CN115683661A - Automobile fault detection method, device, equipment and storage medium - Google Patents
Automobile fault detection method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN115683661A CN115683661A CN202211395160.1A CN202211395160A CN115683661A CN 115683661 A CN115683661 A CN 115683661A CN 202211395160 A CN202211395160 A CN 202211395160A CN 115683661 A CN115683661 A CN 115683661A
- Authority
- CN
- China
- Prior art keywords
- matrix
- kernel
- principal component
- automobile
- data
- 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
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention relates to the technical field of automobile fault detection, in particular to an automobile fault detection method, device, equipment and storage medium. The method comprises the steps of firstly collecting data of each moment of a detected automobile in real time, forming a current sample matrix by using the data of the current moment in the data of each moment, then applying a dynamic principal component analysis algorithm to the current sample matrix to obtain an augmented matrix of the current sample matrix, then applying a kernel principal component analysis algorithm to the augmented matrix to obtain a kernel matrix, then constructing the principal component matrix, and finally determining the detection result of the automobile under the combined action of the principal component matrix and a feature vector of the kernel matrix. The data used for judging the detection result is from the detected automobile, the objective data of the automobile data is input into the mathematical model formed by the dynamic principal component analysis algorithm and the kernel principal component analysis algorithm, and the automobile fault can be quantitatively detected according to the data output by the data model, so that the detection accuracy is improved.
Description
Technical Field
The invention relates to the technical field of automobile fault detection, in particular to an automobile fault detection method, device, equipment and storage medium.
Background
The automobile fault detection includes a test logic detection method (a logic detection method for detecting whether an error exists through a logic relation between statistical data, wherein the logic relation of the statistical data is a reflection of the logic relation of the objective existence of things and the motion development of the things) and a physical detection method. The former has huge content due to the repeated occurrence of a large amount of test logics, and changes corresponding functional logics aiming at different automobiles, and because the functional logics are complex, functional logic errors are easily caused in the changing process, so that the accuracy of detecting automobile faults by using the changed functional logics is reduced. The physical detection method comprises a detection diagnosis method based on human experience, an instrument detection diagnosis method and a self-diagnosis method, the detection diagnosis method based on human experience needs maintenance personnel to position faults according to previous work experience and simple tests, and the accuracy of a detection result is poor due to lack of support of objective data; the instrument detection diagnosis method utilizes parameters, states, curves and waveforms of instrument measurement data to diagnose faults. Maintenance personnel still need to deal with many types of faults that have not been incorporated into the mathematical model; the self-diagnosis method cross-checks a signal level of an Electronic Control Unit (ECU) using a self-diagnosis system of a vehicle, and reference values thereof are stored in a memory. If the level of the signal exceeds the allowable limit, the ECU recognizes the signal as a fault signal and transmits a fault code to the memory. However, the self-diagnosis system of the vehicle has not reached a level of ensuring an accurate detection rate.
In conclusion, the existing automobile fault detection method is low in accuracy.
Thus, there is a need for improvements and enhancements in the art.
Disclosure of Invention
In order to solve the technical problems, the invention provides a method, a device, equipment and a storage medium for detecting automobile faults, and solves the problem that the existing automobile fault detection method is low in accuracy.
In order to achieve the purpose, the invention adopts the following technical scheme:
in a first aspect, the present invention provides a method for detecting a failure of a vehicle, wherein the method comprises:
applying a dynamic principal component analysis algorithm to a current sample matrix to obtain an augmentation matrix corresponding to the current sample matrix, wherein the current sample matrix is formed by current time data of a detected automobile;
applying a kernel principal component analysis algorithm to the augmentation matrix to obtain a kernel matrix corresponding to the augmentation matrix;
obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix and the kernel matrix;
and obtaining the detection result of the detected automobile according to the characteristic vectors of the principal component matrix and the kernel matrix.
In one implementation, the applying a dynamic principal component analysis algorithm to the current sample matrix to obtain an augmented matrix corresponding to the current sample matrix, where the current sample matrix is formed by current time data of the detected automobile, includes:
setting a model order of the dynamic principal component analysis algorithm;
obtaining the number of matrix columns according to the model order;
obtaining the number of matrix rows according to the model order and the total number of samples, wherein the total number of samples is the number of samples contained in a sample library where the current sample is located;
sequentially selecting samples positioned before the current sample from the sample library until the number corresponding to the selected samples reaches the matrix column number, wherein the generation time of data included in the samples positioned before the current sample is positioned before the generation time of the data included in the current sample, and the selected samples are marked as a first sample group;
sequentially selecting samples positioned before the current sample from the sample library until the number corresponding to the selected samples reaches the number of the matrix rows, and marking the selected samples as a second sample group;
and constructing an augmentation matrix according to the first sample group and the second sample group.
