CN109203995A - A kind of trailer system major loop ground fault diagnosis method and system - Google Patents
A kind of trailer system major loop ground fault diagnosis method and system Download PDFInfo
- Publication number
- CN109203995A CN109203995A CN201810752527.8A CN201810752527A CN109203995A CN 109203995 A CN109203995 A CN 109203995A CN 201810752527 A CN201810752527 A CN 201810752527A CN 109203995 A CN109203995 A CN 109203995A
- Authority
- CN
- China
- Prior art keywords
- residual
- data
- residual signals
- real time
- statistic
- 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
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L3/00—Electric devices on electrically-propelled vehicles for safety purposes; Monitoring operating variables, e.g. speed, deceleration or energy consumption
- B60L3/0023—Detecting, eliminating, remedying or compensating for drive train abnormalities, e.g. failures within the drive train
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L3/00—Electric devices on electrically-propelled vehicles for safety purposes; Monitoring operating variables, e.g. speed, deceleration or energy consumption
- B60L3/12—Recording operating variables ; Monitoring of operating variables
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L2200/00—Type of vehicles
- B60L2200/26—Rail vehicles
Abstract
The present invention relates to trailer system fault diagnosis fields, disclose a kind of trailer system major loop ground fault diagnosis method and system, to consider the correlation in major loop between variable, quickly and accurately judge whether that ground fault occurs；This method includes obtaining measurable variable related with DC link voltage, and measurable variable is grouped, and chooses the data matrix of the corresponding history normal data composition respective sets number of every group of measurable variable, and establishes residual generation；The residual signals of history normal data are calculated, the residual signals statistic of history normal data is then calculated according to residual signals, then setting control limit；Calculate the residual signals of real time data, then the residual signals statistic of real time data is calculated according to the residual signals of real time data, by the residual signals statistic of real time data and control limit be compared, when real time data residual signals statistic be more than control prescribe a time limit, be considered as generation ground fault；Conversely, being considered as normal operation.
Description
Technical field
The present invention relates to trailer system fault diagnosis field more particularly to a kind of trailer system major loop ground fault diagnosises
Method and system.
Background technique
With the fast development of railway electrification, trailer system is had become in modern railway transportation train, such as electric power machine
Vehicle and bullet train, an indispensable important subsystem.Compared with traditional trailer system, electric propulsion have high energy efficiency,
The advantages that low emission and economic operation, typical trailer system major loop is by pre-charge circuit, rectifier, intermediate dc ring
Section, inverter and traction electric machine composition.Due to train operation environment complexity and badly, such as abrasion, aging, high temperature and electrostatic,
Trailer system is easy to that ground fault occurs, and ground fault will lead to big short circuit current, easily damage critical component, shadow
It rings train to operate normally, or even the personal safety of meeting crisis passenger.
In order to detect the generation of major loop ground failure, it often will be installed earthing detection hardware circuit in system and (also referred to as connect
Ground measure loop), which is mainly to pass through the voltage of sensor measurement intermediate DC link, then by the voltage of measurement with according to
Realization fault detection is compared according to the threshold value that experience is arranged.Although the circuit is able to detect the generation of some ground faults,
Be such method threshold value setting often have conservative, also, when some small faults generation insufficient for voltage
When more than threshold value, cannot effectively it be detected；Meanwhile such method can not also the generation to failure be effectively isolated.
Summary of the invention
It is an object of that present invention to provide a kind of trailer system major loop ground fault diagnosis method and system, to consider to lead back
Correlation in road between variable quickly and accurately judges whether that ground fault occurs.
To achieve the above object, the present invention provides a kind of trailer system major loop ground method for diagnosing faults, including with
Lower step:
S1: obtaining measurable variable related with DC link voltage, the measurable variable be grouped, and chooses every
The data matrix of the corresponding history normal data composition respective sets number of group measurable variable, comprehensive all data matrixes are established residual
Difference generator；
S2: using the history normal data as the input of the residual generation, the history normal data is calculated
Then residual signals calculate the residual signals statistic of the history normal data according to the residual signals, and according to described
Residual signals statistic setting control limit；
S3: the corresponding real time data of every group of measurable variable is obtained, is occurred the real time data as the residual error
The input of device calculates the residual signals of the real time data, then calculates the reality according to the residual signals of the real time data
When data residual signals statistic, by the residual signals statistic of the real time data and the control limit be compared, when
The residual signals statistic of the real time data is more than that the control is prescribed a time limit, and is considered as generation ground fault；Conversely, being considered as normal fortune
Row.
