Detailed Description
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The invention belongs to the technical field of battery management, and discloses a battery system early fault diagnosis method which can detect, distinguish and locate whether a battery system has a battery short circuit fault, an electric connection fault and a voltage sensor fault. The method mainly comprises the following steps of monitoring the voltage of a battery cell and the voltage of a connecting resistor in real time by using a non-redundant cross measurement topology, setting a time window and a 'first in first out' data cache, arranging detection data in fixed time into an Mk matrix, and carrying out feature matching on Mk according to a statistical method to obtain a fault type and a fault position.
Example 1
The embodiment provides a battery pack fault early warning method;
As shown in fig. 1, the battery pack fault early warning method includes:
s101, acquiring voltage data of a battery pack, wherein the voltage data comprises single voltage and combined voltage, and the combined voltage refers to two adjacent single batteries and a resistor connected between the two single batteries;
s102, expanding the voltage data of the battery pack by adopting a sliding window to obtain a two-dimensional data analysis matrix;
S103, carrying out standardization processing on the two-dimensional data analysis matrix to obtain a standardized matrix, and further obtaining a covariance matrix corresponding to the standardized matrix, wherein the covariance matrix corresponding to the standardized matrix comprises five submatrices;
s104, calculating the expectation of the maximum eigenvalue of each sub-matrix, determining the expected value of a fault indicator of the voltage sensor fault according to the expectation of the maximum eigenvalue of each sub-matrix, and determining whether the voltage sensor has the fault according to the expected value of the fault indicator.
Further, the method further comprises:
S105, carrying out normalization processing on the two-dimensional data analysis matrix to obtain a normalization matrix, and further obtaining a covariance matrix corresponding to the normalization matrix, wherein the covariance matrix corresponding to the normalization matrix comprises three new submatrices;
S106, calculating the expectation of the maximum eigenvalue of each new submatrix, determining an early battery short-circuit fault threshold and an early electric connection fault threshold according to the expectation of the maximum eigenvalue of each new submatrix, determining whether the battery pack has battery short-circuit fault according to the early battery short-circuit fault threshold, and determining whether the battery pack has electric connection fault according to the early electric connection fault threshold.
Further, the method further comprises:
And S107, determining the position of the fault through the minimum value of the feature vector element corresponding to the maximum feature value.
Further, as shown in fig. 2, the step S101 of acquiring voltage data of the battery pack, where the voltage data includes a cell voltage and a combined voltage, where the combined voltage refers to two adjacent cells and a resistor connected between the two cells, and specifically includes:
The battery pack comprises a battery pack negative electrode, a 1 st connecting point, a resistor R0,1, a 2 nd connecting point, a 1 st battery cell, a 3 rd connecting point, a resistor R1,2, a 4 th connecting point, a 2 nd battery cell, a 5 th connecting point, a resistor R2,3, a 6 th connecting point, a 3 rd battery cell, a 7 th connecting point, a resistor R3,4, an 8 th connecting point and a 4 th battery cell which are sequentially connected, wherein n is more than or equal to 1;
A first voltage sensor is connected between the 2 nd connecting point and the 5 th connecting point, and the voltage acquired by the first voltage sensor is thatA second voltage sensor is connected between the 4 th connecting point and the 7 th connecting point, and the voltage acquired by the second voltage sensor is thatA third voltage sensor is connected between the 6 th connecting point and the 9 th connecting point, and the voltage acquired by the third voltage sensor is thatAnd so on;
An nth voltage sensor is connected between the 2nth connection point and the 2n+3th connection point, and the voltage acquired by the nth voltage sensor is that
Obtaining a measurement data vectorWherein n is the number of batteries connected in series;
According to the measurement topology and kirchhoff voltage law, the decomposition formula of the measurement data vector is known as:
Wherein,Is a vector of voltage values measured by the voltage sensor,Is a battery terminal voltage vector, Rn,n+1 is an equivalent connection resistance between the nth single battery and the (n+1) th single battery, Rb,1 is a resistance connected with the negative electrode of the battery pack, Rn,b is a resistance connected with the positive electrode of the battery pack, and each voltage sensor is used for measuring the voltage of 2 adjacent single batteries and the equivalent connection resistance between the adjacent single batteries.
It should be appreciated that the change in the voltage detection amount caused by the current change remains highly consistent with the current, while the change in the voltage detection amount caused by the fault has poor correlation with the current, a characteristic that can be used to detect and distinguish between battery system faults, each cell comprising one cell or a plurality of cells connected in parallel.
Further, the step S102 of expanding the voltage data of the battery pack by a sliding window to obtain a two-dimensional data analysis matrix specifically comprises:
Setting the size and the sliding step length w of a sliding window, sliding the sliding window on a measured data vector Vs, obtaining a two-dimensional data analysis submatrix by one step after sliding, and finally obtaining a two-dimensional data analysis matrix Mk:
Mk＝[Vs|k-w+1,Vs|k-w+2,...,Vs|k]T (2)
Wherein,
Vs|k represents the es state at the kth time within the time window.
Abstracting the effect of the error as a vector e, reducing equation (1) to m=ax1+Bx2+e＝m* +e, where m* represents the fault-free part of vector m, e being the measurement error, can be considered to be satisfiedWhite noise of (a), and
Further, the step S103 is to perform standardization processing on the two-dimensional data analysis matrix to obtain a standardized matrix, and further obtain a covariance matrix corresponding to the standardized matrix, wherein the covariance matrix corresponding to the standardized matrix comprises five submatrices, and specifically comprises the following steps:
The two-dimensional data analysis matrix Mk is standardized to enableRepresenting no fault, no error component, where
Wherein,Is a standard deviation diagonal array of Yk,AndThe average vectors of x1 and x2, respectively.
Adding the fault amount can obtain:
Wherein,Representing no fault component, fk representing a fault component in a time window, which includes characteristics of fault duration, ζ representing a fault weight, only the fault channel being non-zero, and ζ2 =1; As an error component, represents the normal measurement error of the voltage sensor.
Covariance matrix of (2) isFor analysis convenience, it was split into 5 parts,
Wherein the method comprises the steps of
It should be appreciated that the voltage sensor failure modes match. When a certain voltage sensor fails, it can be considered that the detected value thereof overlaps with the failure component f, and then m=m* +ζf exists. m is an actual detection value, m* is a fault-free component, ζf is a fault component, ζ is a fault weight, only the fault channel is non-zero, and ζζ2 =.
It should be appreciated that the fault detection indicator is directly related to the eigenvalue distribution of the covariance matrix Ck. Thus, the expected deployment of eigenvalues is studied to gain insight into the correlation between the fault magnitude and the detection threshold. The lower bound of each submatrix { Ci } is derived separately.
Further, the step S104 of calculating the expectation of the maximum eigenvalue of each submatrix specifically comprises:
(1)C1,k
according to the matrix multiplication switching law, C1,k is expressed as:
Wherein,
Through eigenvalue decomposition, the method can obtain:
Wherein,To be arranged in descending orderIs a diagonal matrix of eigenvalues of (a),Is the corresponding feature vector.
Eigenvalues of right part of upper equal signThe method meets the following conditions:
When (when)In the time-course of which the first and second contact surfaces,At the position ofInternal strictly monotonically increasing, it can be seen that
For maximum eigenvalues
Namely:
Then the first time period of the first time period,Is satisfied by the following:
Where tf is the duration of the fault within the sliding window.
According to the Weyl inequality, the lower limit of the maximum eigenvalue of Ck is:
the Weyl inequality is defined as follows:
Let a, B e Mn be a Hermite matrix, let the individual eigenvalues λi(A),λi(B),λi (a+b) all be arranged in ascending order, then for each k=1, 2..n there is:
λk(A)+λ1(B)≤λk(A+B)≤λk(A)+λn(B)
(2)C2,k+C3,k
Since ζl =1 in the initial fault condition, C2,k+C3,k can be expanded as:
Wherein,Is thatIs the ith column of (2). Order theThe characteristic polynomial of C2,k+C3,k can be expressed as
The method is characterized by comprising the following steps of:
Then, the non-zero eigenvalues of C2,k+C3,k are:
wherein the method comprises the steps ofThus, the first and second substrates are bonded together,Is satisfied by the following:
Thus, the first and second substrates are bonded together,
(3)C4,k
Unfolding C4,k can obtain:
Its characteristic valueSatisfy the following requirements
Wherein,AndRespectively areAndIs the j-th column of (2).
Order theAnd:
Then
Wherein the method comprises the steps ofThus, the first and second substrates are bonded together,
Due toThus, the first and second substrates are bonded together,
Thus, it is possible to obtain:
(4)C5,k
since the noise level of each voltage channel is almost the same, letIs the minimum characteristic value of C5,k,The minimum eigenvalue of normalized (n-1)-1ET E can be obtained:
I.e.
According to Marchenko-Pastur law in the large-dimensional random matrix theory,Because of the fact that n, w → +. Infinity of the two points, c= (n-1)/w e (0, 1), thus:
Designing fault indicators as
Wherein lambdai,k is within the time windowEigenvalues of covariance matrix Ck,For its average value over a time window,For its standard deviation within a time window
Further, S104, determining a fault indicator expected value of the voltage sensor fault according to the expected maximum eigenvalue of each submatrix, specifically comprising:
The expected value of the fault indicator of the voltage sensor fault is as follows, which are available from the formulas (5), (8), (10), (11) and (12):
Wherein,Is the expected value of lambdaj,k.
Further, S104, determining whether the voltage sensor fails according to the expected value of the fault indicator, wherein the method specifically comprises the step of indicating that the voltage sensor fails in the system if the value of the fault indicator is smaller than the expected value of the fault indicator.
Further, S105, normalizing the two-dimensional data analysis matrix to obtain a normalized matrix, further obtaining a covariance matrix corresponding to the normalized matrix, wherein the covariance matrix corresponding to the normalized matrix comprises three new submatrices
Assuming that an incipient short-circuit fault f occurs in the first channel x1,l, the differential measurement vector is described as:
m=A(x1+ξf)+Bx2+e;
the sample matrix Mk in the sliding window is normalized to:
Is expressed as covariance matrix of (2)Wherein the method comprises the steps of
The short circuit fault is located at cell voltage vector x1 and the connection fault is located at contact resistance voltage vector x2. The locations of the two electrical faults are similar, and the statistical analysis method is also similar.
Bonding (12) (14) is provided with
Further, S106, calculating the expectation of the maximum eigenvalue of each new submatrix, specifically comprising:
If it is desired toThenWhere i ε {1,., n-1}, δsc is the detection threshold for a short circuit fault.
From the additivity of the matrix trace, it is possible to obtainSince the characteristic value is expected to satisfyThe method can obtain:
In the case of the C2,k,From the matrix trace and the desired switching law, it is possible to:
for C3,k, according to the switching law of the matrix trace,Wherein,Representation ofIs the first column of (c), and therefore,Then there are:
Further, S106, determining an early battery short-circuit fault threshold and an early electrical connection fault threshold according to the expectation of the maximum eigenvalue of each new submatrix, determining whether the battery pack has a battery short-circuit fault according to the early battery short-circuit fault threshold, and determining whether the battery pack has an electrical connection fault according to the early electrical connection fault threshold, wherein the method specifically comprises the following steps:
for early battery short-circuit failure, there are combinations (15) (16) (17) (18)
The mathematical representation of an electrical connection fault is similar to a short circuit fault. Thus, for early electrical connection failure, there is
If equation (19) holds, it indicates that a battery short-circuit fault has occurred;
If equation (20) holds, it indicates that an electrical connection failure has occurred.
When the fault assessment is below the corresponding detection threshold, a corresponding fault is declared to the system.