In one implementation, the applying a kernel principal component analysis algorithm to the augmented matrix to obtain a kernel matrix corresponding to the augmented matrix includes:
subtracting a column matrix from a row matrix of the augmented matrix to obtain an intermediate matrix;
calculating the two norms of the intermediate matrix;
and constructing a kernel matrix according to the two norms of the intermediate matrix.
In one implementation, a row matrix of the augmented matrix is subtracted to obtain an intermediate matrix.
In one implementation, the obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix, and the kernel matrix includes:
accumulating each characteristic value to obtain a total accumulated sum;
sequencing each characteristic value in a descending order to obtain a sequence formed by each characteristic value;
sequentially screening the characteristic values from the head of the sequence until the ratio of the local accumulation sum and the total accumulation sum corresponding to the screened characteristic values is greater than a threshold value;
and obtaining a principal component matrix according to the eigenvector and the kernel matrix corresponding to the screened eigenvalue.
In one implementation, the obtaining a detection result of the detected vehicle according to the feature vectors of the principal component matrix and the kernel matrix includes:
obtaining a data fluctuation index of the detected automobile according to each feature vector of the kernel matrix and the kernel matrix, wherein a user of the data fluctuation index represents the data stability of the detected automobile;
and obtaining the detection result of the detected automobile according to the principal component matrix and the data fluctuation index.
In one implementation, the obtaining a detection result of the detected vehicle according to the principal component matrix and the data fluctuation indicator includes:
multiplying the principal component matrix by a transposed matrix of the principal component matrix to obtain a first matrix;
multiplying a matrix formed by each feature vector of the kernel matrix by the kernel matrix to obtain an intermediate matrix;
multiplying the intermediate matrix by a transposed matrix of the intermediate matrix to obtain a second matrix;
subtracting the first matrix from the second matrix to obtain a correlation matrix of the detected automobile, wherein elements in the correlation matrix are used for representing the correlation degree between the data of the detected automobile;
comparing the data fluctuation index with a set fluctuation control limit value to obtain a first comparison result;
comparing each element of the incidence matrix of the detected automobile with a set incidence control limit value respectively to obtain a second comparison result;
and obtaining the detection result of the detected automobile according to the first comparison result and the second comparison result.
In one implementation, the set associated control limit is calculated by:
calculating the average value and the standard deviation corresponding to each element value in the incidence matrix of the fault-free automobile;
and obtaining the associated control limit value according to the average value and the standard deviation.
In a second aspect, an embodiment of the present invention further provides an automobile fault detection apparatus, where the apparatus includes the following components:
the augmented matrix calculation module is used for applying a dynamic principal component analysis algorithm to a current sample matrix to obtain an augmented matrix corresponding to the current sample matrix, and the current sample matrix is formed by current time data of a detected automobile;
the kernel matrix calculation module is used for applying a kernel principal component analysis algorithm to the augmentation matrix to obtain a kernel matrix corresponding to the augmentation matrix;
the principal component matrix calculation module is used for obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix and the kernel matrix;
and the detection module is used for obtaining the detection result of the detected automobile according to the characteristic vectors of the principal component matrix and the kernel matrix.
In a third aspect, an embodiment of the present invention further provides a terminal device, where the terminal device includes a memory, a processor, and an automobile fault detection program that is stored in the memory and is capable of running on the processor, and when the processor executes the automobile fault detection program, the steps of the automobile fault detection method are implemented.
In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, where a vehicle fault detection program is stored on the computer-readable storage medium, and when the vehicle fault detection program is executed by a processor, the steps of the vehicle fault detection method are implemented.
Has the beneficial effects that: the method comprises the steps of firstly collecting data of each moment of a detected automobile in real time, forming a current sample matrix by using the current moment data in the data of each moment, then applying a dynamic principal component analysis algorithm to the current sample matrix to obtain an augmented matrix of the current sample matrix, then applying a kernel principal component analysis algorithm to the augmented matrix to obtain a kernel matrix, then constructing a principal component matrix according to a feature vector and a feature value of the kernel matrix and the kernel matrix, and finally determining the detection result of the automobile under the combined action of the principal component matrix and the feature vector of the kernel matrix.