Preferably, when being judged as generation ground fault, the method also includes:
S4: the corresponding historical failure data of every group of measurable variable is chosen, using the historical failure data as the residual error
The input of generator calculates the historical failure data residual signals, is set according to the historical failure data residual signals residual
Poor direction, and establish residual error direction library；
S5: the residual signals of real time data are compared with each of residual error direction library residual error direction, really
Determine fault type.
Preferably, in the S4, residual error direction, including following step are set according to the historical failure data residual signals
It is rapid:
S41: fault residual matrix is established according to the historical failure data and the historical failure residual signals；
S42: carrying out unusual decomposition to the fault residual matrix, the set of eigenvectors of corresponding fault type is obtained, by institute
State the residual error direction that feature vector concentrates first feature vector direction to be set as corresponding failure.
Preferably, in the S5, by each of the residual signals of real time data and the residual error direction library residual error side
To being compared, fault type is determined, comprising the following steps:
The angle in the residual signals and residual error direction library of the real time data between each residual error direction is calculated,
The obtained corresponding residual error direction of minimum angle is chosen, which is earth fault type.
Preferably, measurable variable described in the S1 includes voltage, voltage on line side, the rectifier input of intermediate DC link
Electric current and current on line side, wherein the voltage of the intermediate DC link, the voltage on line side are voltage quantities；The rectifier
Input current, the current on line side are current variable.
Preferably, the S1 specifically includes the following steps:
S11: obtaining measurable variable is the voltage quantities of intermediate DC link, voltage on line side variable, rectifier input current
With current on line side variable, the history normal data of each corresponding variable is chosen, by the history normal data of all voltage quantities
It is divided into the first data set, the history normal data of all current variables is divided into the second data set；
S12: the first covariance matrix of first data set, the second covariance square of second data set are calculated
Battle array and the Cross-covariance between first data set and second data set, and according to first covariance
Matrix, second covariance matrix and the Cross-covariance establish first data set and second data set
Between correlation matrix；
S13: residual generation, the calculation formula of the residual generation are established according to the correlation matrix are as follows:
In formula,Indicate the first data set,Indicate that the second data set, T indicate that the transposed operator of correlation matrix, Σ indicate
The correlation matrix is carried out to the diagonal matrix obtained after unusual decomposition, L and J indicate canonical correlation vector, wherein Γ indicates the left singular vector of the correlation matrix, R indicate the right side of the correlation matrix it is unusual to
Amount,Indicate minus 1/2nd power of second correlation matrix,Indicate minus 1/the 2 of first correlation matrix
Power.
Preferably, the step S2 specifically includes the following steps:
S21: residual signals statistic is established according to the residual generation are as follows:
In formula,Indicate the inverse matrix of the covariance matrix of corresponding residual signals；
S22: calculating the residual signals statistic of history normal data, then calculates the residual error using Monte Carlo method
The corresponding threshold value of signal statistics amount determines control limit.
As a general technical idea, the present invention also provides a kind of trailer system major loop ground fault diagnosis system,
Including memory, processor and store the computer program that can be run on a memory and on a processor, the processor
The step of realizing the above method when executing the computer program.
The invention has the following advantages:
Trailer system major loop ground fault diagnosis method and system provided by the invention, first in analysis system it is all with
Correlation between the related measurable variable of DC link voltage；It is then based on the correlation and establishes residual generation, it will
The corresponding history normal data of all measurable variables obtains the residual error letter of history normal data as the input of the residual generation
Number, and according to the residual signals statistic of residual signals calculating history normal data, further according to the system of the residual signals
Metering setting control limit；Input of the real time data for all measurable variables that will acquire again as the residual generation, calculates
To the residual signals of real time data, the residual signals statistic of real time data is calculated according to the residual signals of real time data, it will be real
When data residual signals statistic and the control limit be compared, when the residual signals statistic of real time data is more than the control
In limited time, it is considered as generation ground fault；Conversely, being considered as normal operation；Method of the invention has fully considered that the master of trailer system returns
Correlation in road between variable can quickly and accurately judge whether that ground fault occurs.
