described method includes:
The voltage data of battery cell is pre-processed, voltage data matrix is obtained;
Clustering processing is carried out to the voltage data matrix using clustering algorithm, obtains cluster result, the cluster result includes
Preliminary abnormal the dot matrix T, the core dot matrix Q being made of core point being made of boundary point and abnormal point;
According to the cluster result, exceptional value score is calculated using rejecting outliers algorithm;
Compare the size of the exceptional value score and threshold value, if the exceptional value score is less than the threshold value, determines the exception
The corresponding battery cell of value score breaks down, and exports the number of the battery cell to break down.
2. the power battery fault detection method according to claim 1 based on high dimensional data diagnosis, which is characterized in that institute
It states and the voltage data matrix of battery cell is pre-processed, obtain voltage data matrix, specifically include:
It deletes and repeats the data at record moment and the data of misregistration in the voltage data of the battery cell, after obtaining cleaning
Battery cell voltage data;
Utilize formula Δ Ui=Ui+1-UiFirst-order difference processing is carried out to battery cell voltage data after the cleaning, is obtained described
Voltage data matrix;
Wherein, UiIndicate the voltage data of the battery cell of i-th acquisition, Δ UiIndicate the electricity acquired twice in succession
The difference of the voltage data of pond monomer, i indicate the number of the voltage data acquisition of the battery cell.
3. the power battery fault detection method according to claim 1 based on high dimensional data diagnosis, which is characterized in that institute
It states and clustering processing is carried out to the voltage data matrix using clustering algorithm, obtain cluster result, specifically include:
Initialization area threshold value and quantity threshold;
The region that the number of point in the determining voltage data matrix described in the region threshold is greater than the quantity threshold is
Nucleus is made of the core dot matrix Q the point of the nucleus;
It is described to determine that the number of the point in voltage data matrix described in the region threshold and outside the nucleus is less than
The region of quantity threshold is borderline region, and the point of the borderline region is the boundary point;
The point for determining the voltage data matrix described other than the region threshold is the abnormal point;The boundary point with it is described different
Often point constitutes the preliminary abnormal dot matrix T;Determine the core dot matrix Q and preliminary exception dot matrix T for cluster knot
Fruit;
Judge whether the cluster result reaches cluster target, obtains judging result;
When judging result expression is, the cluster result is exported;
When the judging result indicates no, the region threshold and the quantity threshold are adjusted, and is back to step and " determines
The region that the number of point in the region threshold in the voltage data matrix is greater than the quantity threshold is nucleus, by
The point of the nucleus constitutes the core dot matrix Q ".
4. the power battery fault detection method according to claim 3 based on high dimensional data diagnosis, which is characterized in that institute
State 10%~20% that cluster target is the number that the number of the preliminary abnormal point accounts for the point in the voltage data matrix.
5. the power battery fault detection method according to claim 1 based on high dimensional data diagnosis, which is characterized in that institute
It states according to the cluster result, calculates exceptional value score using rejecting outliers algorithm, specifically include:
The preliminary exception dot matrix T and core dot matrix Q is mapped into higher dimensional space;
Calculate the angle that any two point is formed in the point and the core dot matrix Q of the preliminary abnormal dot matrix T;
According to the angle calcu-lation included angle cosine value, included angle cosine value set is obtained;
The variance for calculating the cosine value set determines that the variance is the exceptional value score of the point of the sample data P.
6. a kind of power battery fault detection system based on high dimensional data diagnosis, which is characterized in that the system comprises
Pretreatment unit is pre-processed for the voltage data to battery cell, obtains voltage data matrix;
Clustering processing unit, for, to voltage data matrix progress clustering processing, obtaining cluster result using clustering algorithm,
The cluster result includes the preliminary abnormal dot matrix T being made of boundary point and abnormal point, the core point square being made of core point
Battle array Q;
Exceptional value score calculating unit, for calculating exceptional value score using rejecting outliers algorithm according to the cluster result;
Fail battery determination unit, for the size of the exceptional value score and threshold value, if the exceptional value score is less than
The threshold value determines that the corresponding battery cell of the exceptional value score breaks down, and exports the battery cell to break down
Number.
7. the power battery fault detection system according to claim 6 based on high dimensional data diagnosis, which is characterized in that institute
Pretreatment unit is stated to specifically include:
Data cleansing subelement, data and the record that the record moment is repeated in the voltage data for deleting the battery cell are wrong
Data accidentally, battery cell voltage data after being cleaned;
First-order difference subelement, for utilizing formula Δ Ui=Ui+1-UiOne is carried out to battery cell voltage data after the cleaning
Order difference processing, obtains the voltage data matrix;
Wherein, UiIndicate the voltage data of the battery cell of i-th acquisition, Δ UiIndicate the electricity acquired twice in succession
The difference of the voltage data of pond monomer, i indicate the number of the voltage data acquisition of the battery cell.