Because the correlation among all data in the augmentation matrix is small, the data with small correlation is used as the basis for detecting the automobile fault, and the accuracy of the detection result can be improved. In addition, the invention combines the dynamic principal component analysis algorithm and the kernel principal component analysis algorithm, which can improve the detection accuracy for the following reasons: in the actual industrial process, most system processes are nonlinear, the nonlinearity cannot be greatly influenced only in few cases, but the nonlinearity of most systems is not negligible, the nonlinearity is caused by directly adopting a linearization method for processing, a large error is caused, and the kernel principal component analysis can solve the nonlinearity problem of data. Similarly, the dynamics also needs to consider the influence of the time sequence of the process variables on the system, the traditional principal component analysis considers a static model, and the time observation value of one variable is independent of the previous and next time values and is independent of each other. The method does not accord with the characteristics of the actual process industry, the process of the method is slowly changed, dynamic relations exist generally, and the dynamic problem of data can be solved through dynamic principal component analysis. In summary, the invention adopts the combination of two algorithms to improve the detection accuracy. In addition, the data used for judging the detection result is from the detected automobile, the objective data of the automobile data is input into a mathematical model formed by a dynamic principal component analysis algorithm and a kernel principal component analysis algorithm, and the automobile fault can be quantitatively detected according to the data output by the data model, so that the detection accuracy is improved.
Drawings
FIG. 1 is an overall flow diagram of the present invention;
FIG. 2 is a simulated T for the temperature of the engine of the vehicle according to the embodiment of the present invention 2 A schematic view;
FIG. 3 is a schematic diagram of simulated SPE for vehicle engine temperature in an embodiment of the present invention;
FIG. 4 is a simulated T for the temperature of the motor of the vehicle in the embodiment of the present invention 2 A schematic view;
FIG. 5 is a schematic diagram of simulated SPE for the temperature of the automobile motor according to the embodiment of the present invention;
fig. 6 is a schematic block diagram of an internal structure of a terminal device according to an embodiment of the present invention.
Detailed Description
The technical scheme of the invention is clearly and completely described below by combining the embodiment and the attached drawings of the specification. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
Research shows that the automobile fault detection comprises a test logic detection method and a physical detection method. The former has huge content due to the repeated occurrence of a large amount of test logics, and changes corresponding functional logics aiming at different automobiles, and because the functional logics are complex, functional logic errors are easily caused in the changing process, so that the accuracy of detecting automobile faults by using the changed functional logics is reduced. The physical detection method comprises a detection diagnosis method based on human experience, an instrument detection diagnosis method and a self-diagnosis method, the detection diagnosis method based on human experience needs maintenance personnel to position faults according to previous work experience and simple tests, and the accuracy of a detection result is poor due to lack of support of objective data; the instrument detection diagnosis method utilizes parameters, states, curves and waveforms of instrument measurement data to diagnose faults. Maintenance personnel still need to deal with many types of faults that have not been incorporated into the mathematical model; the self-diagnosis method cross-checks a signal level of an Electronic Control Unit (ECU) using a self-diagnosis system of a vehicle, and reference values thereof are stored in a memory. If the level of the signal exceeds the allowable limit, the ECU recognizes the signal as a failure signal and transmits a failure code to the memory. However, the self-diagnosis system of the vehicle has not yet reached a level to ensure an accurate detection rate.
In order to solve the technical problems, the invention provides a method, a device, equipment and a storage medium for detecting automobile faults, and solves the problem that the existing automobile fault detection method is low in accuracy. When the method is specifically implemented, firstly, a dynamic principal component analysis algorithm is applied to a current sample matrix to obtain an augmentation matrix corresponding to the current sample matrix; then applying a kernel principal component analysis algorithm to the augmentation matrix to obtain a kernel matrix corresponding to the augmentation matrix; then, obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix and the kernel matrix; and finally, obtaining the detection result of the detected automobile according to the characteristic vectors of the principal component matrix and the kernel matrix.