Below with reference to accompanying drawings, the present invention is described in further detail.
Detailed description of the invention
The attached drawing constituted part of this application is used to provide further understanding of the present invention, schematic reality of the invention
It applies example and its explanation is used to explain the present invention, do not constitute improper limitations of the present invention.In the accompanying drawings:
Fig. 1 is the trailer system major loop ground method for diagnosing faults flow chart of the preferred embodiment of the present invention；
Fig. 2 is certain type locomotive traction system major loop of the preferred embodiment of the present invention；
Fig. 3 is that the trailer system major loop ground fault diagnosis result of the preferred embodiment of the present invention and conventional detection circuitry are examined
Result is surveyed to the testing result comparison diagram of failure F1；
Fig. 4 is that the trailer system major loop ground fault diagnosis result of the preferred embodiment of the present invention and conventional detection circuitry are examined
Result is surveyed to the testing result comparison diagram of failure F3.
Specific embodiment
The embodiment of the present invention is described in detail below in conjunction with attached drawing, but the present invention can be defined by the claims
Implement with the multitude of different ways of covering.
Embodiment 1
Referring to Fig. 1, the present embodiment provides a kind of trailer system major loop ground method for diagnosing faults, comprising the following steps:
S1: measurable variable related with DC link voltage is obtained, measurable variable is grouped, choosing every group can
The data matrix of the corresponding history normal data composition respective sets number of variable is surveyed, comprehensive all data matrixes establish residual error
Device；
S2: using history normal data as the input of residual generation, the residual signals of history normal data are calculated, then
The residual signals statistic of history normal data is calculated according to residual signals, and control limit is set according to residual signals statistic；
S3: obtaining the corresponding real time data of every group of measurable variable, using real time data as the input of residual generation, calculates
Then the residual signals of real time data calculate the residual signals statistic of real time data according to the residual signals of real time data, will
The residual signals statistic of real time data is compared with control limit, when the residual signals statistic of real time data is more than control limit
When, it is considered as generation ground fault；Conversely, being considered as normal operation.
Through the above steps, the correlation in the major loop of trailer system between variable can be fully considered, quickly and accurately
Judge whether the major loop occurs ground fault, it is convenient and practical without increasing new measuring device.
Further, when being judged as generation ground fault, method provided by the invention is further comprising the steps of:
S4: the corresponding historical failure data of every group of measurable variable is chosen, using historical failure data as residual generation
Input calculates historical failure data residual signals, sets residual error direction according to historical failure data residual signals, and establish residual error
Direction library；
S5: the residual signals of real time data are compared with each of residual error direction library residual error direction, determine event
Hinder type.
In summary step S4 and step S5 can accurately obtain failure on the basis of judging whether to occur ground fault
Type realizes Fault Isolation, provides more accurate testing result for real work.
It should be noted that measurable variable includes voltage, voltage on line side, rectifier the input electricity of intermediate DC link in S1
Stream and current on line side, wherein voltage, the voltage on line side of intermediate DC link are voltage quantities；Rectifier input current, net side electricity
Stream is current variable.
In practical application, in above-mentioned steps, trailer system major loop ground method for diagnosing faults of the invention can be with
It optimizes, the embodiment after optimization is as follows:
S1: measurable variable related with DC link voltage is obtained, measurable variable is grouped, choosing every group can
The data matrix of the corresponding history normal data composition respective sets number of variable is surveyed, comprehensive all data matrixes establish residual error
Device, specific as follows:
S11: obtaining measurable variable is the voltage quantities of intermediate DC link, voltage on line side variable, rectifier input current
With current on line side variable, the history normal data of each corresponding variable is chosen, by the history normal data of all voltage quantities
It is divided into the first data set, the history normal data of all current variables is divided into the second data set.Wherein, the first data set
Are as follows:
In formula, YaIndicate the corresponding history normal data of voltage quantities, l indicates the history normal data number obtained, N table
Show that independent sample is counted；Then ya(i), i=1,2 ..., N indicates vector yaMeasured value of the middle l variable at the ith sample moment,
Second data set are as follows:
In formula, YbIndicate the corresponding history normal data of current variable, m indicates the history normal data number obtained, yb
(i), i=1,2 ..., N indicates vector ybMeasured value of the middle l variable at the ith sample moment.