Further, S107, determining the position of the fault through the minimum value of the feature vector element specifically comprises:
when a voltage sensor fault occurs, setting the minimum element index in the feature vector vm,k as i, and then generating the fault on the ith voltage sensor;
When a short circuit fault occurs, a pair of adjacent elements with indexes of { i-1, i } which are smaller than the average value in the characteristic vector vm,k are found, and the fault occurs in the ith battery cell;
When an electrical connection failure occurs, the smallest element index in the feature vector vm,k is set to i, and the failure occurs between the i-th battery cell and the i+1th battery cell.
It should be appreciated that, based on the analysis results, the contribution of the ith row of the covariance matrix to the eigenvalue λm,k may be measured by the ith term of the corresponding eigenvector vm,k. After the fault determination, the location of the fault occurrence can be determined by normalizing the contribution of the outlier lambda1,k.
The invention tests in a battery system consisting of 5 strings of batteries, and fault detection and positioning results are shown in fig. 3 (a), 3 (b), 4 (a), 4 (b), 5 (a) and 5 (b).
Example two
The embodiment provides a battery pack fault early warning system, which comprises:
The acquisition module is configured to acquire voltage data of the battery pack, wherein the voltage data comprises a single voltage and a combined body voltage, and the combined body refers to two adjacent single batteries and a resistor connected between the two single batteries;
the expansion module is configured to expand the voltage data of the battery pack by adopting a sliding window to obtain a two-dimensional data analysis matrix;
The processing module is configured to perform standardization processing on the two-dimensional data analysis matrix to obtain a standardized matrix, and further obtain a covariance matrix corresponding to the standardized matrix, wherein the covariance matrix corresponding to the standardized matrix comprises five submatrices;
The voltage sensor fault determination module is configured to calculate a desire for a maximum eigenvalue of each sub-matrix, determine a fault indicator desire value for a voltage sensor fault based on the desire for the maximum eigenvalue of each sub-matrix, and determine whether the voltage sensor has a fault based on the fault indicator desire value.
Here, it should be noted that the above-mentioned obtaining module, expanding module, processing module and determining module correspond to steps S101 to S104 in the first embodiment, and the above-mentioned modules are the same as examples and application scenarios implemented by the corresponding steps, but are not limited to the disclosure of the first embodiment. It should be noted that the modules described above may be implemented as part of a system in a computer system, such as a set of computer-executable instructions.
The foregoing embodiments are directed to various embodiments, and details of one embodiment may be found in the related description of another embodiment.
The proposed system may be implemented in other ways. For example, the system embodiments described above are merely illustrative, such as the division of the modules described above, are merely a logical function division, and may be implemented in other manners, such as multiple modules may be combined or integrated into another system, or some features may be omitted, or not performed.
Example III
The embodiment also provides an electronic device comprising one or more processors, one or more memories, and one or more computer programs, wherein the processors are connected with the memories, the one or more computer programs are stored in the memories, and when the electronic device is operated, the processors execute the one or more computer programs stored in the memories, so that the electronic device executes the method in the embodiment.
It should be understood that in this embodiment, the processor may be a central processing unit CPU, and the processor may also be other general purpose processors, digital signal processors DSP, application specific integrated circuits ASIC, off-the-shelf programmable gate array FPGA or other programmable logic device, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory may include read only memory and random access memory and provide instructions and data to the processor, and a portion of the memory may also include non-volatile random access memory. For example, the memory may also store information of the device type. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or by instructions in the form of software.
The method in the first embodiment may be directly implemented as a hardware processor executing or implemented by a combination of hardware and software modules in the processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in a memory, and the processor reads the information in the memory and, in combination with its hardware, performs the steps of the above method. To avoid repetition, a detailed description is not provided herein.
Those of ordinary skill in the art will appreciate that the elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
Fourth embodiment the present embodiment also provides a computer-readable storage medium storing computer instructions that, when executed by a processor, perform the method of the first embodiment.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.