For example, the current data of the automobile, such as the engine torque, the engine speed, and the engine coolant temperature at the current time of the automobile, are collected to form a current sample matrix, a dynamic principal component analysis algorithm is applied to the current sample matrix to generate an augmentation matrix, the current sample matrix is one element of the augmentation matrix, other elements of the augmentation matrix are matrices formed by the torque, the speed, and the coolant temperature generated at the time before the current time of the engine, a kernel matrix of the augmentation matrix is calculated and subjected to decentralized processing, an element value of the kernel matrix is a difference between values of elements in the augmentation matrix, and finally, the principal component matrix is obtained according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix, and whether the engine fails is judged according to each element in the principal component matrix.
Exemplary method
The method for detecting the automobile fault can be applied to terminal equipment, and the terminal equipment can be terminal products with a computing function, such as computers and the like. In this embodiment, as shown in fig. 1, the method for detecting a vehicle fault specifically includes the following steps S100, S200, S300, and S400:
s100, applying a dynamic principal component analysis algorithm to a current sample matrix to obtain an augmentation matrix corresponding to the current sample matrix, wherein the current sample matrix is formed by current time data of the detected automobile.
For example, the torque a of the engine of the motor vehicle at the current time t is calculated t Rotational speed v t Cooling liquid temperature w t Torque a at time t-1 before the present time t t-1 Rotational speed v t-1 Cooling liquid temperature w t-1 Torque a at time t-2 t-2 Rotational speed v t-2 Temperature w of coolant t-2 . Torque a t Rotational speed v t Cooling liquid temperature w t Arranged in sequence to form a row matrix x t (ii) a Torque a t-1 Rotational speed v t-1 Cooling liquid temperature w t-1 Arranged in sequence to form a row matrix x t-1 (ii) a Torque a t-2 Rotational speed v t-2 Cooling liquid temperature w t-2 Arranged in sequence to form a row matrix x t-2 。x t-2 、x t-1 、x t Arranged in sequence to form a total sample matrix x = [ x ] t-2 ,x t-1 ,x t ]。
a t 、v t 、w t 、a t-1 、v t-1 、w t-1 、a t-2 、v t-2 、w t-2 Are all values after being subjected to a normalization process, e.g. calculation
Wherein
Is a' t 、a′ t-1 、a′ t-2 σ is a' t 、a′ t-1 、a′ t-2 Standard deviation of the three, a' t 、a′ t-1 、a′ t-2 Are respectively a t 、a t-1 、a t-2 The values before the normalization process. For v t 、w t 、a t-1 、v t-1 、w t-1 、a t-2 、v t-2 、w t-2 The same method as above is used for calculation.
In one embodiment, step S100 comprises the following steps S101, S102, S103, S104, S105, S106:
s101, setting a model order of the dynamic principal component analysis algorithm DPCA.
The model order S of the DPCA is determined to take into account the dynamic relationship between the data at the current time and the data at the past time. The present embodiment takes S =2 from the viewpoint of the time cost of calculation.
And S102, obtaining the number of the matrix columns according to the model order S.
In this embodiment, the number of matrix columns included in the augmented matrix to be configured is equal to S +1.
S103, obtaining the number of matrix rows according to the model order and the total number n of samples, wherein the total number of samples is the number of samples contained in the sample library where the current sample is located.
In this embodiment, the number of matrix rows included in the augmented matrix to be configured is equal to n-S +1.
And S104, sequentially selecting samples before the current sample from the sample library until the number corresponding to the selected samples reaches the matrix column number, wherein the generation time of the data included in the samples before the current sample is before the generation time of the data included in the current sample, and the selected samples are marked as a first sample group.
If the current sample is x t Then the first group of samples comprises samples x t ,x t-1 ,…,x t-S 。
And S105, sequentially selecting samples positioned before the current sample from the sample library until the number corresponding to the selected samples reaches the number of the matrix rows, and marking the selected samples as a second sample group.
If the current sample is x t Then the first group of samples comprises samples x t ,x t-1 ,…,x t+S-n
And S106, constructing an augmentation matrix X (S) according to the first sample group and the second sample group.
S200, applying a kernel principal component analysis algorithm to the augmentation matrix to obtain a kernel matrix corresponding to the augmentation matrix.
Step S200 includes steps S201, S202, S203 as follows:
s201, subtraction is carried out on the row matrix of the augmentation matrix to obtain an intermediate matrix.
And S202, calculating the two norms of the intermediate matrix.
And S203, constructing a kernel matrix K according to the two norms of the intermediate matrix.