8. the power battery fault detection system according to claim 6 based on high dimensional data diagnosis, which is characterized in that institute
Clustering processing unit is stated to specifically include:
Threshold value initializes subelement, is used for initialization area threshold value and quantity threshold;
Nucleus determines subelement, and the number for determining the point in the voltage data matrix described in the region threshold is big
It is nucleus in the region of the quantity threshold, the core dot matrix Q is made of the point of the nucleus;
Boundary Region area determines subelement, for determining voltage data square described in the region threshold and outside the nucleus
The region that the number of point in battle array is less than the quantity threshold is Boundary Region area, and the point of the borderline region is the boundary point;
Abnormal point determines subelement, for determining that the point of the voltage data matrix described other than the region threshold is the exception
Point;The boundary point and the abnormal point constitute the preliminary abnormal dot matrix T;Determine the core dot matrix Q and it is described just
Step exception dot matrix T is cluster result;
Cluster result judgment sub-unit obtains judging result for judging whether the cluster result reaches cluster target;
Cluster result exports subelement, for exporting the cluster result when judging result expression is;
Adjusting thresholds subelement, for adjusting the region threshold and the quantity threshold when the judging result indicates no,
And it is back to step and " determines that the number of the point in the voltage data matrix described in the region threshold is greater than the quantity threshold
Region be nucleus, the core dot matrix Q " is made of the point of the nucleus.
9. the power battery fault detection system according to claim 8 based on high dimensional data diagnosis, which is characterized in that institute
State 10%~20% that cluster target is the number that the number of the preliminary abnormal point accounts for the point in the voltage data matrix.
10. the power battery fault detection system according to claim 6 based on high dimensional data diagnosis, which is characterized in that
The exceptional value score calculating unit specifically includes:
Subelement is mapped, for the preliminary exception dot matrix T and core dot matrix Q to be mapped to higher dimensional space;
Angle calcu-lation subelement, the point for calculating the preliminary abnormal dot matrix T and any two in the core dot matrix Q
The angle that point is formed;
Included angle cosine value computation subunit, for obtaining included angle cosine value set according to the angle calcu-lation included angle cosine value;
Exceptional value score determines subelement, for calculating the variance of the cosine value set, determines that the variance is the sample
The exceptional value score of the point of data P.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811336819.XA CN109765490B (en) | 2018-11-12 | 2018-11-12 | Power battery fault detection method and system based on high-dimensional data diagnosis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811336819.XA CN109765490B (en) | 2018-11-12 | 2018-11-12 | Power battery fault detection method and system based on high-dimensional data diagnosis |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109765490A true CN109765490A (en) | 2019-05-17 |
CN109765490B CN109765490B (en) | 2020-08-04 |
Family
ID=66449573
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811336819.XA Active CN109765490B (en) | 2018-11-12 | 2018-11-12 | Power battery fault detection method and system based on high-dimensional data diagnosis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109765490B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110781569A (en) * | 2019-11-08 | 2020-02-11 | 桂林电子科技大学 | Multi-resolution grid division based anomaly detection method and system |
CN111157850A (en) * | 2020-01-15 | 2020-05-15 | 上海电力大学 | Mean value clustering-based power grid line fault identification method |
CN111929591A (en) * | 2020-08-21 | 2020-11-13 | 彩虹无线(北京)新技术有限公司 | Fault battery detection method, device, equipment and computer storage medium |
CN112213657A (en) * | 2020-09-29 | 2021-01-12 | 江西江铃集团新能源汽车有限公司 | Battery early warning method and system |
CN112319309A (en) * | 2020-11-04 | 2021-02-05 | 北京理工大学 | Fault diagnosis method and system for power battery |
CN113376534A (en) * | 2021-05-12 | 2021-09-10 | 山东大学 | Phase plane diagnosis method for early fault of power battery and advanced early warning system |
WO2022030971A1 (en) * | 2020-08-03 | 2022-02-10 | 주식회사 엘지에너지솔루션 | Battery diagnosis device, battery pack, battery system, and battery diagnosis method |
CN114264957A (en) * | 2021-12-02 | 2022-04-01 | 东软集团股份有限公司 | Abnormal monomer detection method and related equipment thereof |
CN115267541A (en) * | 2022-07-29 | 2022-11-01 | 中国汽车工程研究院股份有限公司 | New energy automobile battery abnormity detection method based on countermeasure learning |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106021789A (en) * | 2016-06-01 | 2016-10-12 | 北京交通大学 | Fuzzy-intelligence-based rail car suspension system fault classification method and system |
US20160349330A1 (en) * | 2015-06-01 | 2016-12-01 | Verizon Patent And Licensing Inc. | Systems and methods for determining vehicle battery health |
JP2017092028A (en) * | 2015-11-02 | 2017-05-25 | 三星電子株式会社Samsung Electronics Co.,Ltd. | Battery management method and battery management apparatus |