Then, average value processing is carried out to the first data set and the second data set, obtains new data set are as follows:
S12: calculate the covariance matrix of the first data set, the covariance matrix of the second data set and the first data set and
Cross-covariance between second data set, and according to the first covariance matrix, the second covariance matrix and cross covariance
Matrix establishes the correlation matrix between the first data set and the second data set.
Specifically, data set is calculatedCovariance matrix matrix are as follows:
In formula, T indicates that the transposed operator of correlation matrix, N indicate the total sample number in data set；
Calculate data setCovariance matrix matrix are as follows:
Calculate data setWith data setCross-covariance matrix are as follows:
Further establish the correlation matrix Ω between the first data set and the second data set are as follows:
Then, singular value decomposition is carried out to above-mentioned correlation matrix are as follows:
Ω=Γ Σ RT；
In formula, Γ indicates that the left singular vector of correlation matrix, R indicate the right singular vector of correlation matrix, and Σ indicates diagonal
Matrix.
S13: according to correlation matrix, establishing residual generation using Canonical Correlation Analysis, wherein residual generation
Calculation formula are as follows:
In formula,Indicate the first data set,Indicate that the second data set, T indicate that the transposed operator of correlation matrix, Σ indicate
Correlation matrix is carried out to the diagonal matrix obtained after unusual decomposition, L and J indicate canonical correlation vector, wherein Γ indicates that the left singular vector of correlation matrix, R indicate the right singular vector of correlation matrix,Indicate the second phase
Minus 1/2nd power of matrix is closed,Indicate minus 1/2nd power of the first correlation matrix.
S2: using history normal data as the input of residual generation, the residual signals of history normal data and residual are calculated
Difference signal statistic sets control limit according to the residual signals statistic of history normal data, specific as follows:
S21: residual signals statistic is established according to residual generation are as follows:
In formula,Indicate the inverse matrix of the covariance matrix of corresponding residual signals；Specifically, covariance matrix ΣrMeter
Calculate formula are as follows:
Σr=I- Σ ΣT；
In formula, I indicates the unit matrix of covariance matrix.
S22: calculating the residual signals statistic of history normal data, then calculates residual signals using Monte Carlo method
The corresponding threshold value of statistic determines control limit.
Wherein, the specifically used step of above-mentioned Monte Carlo method in the present embodiment are as follows:
The simulation times of Monte Carlo method are set first as K, set initial threshold asEnable current iteration number s=
0；
The data set obtained when choosing the s times iteration are as follows:
The statistic of residual generation is denoted as Js, counting function is denoted asThe then value of the technology function are as follows:
Setting estimates the formula of rate of false alarm value after K Monte-Carlo Simulation are as follows:
pFARIndicate specified maximum allowable rate of false alarm, ε is a constant, and ε > 0；
Wherein, ifEstablishment then enablesAnd return to present threshold valueFor control limit.If not,
It enablesAnd the value of count function again, wherein Δ is iteration tolerance.
S3: obtaining the corresponding real time data of every group of measurable variable, using real time data as the input of residual generation, calculates
The residual signals and residual signals statistic of real time data compare the residual signals statistic of real time data and control limit
Compared with, when real time data residual signals statistic be more than control prescribe a time limit, be considered as generation ground fault；Conversely, being considered as no ground event
Barrier.
Specifically, for obtaining the real time data of practical first sampling instant, the specific steps are as follows:
The measurement data set of collected 1st sampling instant is denoted asThen it is gone
Average value processing obtains data setWithBy data setWithAs the input of residual generation, calculate real-time
The residual signals of data are as follows:
Further calculate the statistic of the residual signals of real time data are as follows:
By T2(1) the control limit determined with above-mentioned steps S22 is compared, if T2It (1) is more than control limit, then it is assumed that occur
Ground fault, if T2(1) it is no more than control limit, then it is assumed that ya(1) and ybIt (1) is normal data.