The kernel matrix K is the matrix after the matrix K is processed by decentralization, and the element K (i, j) of the ith row and the jth column in the matrix K is:
k (i, j) is a momentThe form of the matrix is taken as the element of the matrix k, X (S) i To augment the ith row of matrix X (S), X (S) j Is the jth row of the augmented matrix X (S), c is a constant, | | | | |2 is a two-norm, X (S) ι -x(S) j Is an intermediate matrix.
Performing decentralized processing on the matrix K to obtain a kernel matrix K:
K＝k-I n k-kI n +I n kI n
I n the representative element is
N × n matrix.
S300, obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix and the kernel matrix.
In one embodiment, step S300 includes steps S301, S302, S303, S304 as follows:
s301, accumulating all the characteristic values to obtain a total accumulated sum.
S302, sequencing each characteristic value in a descending order to obtain a sequence formed by each characteristic value.
S303, sequentially screening the characteristic values from the head of the sequence until the ratio of the local accumulation sum corresponding to the screened characteristic values to the total accumulation sum is greater than a threshold value.
S304, obtaining a principal component matrix according to the eigenvector corresponding to the screened eigenvalue and the kernel matrix.
S301 to S304 form a principal component matrix T according to the following formula A :
T A ＝α A K
λ i Is the ith eigenvalue of the kernel matrix K, s is the total number of eigenvalues,
in the form of a total sum of the sums,
is a local sum of sums, λ j 85% of j-th characteristic value in the sequence obtained by descending order arrangement is a threshold value alpha A Is a matrix of a eigenvectors.
For example, the kernel matrix K has a total of five eigenvalues λ 1 、λ 2 、λ 3 、λ 4 、λ 5 After descending order is λ 5 、λ 2 、λ 4 、λ 1 、λ 3 When lambda is 5 、λ 2 Sum (local cumulative sum) and λ 1 、λ 2 、λ 3 、λ 4 、λ 5 (Total cumulative sum) less than 85%, and when lambda is 5 、λ 2 、λ 4 Sum (local cumulative sum) and λ 1 、λ 2 、λ 3 、λ 4 、λ 5 (the total cumulative sum) is greater than 85%, then the characteristic value (as principal element) is selected as λ 5 、λ 2 、λ 4 The three eigenvectors (in this case, the value of A is 3) of the kernel matrix corresponding to these three eigenvalues constitute α A 。
S400, obtaining the detection result of the detected automobile according to the characteristic vectors of the principal component matrix and the kernel matrix.
In one embodiment, step S400 includes steps S401 to S408 as follows:
s401, obtaining a data fluctuation index T of the detected automobile according to each feature vector alpha of the kernel matrix and the kernel matrix K 2 And the data fluctuation index user represents the data stability of the detected automobile.
T＝αK
T 2 ＝T A Λ -1 T A T
A diagonal matrix of a principal element.
For example, the current samples include engine torque, engine speed, engine coolant temperature, and the matrix T 2 The internal element value corresponding to the engine torque reflects the fluctuation of the engine torque at each moment, and the larger the element value is, the more unstable the engine torque is, and the engine is reflected on the engine, namely the engine fails.
S402, the principal component matrix T A Multiplying by a transpose T of the principal component matrix A T To obtain a first matrix T A T A T 。
And S403, multiplying the matrix alpha formed by the feature vectors of the kernel matrix by the kernel matrix K to obtain an intermediate matrix T = alpha K.
S404, multiplying the intermediate matrix by a transpose matrix T of the intermediate matrix T T Obtaining a second matrix Gamma T 。
S405, subtracting the first matrix from the second matrix to obtain a correlation matrix SPE of the detected automobile, wherein elements in the correlation matrix are used for representing the correlation degree between the data of the detected automobile.
S406, the data fluctuation index T is used 2 With set fluctuation control limit
And comparing to obtain a first comparison result.
In one embodiment, β has a value of 99%.
S407, respectively connecting each element of the SPE of the detected automobile with a set association control limit value
And comparing to obtain a second comparison result.
For example, one element of the SPE matrix represents the degree of correlation between the engine speeds at different times, and when the element is greater than the corresponding speed
When the value is equal, the rotating speed is abnormal, namely the engine is in failure.
In one embodiment, control limits are associated
The calculation of (c) is as follows:
and calculating the average value mSPE and the standard deviation vSPE corresponding to each element value in the incidence matrix of the fault-free automobile.