CN107025384A (en) * | 2015-10-15 | 2017-08-08 | 赵乐平 | A kind of construction method of complex data forecast model |
CN107340475A (en) * | 2016-04-29 | 2017-11-10 | 株式会社日立制作所 | Battery fault detection method and battery fault detection device |
CN107755295A (en) * | 2017-10-19 | 2018-03-06 | 杭州电子科技大学 | A kind of lead-acid accumulator sorting technique based on charging and discharging curve |
-
2018
- 2018-11-12 CN CN201811336819.XA patent/CN109765490B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160349330A1 (en) * | 2015-06-01 | 2016-12-01 | Verizon Patent And Licensing Inc. | Systems and methods for determining vehicle battery health |
CN107025384A (en) * | 2015-10-15 | 2017-08-08 | 赵乐平 | A kind of construction method of complex data forecast model |
JP2017092028A (en) * | 2015-11-02 | 2017-05-25 | 三星電子株式会社Samsung Electronics Co.,Ltd. | Battery management method and battery management apparatus |
CN107340475A (en) * | 2016-04-29 | 2017-11-10 | 株式会社日立制作所 | Battery fault detection method and battery fault detection device |
CN106021789A (en) * | 2016-06-01 | 2016-10-12 | 北京交通大学 | Fuzzy-intelligence-based rail car suspension system fault classification method and system |
CN107755295A (en) * | 2017-10-19 | 2018-03-06 | 杭州电子科技大学 | A kind of lead-acid accumulator sorting technique based on charging and discharging curve |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110781569A (en) * | 2019-11-08 | 2020-02-11 | 桂林电子科技大学 | Multi-resolution grid division based anomaly detection method and system |
CN110781569B (en) * | 2019-11-08 | 2023-12-19 | 桂林电子科技大学 | Abnormality detection method and system based on multi-resolution grid division |
CN111157850A (en) * | 2020-01-15 | 2020-05-15 | 上海电力大学 | Mean value clustering-based power grid line fault identification method |
WO2022030971A1 (en) * | 2020-08-03 | 2022-02-10 | 주식회사 엘지에너지솔루션 | Battery diagnosis device, battery pack, battery system, and battery diagnosis method |
CN111929591A (en) * | 2020-08-21 | 2020-11-13 | 彩虹无线(北京)新技术有限公司 | Fault battery detection method, device, equipment and computer storage medium |
CN112213657A (en) * | 2020-09-29 | 2021-01-12 | 江西江铃集团新能源汽车有限公司 | Battery early warning method and system |
CN112319309A (en) * | 2020-11-04 | 2021-02-05 | 北京理工大学 | Fault diagnosis method and system for power battery |
CN112319309B (en) * | 2020-11-04 | 2021-12-31 | 北京理工大学 | Fault diagnosis method and system for power battery |
CN113376534A (en) * | 2021-05-12 | 2021-09-10 | 山东大学 | Phase plane diagnosis method for early fault of power battery and advanced early warning system |
CN113376534B (en) * | 2021-05-12 | 2022-04-05 | 山东大学 | Phase plane diagnosis method for early fault of power battery and advanced early warning system |
CN114264957A (en) * | 2021-12-02 | 2022-04-01 | 东软集团股份有限公司 | Abnormal monomer detection method and related equipment thereof |
CN115267541A (en) * | 2022-07-29 | 2022-11-01 | 中国汽车工程研究院股份有限公司 | New energy automobile battery abnormity detection method based on countermeasure learning |
Also Published As
Publication number | Publication date |
---|---|
CN109765490B (en) | 2020-08-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109765490A (en) | A kind of power battery fault detection method and system based on high dimensional data diagnosis | |
CN104849050B (en) | A kind of Fault Diagnosis of Roller Bearings based on compound multiple dimensioned arrangement entropy | |
CN106353690B (en) | Utilize the method for Petri network diagnosis lithium battery failure | |
CN110018425A (en) | A kind of power battery fault diagnosis method and system | |
CN102435910B (en) | Power electronic circuit health monitoring method based on support vector classification | |
CN101828918B (en) | Electrocardiosignal R peak detection method based on waveform characteristic matching | |
Costa et al. | Monitoring process mean and variability with one non-central chi-square chart | |
CN104868180B (en) | Cell method for group matching and system | |
CN102805620B (en) | Method and device for detecting atrial fibrillation | |
CN106526488A (en) | Fault diagnosis method of sensors in tandem type power battery pack | |
CN105550653A (en) | ECG lead intelligent selection method and system | |
CN113933732A (en) | New energy automobile power battery health state analysis method, system and storage medium | |
CN109670714B (en) | Ship gas turbine comprehensive state evaluation method based on membership degree analysis | |
CN110261746B (en) | Cable defect detection method based on periodic attenuation characteristics of oscillating wave voltage | |
CN108414896B (en) | Power grid fault diagnosis method | |
CN107167737A (en) | A kind of battery system method for safety monitoring | |
CN109365325B (en) | Manufacturing method and system of battery | |
CN104237803A (en) | Method for detecting electrical property of battery of electric vehicle on basis of battery working condition simulation | |
CN113866641A (en) | Fault detection method and device for lithium ion battery | |
CN115166563A (en) | Power battery aging state evaluation and decommissioning screening method and system | |
CN115656861A (en) | Power battery diagnosis device and method based on equalization circuit | |
CN111881594A (en) | Non-stationary signal state monitoring method and system for nuclear power equipment | |
CN113076354A (en) | User electricity consumption data analysis method and device based on non-invasive load monitoring | |
CN117102082A (en) | Sorting method and system for liquid metal batteries | |
CN110133530A (en) | Battery state detecting device and method |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |