CN115718262A - Power battery sampling abnormal risk identification method and fault judgment method - Google Patents
Power battery sampling abnormal risk identification method and fault judgment method Download PDFInfo
- Publication number
- CN115718262A CN115718262A CN202211493594.5A CN202211493594A CN115718262A CN 115718262 A CN115718262 A CN 115718262A CN 202211493594 A CN202211493594 A CN 202211493594A CN 115718262 A CN115718262 A CN 115718262A
- Authority
- CN
- China
- Prior art keywords
- sampling
- abnormal
- battery
- power battery
- risk
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Secondary Cells (AREA)
Abstract
The invention relates to the technical field of risk identification, and discloses a power battery sampling abnormity risk identification method and a fault judgment method, which comprise the following steps: step 1: collecting basic data of a battery; and 2, step: carrying out data cleaning and data preprocessing on the basic data of the battery; and 3, step 3: extracting an abnormal safety factor Cr from the sampled battery basic data processed in the step 2; the sampling abnormal safety factor Cr is a related projection variance parameter of the differential pressure change speed of the reference cell on the current change speed; and 4, step 4: carrying out nonlinear characteristic conversion and amplification processing on the abnormal safety factor Cr to obtain an abnormal factor Sf; and 5: and quantizing the sampling abnormal safety elements by using the variance entropy, and obtaining quantization characteristics. The method can finish early accurate identification of sampling abnormal risks, and can accurately identify and judge sampling abnormal faults.
Description
Technical Field
The invention relates to the technical field of risk identification, in particular to a power battery sampling abnormity risk identification method and a fault judgment method.
Background
With the increasing number of new energy automobiles, more and more new energy automobiles have obvious safety problems. Among them, the battery safety problem in new energy vehicles is one of the main safety problems of the new energy vehicles nowadays.
In the battery failure problem, sampling abnormality is a common failure phenomenon. The sampling abnormal fault mode refers to the abnormity of the sampling functional module caused by the failure of the single battery sampling module. The single battery sampling module relates to a single battery, a module, a sampling chip, a wire harness, a plug-in unit of the single battery sampling module and the like, wherein the wire harness is an important part of the single battery sampling module and is also a part which is easy to cause abnormal sampling. The abnormal sampling causes are many, including that the wire harness terminal has glue dust, so that the contact impedance is increased, and the voltage sampling is lost; the wiring harness locking terminal elastic sheet deforms, and voltage sampling is lost; the wiring harness connector retreats from the needle, so that voltage sampling is abnormal, and the like, and further the problems of change of resistance of a sampling line contact point, such as falling, breakage, damage and the like are caused.
At present, most of abnormal sampling identification depends on data extracted by a BMS system, but the data reported by the BMS system possibly has a lot of false alarm problems, because the fluctuation of the data is a normal phenomenon, the BMS system reports the fluctuant data, so that a lot of abnormal sampling is easily generated, the situation of false alarm is caused, and the abnormal judgment is not accurate. In addition, due to the fact that the actual operation process environment of the existing new energy automobile is changeable, the scene is complex, the operation data of the existing new energy automobile also has the characteristics of being multi-dimensional, redundant, heterogeneous and strongly coupled, and the like, and due to the fact that the physical characteristics of a battery system and the design and acquisition accuracy of a sensor are influenced, information coupling, redundancy and errors are inevitably generated among different signal data, and therefore data analysis and abnormal recognition judgment are more difficult and errors are larger.
In addition, the types of faults (i.e. types of risks) which are easily generated by the power battery are various, such as abnormal connection, abnormal self-discharge, abnormal capacity, abnormal internal resistance and the like are also included in addition to the sampling abnormity, and the occurrence of the faults intuitively causes abnormal fluctuation of the operation data of the power battery, and the similarity between the fluctuations is very high, so that the existence of the risks is relatively easy to determine based on the operation data.
Disclosure of Invention
The invention aims to provide a power battery sampling abnormal risk identification method and a fault judgment method, which can finish early accurate identification of sampling abnormal risks and can accurately identify and judge sampling abnormal faults.
To achieve the above problems, the present invention provides the following basic solutions:
the first scheme comprises the following steps:
a power battery sampling abnormal risk identification method and a fault judgment method comprise the following steps:
step 1: collecting basic data of a battery;
step 2: carrying out data cleaning and data preprocessing on the basic data of the battery;
and 3, step 3: extracting an abnormal safety factor Cr from the sampled battery basic data processed in the step 2; the sampling abnormal safety factor Cr is a related projection variance parameter of the differential pressure change speed of the reference cell on the current change speed;
and 4, step 4: carrying out nonlinear characteristic conversion and amplification processing on the abnormal safety factor Cr to obtain an abnormal factor Sf; and Sf = e αCr (ii) a Wherein alpha is a signal amplification coefficient;
and 5: quantizing the sampling abnormal safety elements by using the variance entropy to obtain quantization characteristics; the smaller the quantitative characteristic value is, the smaller the fluctuation degree of the sampling abnormal characteristic is and the smaller the abnormal risk is on the time scale.
Scheme II:
a power battery sampling abnormal fault determination method comprises the following steps:
s1: risk identification is carried out by adopting the power battery sampling abnormity risk identification method in the scheme one;
s2: comparing the risk probability value with a judgment threshold value, and taking a corresponding time point when the risk probability value is greater than the judgment threshold value as a high risk point;
s3: extracting abnormal element values and voltage values in a preset time period before and after the high risk point, and drawing a risk image;
s4: and judging sampling abnormal faults according to the fluctuation condition of each numerical value curve in the risk image.
The working principle and the advantages of the invention are as follows:
according to the scheme, by combining mechanism knowledge and data mining related technologies of sampling abnormity, starting from corresponding mechanism change of voltage signal data when the sampling abnormity is generated, by designing characterization characteristics of the sampling abnormity on the voltage signal data, namely corresponding to related projection variance parameters of the differential pressure change speed of a reference battery cell on the current change speed, abnormity which is difficult to capture from complicated data is extracted in a sampling abnormity safety element form, nonlinear characteristic conversion and amplification processing and variance entropy are adopted, the abnormity is processed into dimensionless quantitative characteristic data, fluctuation conditions and risk conditions of the abnormal characteristics on a time scale can be visually represented, further data expression and automatic identification of fault information can be effectively realized, and accurate identification of new energy vehicles and accurate judgment of sampling abnormity faults can be realized on the basis of vehicle operation historical data.
In addition, particularly, when risks are identified and abnormal conditions are judged, abnormal data are not directly captured, fluctuation conditions of specific characteristics of voltage signal data, namely relevant projection variance parameters of differential pressure change speed on current change speed, are further captured in the sampling abnormal generation process or under the condition that the sampling abnormal conditions exist, the parameters actually describe mechanism states corresponding to the sampling abnormal conditions (namely adjacent cell voltages are over-high and over-low, adjacent cell voltages are reversely deviated and the like), and characteristic mining is performed through describing and mining the mechanism states, so that sampling abnormal faults can be directionally and accurately identified.
Moreover, the parameter embodies pure characteristic fluctuation, is relatively not influenced by the size deviation of the numerical value of the data, simultaneously, tiny abnormity which cannot be embodied by a pure voltage numerical value is covered under the numerical coupling error, the parameter can be captured meticulously, early accurate identification of sampling abnormity risks is facilitated, namely the sampling abnormity risks can be identified directionally and accurately in the process of evolution and formation of sampling abnormity, and the sampling abnormity can be excavated directionally in similar numerical value representations. By calculating and extracting the parameter (namely sampling abnormal safety element), redundant information and noise information generated by interference of external factors in the representation signal of the power battery can be effectively eliminated, the signal characteristic of the representation power battery running state can be more accurately extracted, and the obtained safety element can effectively and accurately describe the characteristic change condition of the voltage signal in the running process of the vehicle, namely the power battery running process, so that the sampling abnormal fault can be accurately identified and judged.
Drawings
Fig. 1 is a schematic flow chart of a risk identification method according to an embodiment of a power battery sampling abnormality risk identification method and a fault determination method of the present invention;
fig. 2 is a schematic flow chart of a fault determination method according to an embodiment of the power battery sampling abnormal risk identification method and the fault determination method of the present invention;
fig. 3 is a schematic view of a risk image of an embodiment of a power battery sampling abnormality risk identification method and a fault determination method according to the present invention.
Detailed Description
The following is further detailed by the specific embodiments:
the embodiment is basically as shown in the attached figure 1: a power battery sampling abnormal risk identification method and a fault judgment method comprise the following steps:
step 1: collecting basic data of a battery; the basic data of the battery is obtained by analyzing from a message log of the power battery. Specifically, the battery basic data in this embodiment is obtained by analyzing a packet log of the power battery system that conforms to the GB32960 protocol, and basic data reliability is high.
In addition, in the embodiment, for the analyzed power battery of the new energy automobile in the complex operation environment and with variable operation process environment, the corresponding operation data has the characteristics of multiple dimensions, redundancy, heterogeneity, strong coupling and the like, and is difficult to analyze. The following steps of the scheme can accurately and directionally realize early judgment of the sampling abnormal risk from the complex data.
Step 2: and carrying out data cleaning and data preprocessing on the basic data of the battery. Specifically, the data cleaning is to clean abnormal data in the basic data of the battery, wherein the abnormal data specifically refers to the data of the voltage and current signals of the basic data of the battery, which exceed a specified threshold. The data preprocessing comprises the following steps: (1) Identifying and marking interference pulses, and marking the frame data if the difference value between the voltage data of the current frame and the voltage data of the previous frame exceeds a specified threshold value so as to be distinguished in the subsequent evaluation; (2) discontinuity identification and marking: if the difference value between the current frame timestamp data and the previous frame exceeds a specified threshold value, marking the frame data so as to be convenient for distinguishing in the subsequent evaluation; specifically, during subsequent evaluation, the risk of the marking point is processed by setting 0, which is beneficial to improving the risk identification accuracy. And (3) average filtering to reduce noise data.
And 3, step 3: extracting an abnormal safety factor Cr from the sampled battery basic data processed in the step 2; and the sampling abnormal safety factor Cr is a related projection variance parameter of the differential pressure change speed of the reference battery cell on the current change speed.
The reference cell is selected by adopting the following substeps:
substep 1: calculating a difference matrix Vd of adjacent electric cores of the power battery;
substep 2: constructing a symbolic function; the symbolic function is
Vaa is a range voltage, that is, a difference between a maximum value of a cell voltage and a minimum value of the cell voltage at any time; and alpha is belonged to (0, 1) and is a sampling abnormal characteristic characterization coefficient.
Substep 2.1: and carrying out mean value filtering processing on the symbol function, and smoothing the step symbol function to achieve the effect of signal expansion of the step position.
Substep 3: and calculating the symbol function offset of each battery cell, sequencing, and selecting the battery cell corresponding to the value with the maximum offset as a reference battery cell.
The reference battery cell obtained through the steps is the battery cell which can reflect the characteristic (identification element) fluctuation condition in the power battery system most, and compared with the method for analyzing all the battery cells generally, the method effectively limits the analysis range to one reference battery cell, greatly reduces the data processing workload, and effectively makes up the analysis precision gap after the analysis is simplified by selecting the obtained battery cell electrode.
The sampling anomaly safety factor
In the formula (I), the compound is shown in the specification,
V v representing the vector, I, corresponding to the position of the reference cell in the differential matrix v Representing the current variation velocity vector of the reference cell.
And 4, step 4: carrying out nonlinear characteristic conversion and amplification processing on the abnormal safety factor Cr to obtain an abnormal factor Sf; and Sf = e αCr (ii) a Wherein alpha is a signal amplification coefficient;
and 5: quantizing the sampling abnormal safety elements by using the variance entropy to obtain quantization characteristics; the smaller the quantitative characteristic value is, the smaller the fluctuation degree of the sampling abnormal characteristic is and the smaller the abnormal risk is on the time scale.
Specifically, the quantization feature p =1- λ; wherein λ = E 2 (Sf)/( 2 ) (ii) a And λ is more than or equal to 0 and less than or equal to 1, and the closer λ is to 1, the smaller p is, the smaller the fluctuation degree of the sampling abnormal features is on the time scale, and the smaller the abnormal risk is.
Step 6: and analyzing to obtain a risk probability value based on the quantitative characteristics.
Specifically, discrete integration is performed on p on a time scale, a discrete integration function Sp = ∑ p is obtained, sp is a monotonously increasing curve, the amplitude of the slope of the discrete integration function Sp curve is used as a risk probability value, the greater the risk probability value is, the greater the fluctuation degree of the sampled abnormal features is, and the greater the abnormal risk is.