Obtaining the associated control limit value according to the mean value mSPE and the standard deviation vSPE
Wherein g = vsspe/(2 msspe), χ h ＝(2*mSPE 2 )/vSPE。
For example, data (such as the temperature and the torque of an engine) of an automobile without a fault at each moment is collected, an augmentation matrix, a kernel matrix, a principal component matrix and a correlation matrix corresponding to the automobile without the fault are sequentially calculated, mSPE and vSPE corresponding to the same data variable in a plurality of correlation matrices, such as the temperature mSPE and the temperature vSPE, are solved, and the temperature mSPE and the temperature vSPE are brought into the automobile without the fault
In the calculation formula, the associated control limit value of the temperature is obtained
S408, obtaining the detection result of the detected automobile according to the first comparison result and the second comparison result.
The accuracy of the method for detecting the automobile fault is described by taking XMQ6127AGCHEVN61 hybrid bus as an example:
the sampling period is 1s, 10000 normal sample data are used for off-line training, and 9262 samples are used for on-line testing. A fault is introduced starting with the 1001 st online sample. In fig. 2 to 5, the abscissa represents the sampling time and the ordinate represents T 2 Or SPE value. Straight line represents T 2 Or the control limit value corresponding to SPE, the curve representing the T of the detected car 2 Or SPE value.
Fig. 2 and 3 are fault detection diagrams at the time of engine over-temperature. It can be seen that starting from the 1001 st sample in fig. 2 and 3, the curve starts to be higher than a straight line, indicating that the method of the present invention successfully detected a fault.
Fig. 4 and 5 are fault detection diagrams when the motor is over-temperature. It can be seen that starting from the 1001 st sample in fig. 4 and 5, the curve line where at least the SPE statistic is located starts to be higher than the straight line, indicating that the method of the present invention successfully detects a fault.
In conclusion, the data with small correlation in the augmentation matrix of the invention is used as the basis for detecting the automobile fault, so that the accuracy of the detection result can be improved. In addition, the invention combines the dynamic principal component analysis algorithm and the kernel principal component analysis algorithm. In addition, the data used for judging the detection result is from the detected automobile, the objective data of the automobile data is input into a mathematical model formed by a dynamic principal component analysis algorithm and a kernel principal component analysis algorithm, and the automobile fault can be quantitatively detected according to the data output by the data model, so that the detection accuracy is improved.
The method is also suitable for fault detection of the hybrid electric vehicle, adopts a principal component analysis algorithm in data driving, establishes an accurate data model, and optimizes timeliness and accuracy of fault detection compared with an automobile self-diagnosis system based on artificial experience and non-real time.
When the detected automobile is a hybrid automobile, the dynamic problem and the nonlinear problem of the data of the hybrid automobile are considered, and the accuracy of fault detection is improved by adopting a dynamic kernel principal component analysis algorithm.
Exemplary devices
The embodiment also provides a vehicle fault detection device, which comprises the following components:
the augmented matrix calculation module is used for applying a dynamic principal component analysis algorithm to a current sample matrix to obtain an augmented matrix corresponding to the current sample matrix, and the current sample matrix is formed by current time data of a detected automobile;
the kernel matrix calculation module is used for applying a kernel principal component analysis algorithm to the augmented matrix to obtain a kernel matrix corresponding to the augmented matrix;
the principal component matrix calculation module is used for obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix and the kernel matrix;
and the detection module is used for obtaining the detection result of the detected automobile according to the characteristic vectors of the principal component matrix and the kernel matrix.
Based on the above embodiment, the present invention further provides a terminal device, and a schematic block diagram thereof may be as shown in fig. 6. The terminal equipment comprises a processor, a memory, a network interface, a display screen and a temperature sensor which are connected through a system bus. Wherein the processor of the terminal device is configured to provide computing and control capabilities. The memory of the terminal equipment comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the terminal device is used for connecting and communicating with an external terminal through a network. The computer program is executed by a processor to implement a method of vehicle fault detection. The display screen of the terminal equipment can be a liquid crystal display screen or an electronic ink display screen, and the temperature sensor of the terminal equipment is arranged in the terminal equipment in advance and used for detecting the operating temperature of the internal equipment.
It will be understood by those skilled in the art that the block diagram of fig. 6 is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation to the terminal device to which the solution of the present invention is applied, and a specific terminal device may include more or less components than those shown in the figure, or may combine some components, or have different arrangements of components.