S4: the corresponding historical failure data of every group of measurable variable is chosen, using historical failure data as residual generation
Input calculates historical failure data residual signals, sets residual error direction according to historical failure data residual signals, and establish residual error
Direction library:
S41: fault residual matrix is established according to the historical failure data and the historical failure residual signals；
S42: carrying out unusual decomposition to the fault residual matrix, the set of eigenvectors of corresponding fault type is obtained, by institute
State the residual error direction that feature vector concentrates first feature vector direction to be set as corresponding failure.
S5: the residual signals of real time data are compared with each of residual error direction library residual error direction, determine event
Hinder type, specifically: the angle in the residual signals and residual error direction library of real time data between each residual error direction is calculated, is selected
The corresponding residual error direction of the minimum angle obtained, the corresponding fault verification in residual error direction are earth fault type.
As the present embodiment preferred embodiment, the present embodiment is returned with certain type locomotive traction system master as shown in Figure 2
For road, method of the invention is further described and is verified.It should be noted that the type locomotive traction system mainly by
Tractive transformer, traction convertor (including it is charge circuit, four-quadrant rectifier, earthing detection circuit, intermediate DC link, inverse
Become device etc.) and traction electric machine three parts composition.In actual operation, there is single-phase AC 25kV alternating current disconnected by pantograph, master
Road device VCB and tractive transformer primary side enter car body, provide alternating current from tractive transformer secondary side winding to convertor circuit.Exchange
Electric current is transformed into direct current under the action of four-quadrant rectifier, after intermediate DC link filters, is converted into using inverter
The three-phase alternating current electric drive traction electric machine that frequency and amplitude can be changed is advanced to control locomotive with friction speed and tractive force.
This example considers that five kinds of common trailer system major loop ground fault points are as shown in table 1:
The common major loop ground fault point of 1 trailer system of table
The present embodiment carries out verification experimental verification using the data set acquired from locomotive traction system onboard sensor to illustrate this
The feasibility and validity of invention.
Firstly, choosing 6 onboard sensor measurement results, according to requiring, voltage on line side is denoted as Un, then two centres are straight
Flow link voltage Us1And Us2Constitute vector ya；By rectifier input current Ir1And Ir2And current on line side InConstitute ybVector.Number
It is divided into four parts: training dataset (fault-free), five class ground fault data sets, threshold learning data set and test number according to collection
According to collection.Sampling period is 100 μ s, Monte-Carlo Simulation number K=1000, and training dataset is made of 2000 samples, five classes
The fault sample number of ground fault data set is 1000, each is made of threshold learning data set 3000 samples, tests number
It is made of according to collection 2000 samples, and failure generation initial time is chosen for the 1046th sample point.
In actual operation, the testing result difference for choosing ground fault F1 and F3 is as shown in Figure 3 and Figure 4, wherein detecting
Index JeCorresponding method is traditional method based on earthing detection circuit, Testing index JT2Corresponding method is of the invention leads
Draw main loop earth fault diagnostic method (CCA).The two methods refer in two performances of fault detection rate and detection delay
Mark is compared as follows shown in table 2:
2 Earth Fault Detection result of table
Shown in table 2 as above, it can be seen that the method for the present invention to the feasibility and validity of ground fault on-line checking, and
Compared with traditional earthing detection loop method, the fault detection rate of the method for the present invention is higher and detection delay is lower, has more preferable
Detection performance.
After ground fault occurs for detection, further progress Fault Isolation.By current residue signal and residual error direction library
In each residual error direction be compared one by one, the corresponding failure in minimum angle direction be possible occur fault type, tool
Body calculates
Corresponding different ground fault, according to above formula, if current failure residual error direction is corresponding with fault direction library
A certain fault direction angle is minimum, shows that broken down type is failure f, that is, realizes Fault Isolation.Wherein, difference connects
The isolation result of earth fault is as shown in table 3 below, and being correctly isolated rate (CIR) with failure is that quantizating index carries out isolation performance evaluation.
The isolation result Fault Isolation result of the different ground faults of table 3
Shown in table 3 as above, specific fault type can be effectively isolated in the method for the present invention.