As shown in fig. 2, a power battery sampling abnormal fault determination method includes the following steps:
s1: risk identification is carried out by adopting the power battery sampling abnormity risk identification method;
s2: comparing the risk probability value with a judgment threshold value, and taking a corresponding time point when the risk probability value is greater than the judgment threshold value as a high risk point; in the present embodiment, the determination threshold is set to 0.5, and the threshold is appropriately set, so that the high risk point can be effectively confirmed.
S3: extracting abnormal element values and voltage values in a preset time period before and after the high risk point, and drawing a risk image; as shown in fig. 3.
S4; and judging sampling abnormal faults according to the fluctuation condition of each numerical value curve in the risk image.
When the fluctuation condition of the voltage value curve shows that the voltages of a plurality of adjacent battery cells deviate towards two opposite directions, the sampling abnormal fault is judged to exist. Specifically, when the voltages of two adjacent cells are deviated towards two opposite directions, that is, compared with other normal cells, the voltages of the two cells are obviously higher and lower, and a single cell in the two cells does not have the problem of high charging and low discharging (high charging voltage and low discharging voltage), it is determined that a sampling abnormal fault exists, and the determination mode is intuitive and effective.
According to the power battery sampling abnormity risk identification method and the fault judgment method provided by the embodiment, mechanism knowledge and data mining related technologies are combined, and the data expression and automatic identification of fault information are effectively realized by designing the characterization characteristics of sampling abnormity on voltage signal data. And by constructing a new energy automobile safety state identification and fault mode judgment technical system of data processing, element extraction, characteristic quantification, risk identification and fault judgment, the accurate identification of new energy automobile risks and the accurate judgment of sampling abnormal faults can be realized based on the automobile operation historical data.
The foregoing is merely an example of the present invention, and common general knowledge in the field of known specific structures and characteristics is not described herein in any greater extent than that known in the art at the filing date or prior to the priority date of the application, so that those skilled in the art can now appreciate that all of the above-described techniques in this field and have the ability to apply routine experimentation before this date can be combined with one or more of the present teachings to complete and implement the present invention, and that certain typical known structures or known methods do not pose any impediments to the implementation of the present invention by those skilled in the art. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several variations and modifications can be made, which should also be considered as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the utility of the patent.
Claims (8)
1. A power battery sampling abnormal risk identification method is characterized by comprising the following steps:
step 1: collecting basic data of a battery;
and 2, step: carrying out data cleaning and data preprocessing on the basic data of the battery;
and step 3: extracting an abnormal safety factor Cr from the sampled battery basic data processed in the step 2; the sampling abnormal safety factor Cr is a related projection variance parameter of the differential pressure change speed of the reference cell on the current change speed;
and 4, step 4: carrying out nonlinear characteristic conversion and amplification processing on the abnormal safety factor Cr to obtain an abnormal factor Sf; and Sf = e αCr (ii) a Wherein alpha is a signal amplification coefficient;
and 5: quantizing the sampling abnormal safety elements by using the variance entropy to obtain quantization characteristics; the smaller the quantitative characteristic value is, the smaller the fluctuation degree of the sampling abnormal characteristic is and the smaller the abnormal risk is on the time scale.
2. The power battery sampling abnormality risk identification method according to claim 1, wherein in step 3, the reference cell is selected by adopting the following substeps:
substep 1: calculating a difference matrix Vd of adjacent battery cells of the power battery;
and substep 2: constructing a symbolic function; the sign function is
Vaa is a range voltage, that is, a difference between a maximum value of a cell voltage and a minimum value of the cell voltage at any time; alpha belongs to (0, 1) and is a sampling abnormal characteristic characterization coefficient;
substep 3: and calculating the symbol function offset of each battery cell, sequencing, and selecting the battery cell corresponding to the maximum offset value as a reference battery cell.
3. The power battery sampling abnormality risk identification method according to claim 2, characterized in that in sub-step 2, the method further comprises sub-step 2.1: and carrying out mean value filtering processing on the symbol function, and smoothing the step symbol function to achieve the effect of signal expansion of the step position.
4. The power battery sampling abnormity risk identification method according to claim 2, characterized in that the sampling abnormity safety element
In the formula (I), the compound is shown in the specification,
V v representing the vector, I, corresponding to the position of the reference cell in the differential matrix v And representing the current change speed vector of the reference cell.
5. The power battery sampling abnormity risk identification method according to claim 1, wherein in step 1, the battery basic data is obtained by parsing from a message log of the power battery.
6. The power battery sampling abnormity risk identification method according to claim 1, further comprising the following steps of 6: and analyzing to obtain a risk probability value based on the quantitative characteristics.