In one embodiment, a terminal device is provided, where the terminal device includes a memory, a processor, and a vehicle failure detection program stored in the memory and executable on the processor, and when the processor executes the vehicle failure detection program, the following operation instructions are implemented:
applying a dynamic principal component analysis algorithm to a current sample matrix to obtain an augmentation matrix corresponding to the current sample matrix, wherein the current sample matrix is formed by current time data of a detected automobile;
applying a kernel principal component analysis algorithm to the augmentation matrix to obtain a kernel matrix corresponding to the augmentation matrix;
obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix and the kernel matrix;
and obtaining the detection result of the detected automobile according to the characteristic vectors of the principal component matrix and the kernel matrix.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, databases, or other media used in embodiments provided herein may include non-volatile and/or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), rambus (Rambus) direct RAM (RDRAM), direct Rambus Dynamic RAM (DRDRAM), and Rambus Dynamic RAM (RDRAM), among others.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for detecting a fault of an automobile, comprising:
applying a dynamic principal component analysis algorithm to a current sample matrix to obtain an augmentation matrix corresponding to the current sample matrix, wherein the current sample matrix is formed by current time data of a detected automobile;
applying a kernel principal component analysis algorithm to the augmentation matrix to obtain a kernel matrix corresponding to the augmentation matrix;
obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix and the kernel matrix;
and obtaining the detection result of the detected automobile according to the characteristic vectors of the principal component matrix and the kernel matrix.
2. The method according to claim 1, wherein the applying a dynamic principal component analysis algorithm to the current sample matrix to obtain an augmentation matrix corresponding to the current sample matrix, the current sample matrix being formed by current time data of the detected vehicle, includes:
setting a model order of the dynamic principal component analysis algorithm;
obtaining the number of matrix columns according to the model order;
obtaining the number of matrix rows according to the model order and the total number of samples, wherein the total number of samples is the number of samples contained in a sample library where the current sample is located;
sequentially selecting samples positioned before the current sample from the sample library until the number corresponding to the selected samples reaches the matrix column number, wherein the generation time of data included in the samples positioned before the current sample is positioned before the generation time of the data included in the current sample, and the selected samples are marked as a first sample group;
sequentially selecting samples positioned before the current sample from the sample library until the number corresponding to the selected samples reaches the number of the matrix rows, and marking the selected samples as a second sample group;
and constructing an augmentation matrix according to the first sample group and the second sample group.
3. The method according to claim 1, wherein the applying a kernel principal component analysis algorithm to the augmented matrix to obtain a kernel matrix corresponding to the augmented matrix comprises:
carrying out subtraction operation on the row matrix of the augmentation matrix to obtain an intermediate matrix;
calculating a two-norm of the intermediate matrix;
and constructing a kernel matrix according to the two norms of the intermediate matrix.
4. The method for detecting vehicle failure according to claim 1, wherein the obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix, and the kernel matrix comprises:
accumulating each characteristic value to obtain a total accumulated sum;
sequencing each characteristic value in a descending order to obtain a sequence formed by each characteristic value;
sequentially screening the characteristic values from the head of the sequence until the ratio of the local accumulation sum and the total accumulation sum corresponding to the screened characteristic values is greater than a threshold value;
and obtaining a principal component matrix according to the eigenvector and the kernel matrix corresponding to the screened eigenvalue.
5. The method for detecting vehicle failure according to claim 1, wherein the obtaining the detection result of the detected vehicle according to the feature vectors of the principal component matrix and the kernel matrix comprises:
obtaining a data fluctuation index of the detected automobile according to each feature vector of the kernel matrix and the kernel matrix, wherein a user of the data fluctuation index represents the data stability of the detected automobile;
and obtaining a detection result of the detected automobile according to the principal component matrix and the data fluctuation index.
6. The method according to claim 5, wherein obtaining the detection result of the detected vehicle according to the principal component matrix and the data fluctuation indicator comprises:
multiplying the principal component matrix by a transposed matrix of the principal component matrix to obtain a first matrix;
multiplying a matrix formed by each feature vector of the kernel matrix by the kernel matrix to obtain an intermediate matrix;
multiplying the intermediate matrix by a transposed matrix of the intermediate matrix to obtain a second matrix;
subtracting the first matrix from the second matrix to obtain a correlation matrix of the detected automobile, wherein elements in the correlation matrix are used for representing the degree of correlation between the data of the detected automobile;
comparing the data fluctuation index with a set fluctuation control limit value to obtain a first comparison result;
comparing each element of the incidence matrix of the detected automobile with a set incidence control limit value respectively to obtain a second comparison result;
and obtaining the detection result of the detected automobile according to the first comparison result and the second comparison result.