Embodiment 2
With above method embodiment correspondingly, the present embodiment provides a kind of trailer system major loop ground fault diagnosis systems
System including memory, processor and stores the computer program that can be run on a memory and on a processor, the processing
The step of device realizes the above method when executing the computer program.
As described above, trailer system major loop ground fault diagnosis method and system provided by the invention, first analysis system
Correlation in system between all measurable variables related with DC link voltage；It is then based on the correlation and establishes residual error
Generator obtains history normal data using the corresponding history normal data of all measurable variables as the input of the residual generation
Residual signals, and according to the residual signals calculate history normal data residual signals statistic, further according to the residual error
The statistic setting control limit of signal；The real time data for all measurable variables that will acquire again is as the defeated of the residual generation
Enter, the residual signals of real time data are calculated, is united according to the residual signals that the residual signals of real time data calculate real time data
The residual signals statistic of real time data and control limit are compared, when the residual signals statistic of real time data by metering
It prescribes a time limit more than the control, is considered as generation ground fault；Conversely, being considered as normal operation；Method of the invention has fully considered traction
Correlation in the major loop of system between variable can quickly and accurately judge whether that ground fault occurs.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (8)
1. a kind of trailer system major loop ground method for diagnosing faults, which comprises the following steps:
S1: measurable variable related with DC link voltage is obtained, the measurable variable is grouped, choosing every group can
The data matrix of the corresponding history normal data composition respective sets number of variable is surveyed, comprehensive all data matrixes establish residual error hair
Raw device；
S2: using the history normal data as the input of the residual generation, the residual error of the history normal data is calculated
Then signal calculates the residual signals statistic of the history normal data according to the residual signals, and according to the residual error
Signal statistics amount setting control limit；
S3: the corresponding real time data of every group of measurable variable is obtained, using the real time data as the residual generation
Input, calculates the residual signals of the real time data, then calculates the real-time number according to the residual signals of the real time data
According to residual signals statistic, by the residual signals statistic of the real time data and the control limit be compared, when described
The residual signals statistic of real time data is more than that the control is prescribed a time limit, and is considered as generation ground fault；Conversely, being considered as normal operation.
2. trailer system major loop ground method for diagnosing faults according to claim 1, which is characterized in that sent out when being judged as
When raw ground fault, the method also includes:
S4: the corresponding historical failure data of every group of measurable variable is chosen, is occurred the historical failure data as the residual error
The input of device calculates the historical failure data residual signals, sets residual error side according to the historical failure data residual signals
To, and establish residual error direction library；
S5: the residual signals of real time data are compared with each of residual error direction library residual error direction, determine event
Hinder type.
3. trailer system major loop ground method for diagnosing faults according to claim 2, which is characterized in that in the S4,
Residual error direction is set according to the historical failure data residual signals, comprising the following steps:
S41: fault residual matrix is established according to the historical failure data and the historical failure residual signals；
S42: carrying out unusual decomposition to the fault residual matrix, the set of eigenvectors of corresponding fault type is obtained, by the spy
First feature vector direction is set as the residual error direction of corresponding failure in sign vector set.
4. trailer system major loop ground method for diagnosing faults according to claim 2, which is characterized in that
In the S5, the residual signals of real time data are compared with each of residual error direction library residual error direction,
Determine fault type, comprising the following steps:
The angle in the residual signals and residual error direction library of the real time data between each residual error direction is calculated, is chosen
The corresponding residual error direction of obtained minimum angle, the corresponding fault verification in residual error direction are earth fault type.
5. trailer system major loop ground method for diagnosing faults according to claim 3 or 4, which is characterized in that the S1
Described in measurable variable include intermediate DC link voltage, voltage on line side, rectifier input current and current on line side, wherein
The voltage of the intermediate DC link, the voltage on line side are voltage quantities；The rectifier input current, the current on line side
For current variable.