7. A power battery sampling abnormal fault judgment method is characterized by comprising the following steps:
s1: risk identification is carried out by adopting the power battery sampling abnormity risk identification method according to any one of claims 1-6;
s2: comparing the risk probability value with a judgment threshold value, and taking a corresponding time point when the risk probability value is greater than the judgment threshold value as a high risk point;
s3: extracting abnormal element values and voltage values in a preset time period before and after the high risk point, and drawing a risk image;
s4: and judging sampling abnormal faults according to the fluctuation condition of each numerical value curve in the risk image.
8. The method for determining the sampling abnormal fault of the power battery according to claim 7, wherein when the fluctuation of the voltage value curve shows that the voltages of the adjacent battery cells are shifted towards two opposite directions, it is determined that the sampling abnormal fault exists.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211493594.5A CN115718262A (en) | 2022-11-25 | 2022-11-25 | Power battery sampling abnormal risk identification method and fault judgment method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211493594.5A CN115718262A (en) | 2022-11-25 | 2022-11-25 | Power battery sampling abnormal risk identification method and fault judgment method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115718262A true CN115718262A (en) | 2023-02-28 |
Family
ID=85256478
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211493594.5A Pending CN115718262A (en) | 2022-11-25 | 2022-11-25 | Power battery sampling abnormal risk identification method and fault judgment method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115718262A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116304582A (en) * | 2023-05-16 | 2023-06-23 | 力高(山东)新能源技术股份有限公司 | Abnormal mark correction method for monotone data in power battery |
-
2022
- 2022-11-25 CN CN202211493594.5A patent/CN115718262A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116304582A (en) * | 2023-05-16 | 2023-06-23 | 力高(山东)新能源技术股份有限公司 | Abnormal mark correction method for monotone data in power battery |
| CN116304582B (en) * | 2023-05-16 | 2023-08-08 | 力高(山东)新能源技术股份有限公司 | Abnormal mark correction method for monotone data in power battery |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109765490B (en) | Power battery fault detection method and system based on high-dimensional data diagnosis | |
| CN111766462B (en) | Non-invasive load identification method based on V-I track | |
| CN113415165B (en) | Fault diagnosis method and device, electronic equipment and storage medium | |
| CN115718262A (en) | Power battery sampling abnormal risk identification method and fault judgment method | |
| CN115220396B (en) | Intelligent monitoring method and system for numerical control machine tool | |
| CN115267589B (en) | Multi-parameter joint diagnosis method for battery faults of electric vehicle | |
| CN115384311A (en) | Method, system, vehicle and storage medium for diagnosing insulation fault of power battery | |
| CN115437358A (en) | Intelligent state monitoring and fault diagnosis system and fault diagnosis method for industrial robot | |
| CN108919041B (en) | Transformer winding state online monitoring method based on cluster analysis | |
| CN115792634A (en) | Battery monomer voltage sampling fault identification method based on cloud online data | |
| CN115144172A (en) | GIS breaker fault online monitoring system and method based on sound | |
| CN115796583A (en) | New energy automobile risk identification model based on self-discharge abnormal characteristics | |
| CN114509703A (en) | Sampling line disconnection judging method, system, detector and storage medium | |
| CN115842393A (en) | New energy automobile risk identification model based on capacity abnormity and decline characteristics | |
| CN112528753A (en) | Preprocessing method for impact vibration signals in rolling process of rolling mill | |
| CN112525337A (en) | Method for preprocessing vibration monitoring data of mechanical press | |
| CN112381130A (en) | Cluster analysis-based power distribution room multivariate data anomaly detection method | |
| CN111735559A (en) | Capacitive edge calculation pressure transmitter acquisition and diagnosis circuit and working method thereof | |
| CN115834364B (en) | Power consumption information monitoring system and method based on power big data | |
| CN116400231B (en) | Battery multi-fault detection method and device of energy storage system and electronic equipment | |
| CN117630459A (en) | Power equipment fault detection method based on composite signal | |
| CN117388690A (en) | Motor fault detection method based on abnormal signal extraction analysis | |
| CN220232294U (en) | STM 32-based multichannel equipment running state analyzer | |
| CN115902637A (en) | Power battery stability evaluation method based on projection correlation | |
| CN115754826A (en) | Power battery connection abnormity risk assessment 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 |