7. The method of claim 6, wherein the set associated control limit is calculated by:
calculating the average value and the standard deviation corresponding to each element value in the incidence matrix of the fault-free automobile;
and obtaining the associated control limit value according to the average value and the standard deviation.
8. An automobile fault detection device, characterized in that the device comprises the following components:
the augmented matrix calculation module is used for applying a dynamic principal component analysis algorithm to a current sample matrix to obtain an augmented matrix corresponding to the current sample matrix, and the current sample matrix is formed by current time data of a detected automobile;
the kernel matrix calculation module is used for applying a kernel principal component analysis algorithm to the augmentation matrix to obtain a kernel matrix corresponding to the augmentation matrix;
the principal component matrix calculation module is used for obtaining a principal component matrix according to each eigenvector of the kernel matrix, each eigenvalue of the kernel matrix and the kernel matrix;
and the detection module is used for obtaining the detection result of the detected automobile according to the characteristic vectors of the principal component matrix and the kernel matrix.
9. A terminal device, characterized in that the terminal device comprises a memory, a processor and a vehicle failure detection program stored in the memory and operable on the processor, and the processor implements the steps of the vehicle failure detection method according to any one of claims 1 to 7 when executing the vehicle failure detection program.
10. A computer-readable storage medium, having a vehicle fault detection program stored thereon, which, when executed by a processor, implements the steps of the vehicle fault detection method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211395160.1A CN115683661A (en) | 2022-11-08 | 2022-11-08 | Automobile fault detection method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211395160.1A CN115683661A (en) | 2022-11-08 | 2022-11-08 | Automobile fault detection method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115683661A true CN115683661A (en) | 2023-02-03 |
Family
ID=85050522
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211395160.1A Pending CN115683661A (en) | 2022-11-08 | 2022-11-08 | Automobile fault detection method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115683661A (en) |
-
2022
- 2022-11-08 CN CN202211395160.1A patent/CN115683661A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10408707B2 (en) | Abnormality diagnosing method and abnormality diagnosing system | |
| CN111104736B (en) | Abnormal data detection method, device, medium and equipment based on time sequence | |
| Harkat et al. | Sensor fault detection, isolation and reconstruction using nonlinear principal component analysis | |
| CN111325159A (en) | Fault diagnosis method and device, computer equipment and storage medium | |
| US20210357727A1 (en) | Abnormality detection device | |
| CN112577515A (en) | Inertial sensor, method for self-calibration of inertial sensor, and computer-readable medium | |
| CN113360304A (en) | Method and computing unit for causal analysis of an abnormal state of a machine | |
| CN114553681A (en) | Equipment state abnormity detection method and device and computer equipment | |
| CN114511025A (en) | Fan fault diagnosis method and device based on weighted multi-sensor fusion filtering | |
| CN109990803B (en) | Method and device for detecting system abnormity and method and device for sensor processing | |
| CN113110961B (en) | Equipment abnormality detection method and device, computer equipment and readable storage medium | |
| CN116627116B (en) | Process industry fault positioning method and system and electronic equipment | |
| CN111382029B (en) | Mainboard abnormity diagnosis method and device based on PCA and multidimensional monitoring data | |
| CN115683661A (en) | Automobile fault detection method, device, equipment and storage medium | |
| CN116907772A (en) | Self-diagnosis and fault source identification method and system of bridge structure monitoring sensor | |
| CN111474476A (en) | Motor fault prediction method | |
| CN115940806A (en) | Photovoltaic power generation system and fault diagnosis method thereof | |
| CN114814707A (en) | Intelligent ammeter stress error analysis method, equipment, terminal and readable medium | |
| CN112231849B (en) | Axle temperature fault detection method based on NEST and SPRT fusion algorithm | |
| CN114877926A (en) | Sensor fault detection and diagnosis method, medium, electronic device and system | |
| Zhu et al. | Multiplicative fault detection and isolation in dynamic systems using data-driven k-Gap metric based kNN Algorithm | |
| CN114192583A (en) | Scada platform-based strip steel rolling process quality monitoring method and system | |
| CN114112374A (en) | Fault diagnosis method of weapon equipment system based on orthogonal sub-block projection | |
| CN117744390A (en) | Flight load inversion method and system based on strain measurement and electronic equipment | |
| CN116663728A (en) | Electric energy metering device error state prediction method and device |
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 |