6. trailer system major loop ground method for diagnosing faults according to claim 5, which is characterized in that the S1 is specific
The following steps are included:
S11: voltage quantities, voltage on line side variable, rectifier input current and the net that measurable variable is intermediate DC link are obtained
Side current variable chooses the history normal data of each corresponding variable, the history normal data of all voltage quantities is divided
For the first data set, the history normal data of all current variables is divided into the second data set；
S12: calculate the first covariance matrix of first data set, the second covariance matrix of second data set, with
And the Cross-covariance between first data set and second data set, and according to first covariance matrix,
Second covariance matrix and the Cross-covariance are established between first data set and second data set
Correlation matrix；
S13: residual generation, the calculation formula of the residual generation are established according to the correlation matrix are as follows:
In formula,Indicate the first data set,Indicate that the second data set, T indicate that the transposed operator of correlation matrix, Σ are indicated institute
It states correlation matrix and carries out the diagonal matrix obtained after unusual decomposition, L and J indicate canonical correlation vector, whereinΓ indicates that the left singular vector of the correlation matrix, R indicate that the right side of the correlation matrix is unusual
Vector,Indicate minus 1/2nd power of second correlation matrix,Indicate first correlation matrix minus 2/
1 power.
7. trailer system major loop ground method for diagnosing faults according to claim 3 or 4, which is characterized in that the step
Rapid S2 specifically includes the following steps:
S21: residual signals statistic is established according to the residual generation are as follows:
In formula,Indicate the inverse matrix of the covariance matrix of corresponding residual signals；
S22: calculating the residual signals statistic of history normal data, then calculates the residual signals using Monte Carlo method
The corresponding threshold value of statistic determines control limit.
8. a kind of trailer system major loop ground fault diagnosis system, including memory, processor and storage are on a memory
And the computer program that can be run on a processor, which is characterized in that the processor is realized when executing the computer program
The step of any the method for the claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810752527.8A CN109203995A (en) | 2018-07-10 | 2018-07-10 | A kind of trailer system major loop ground fault diagnosis method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810752527.8A CN109203995A (en) | 2018-07-10 | 2018-07-10 | A kind of trailer system major loop ground fault diagnosis method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109203995A true CN109203995A (en) | 2019-01-15 |
Family
ID=64990375
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810752527.8A Pending CN109203995A (en) | 2018-07-10 | 2018-07-10 | A kind of trailer system major loop ground fault diagnosis method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109203995A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110879372A (en) * | 2019-12-03 | 2020-03-13 | 中南大学 | Traction system main loop earth fault diagnosis method and system based on feature correlation |
| CN113433917A (en) * | 2021-08-11 | 2021-09-24 | 内蒙古京隆发电有限责任公司 | DCS (distributed control system) for power plant and control method |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101713807A (en) * | 2009-09-15 | 2010-05-26 | 深圳市普禄科智能检测设备有限公司 | High-speed signal acquisition based device and method for recording and analyzing locomotive movably grounded fault |
| KR20110016761A (en) * | 2009-08-12 | 2011-02-18 | 엘에스산전 주식회사 | Ground fault detection circuit for electric vehicle |
| CN105067940A (en) * | 2015-08-06 | 2015-11-18 | 株洲南车时代电气股份有限公司 | Grounding diagnostic method for main loop of locomotive traction converter |
| CN105204496A (en) * | 2015-09-25 | 2015-12-30 | 清华大学 | Sensor fault diagnosing method and system for air braking control system of motor train unit |
| CN106526400A (en) * | 2016-11-28 | 2017-03-22 | 株洲中车时代电气股份有限公司 | Grounding fault diagnosing method and apparatus of DC 600V train power supply system |
| CN106680660A (en) * | 2015-11-09 | 2017-05-17 | 中车大连电力牵引研发中心有限公司 | AC electric locomotive traction auxiliary system grounding fault detection method |
| CN106959397A (en) * | 2017-04-24 | 2017-07-18 | 南京航空航天大学 | A kind of method for designing of small fault diagnostic system for high ferro inverter |
| CN107357275A (en) * | 2017-07-27 | 2017-11-17 | 中南大学 | Non-gaussian industrial process fault detection method and system |
-
2018
- 2018-07-10 CN CN201810752527.8A patent/CN109203995A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20110016761A (en) * | 2009-08-12 | 2011-02-18 | 엘에스산전 주식회사 | Ground fault detection circuit for electric vehicle |
| CN101713807A (en) * | 2009-09-15 | 2010-05-26 | 深圳市普禄科智能检测设备有限公司 | High-speed signal acquisition based device and method for recording and analyzing locomotive movably grounded fault |
| CN105067940A (en) * | 2015-08-06 | 2015-11-18 | 株洲南车时代电气股份有限公司 | Grounding diagnostic method for main loop of locomotive traction converter |
| CN105204496A (en) * | 2015-09-25 | 2015-12-30 | 清华大学 | Sensor fault diagnosing method and system for air braking control system of motor train unit |
| CN106680660A (en) * | 2015-11-09 | 2017-05-17 | 中车大连电力牵引研发中心有限公司 | AC electric locomotive traction auxiliary system grounding fault detection method |
| CN106526400A (en) * | 2016-11-28 | 2017-03-22 | 株洲中车时代电气股份有限公司 | Grounding fault diagnosing method and apparatus of DC 600V train power supply system |
| CN106959397A (en) * | 2017-04-24 | 2017-07-18 | 南京航空航天大学 | A kind of method for designing of small fault diagnostic system for high ferro inverter |
| CN107357275A (en) * | 2017-07-27 | 2017-11-17 | 中南大学 | Non-gaussian industrial process fault detection method and system |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110879372A (en) * | 2019-12-03 | 2020-03-13 | 中南大学 | Traction system main loop earth fault diagnosis method and system based on feature correlation |
| CN110879372B (en) * | 2019-12-03 | 2021-04-23 | 中南大学 | Traction system main loop earth fault diagnosis method and system based on feature correlation |
| CN113433917A (en) * | 2021-08-11 | 2021-09-24 | 内蒙古京隆发电有限责任公司 | DCS (distributed control system) for power plant and control method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fu et al. | Contingency ranking based on severity indices in dynamic security analysis | |
| CN110018425A (en) | A kind of power battery fault diagnosis method and system | |
| CN104836223B (en) | Power grid parameter error and bad data coordinated identification and estimation method | |
| CN104933012B (en) | A kind of measured deviation On-line Fault recognition methods of traction substation transformer | |
| CN106841949B (en) | Method and device for monitoring stator insulation of three-phase asynchronous alternating current motor on line | |
| CN106443292B (en) | A kind of overhead transmission line single-phase earth fault detecting method based on zero sequence current measurement | |
| CN107357275A (en) | Non-gaussian industrial process fault detection method and system | |
| CN103018632B (en) | Small current grounding system single-phase ground fault line selection method based on fisher information | |
| CN109447187A (en) | Method of Motor Fault Diagnosis and system | |
| CN110441695A (en) | A kind of battery pack multiple faults error comprehensive diagnosis method combined based on model and signal processing | |
| CN108562821B (en) | Method and system for determining single-phase earth fault line selection of power distribution network based on Softmax | |
| CN109375010A (en) | Winding in Power Transformer method for diagnosing faults based on vibration distribution characteristic model | |
| CN109203995A (en) | A kind of trailer system major loop ground fault diagnosis method and system | |
| CN106959397A (en) | A kind of method for designing of small fault diagnostic system for high ferro inverter | |
| CN103544542A (en) | Power system transient stability margin predicting method | |
| CN106054023A (en) | Method for estimating system impedances at two sides in single-ended range finding of power transmission line | |
| Du et al. | Detection of single line-to-ground fault using convolutional neural network and task decomposition framework in distribution systems | |
| CN105954615A (en) | State assessment method and assessment system after transformer short circuit | |
| Penido et al. | An analytical zero sequence method to locate fault in distribution systems rich in DG | |
| EP3012643B1 (en) | Method and apparatus for identifying causes for cable overcurrent | |
| Zhang et al. | Fault location method based on SVM and similarity model matching | |
| Altaf et al. | Fault diagnosis in a distributed motor network using Artificial Neural Network | |
| CN103076529A (en) | CT (Current Transformer) secondary circuit intelligent-detection method | |
| CN108828437A (en) | Analog circuit fault feature extracting method based on cloud incidence coefficient matrix | |
| CN112801135A (en) | Fault line selection method and device for power plant service power system based on characteristic quantity correlation |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190115 |
|
| RJ01 | Rejection of invention patent application after publication |