CN109187060A - The detection of train speed sensor abnormal signal and axis locking method for diagnosing faults - Google Patents
The detection of train speed sensor abnormal signal and axis locking method for diagnosing faults Download PDFInfo
- Publication number
- CN109187060A CN109187060A CN201810856916.5A CN201810856916A CN109187060A CN 109187060 A CN109187060 A CN 109187060A CN 201810856916 A CN201810856916 A CN 201810856916A CN 109187060 A CN109187060 A CN 109187060A
- Authority
- CN
- China
- Prior art keywords
- matrix
- speed sensor
- axis locking
- train
- diagnosing faults
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M17/00—Testing of vehicles
- G01M17/08—Railway vehicles
- G01M17/10—Suspensions, axles or wheels
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
- G01P21/00—Testing or calibrating of apparatus or devices covered by the preceding groups
- G01P21/02—Testing or calibrating of apparatus or devices covered by the preceding groups of speedometers
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Train Traffic Observation, Control, And Security (AREA)
Abstract
The present invention relates to a kind of detection of train speed sensor abnormal signal and axis locking method for diagnosing faults, comprising: step S1: carrying out between centers residual computations two-by-two to 4 shaft speed sensor signals of train, extraction obtains 6 stability maintenance state characteristic quantities;Step S2: the principal component model under health condition is obtained using the 6 stability maintenance state characteristic quantities training under non-faulting state, and obtains statistic control limit;Step S3: it inputs based on the obtained 6 stability maintenance state characteristic quantities of actual measurement sensor signal to principal component model, the statistics magnitude of test data is calculated, and judges whether the statistics magnitude exceeds control limit, breaks down if YES then assert and executes step S4, it does not break down if it has not, then assert;Step S4: fault type and abort situation are obtained according to actual measurement sensor signal.Compared with prior art, the present invention is directed to critical component-train shaft end velocity sensor, using specific effective technique algorithm, improves the accuracy of diagnosis, can be effectively reduced train axis locking false alarm rate.
Description
Technical field
The present invention relates to a kind of method for diagnosing faults, more particularly, to a kind of detection of train speed sensor abnormal signal and
Axis locking method for diagnosing faults.
Background technique
With greatly developing for track transportation industry, for rail traffic fault detection and diagnosis technology progressively towards intelligence
Energyization, efficient direction are developed.For the axis locking fault diagnosis of train, existing diagnostic logic is based purely on test gained speed
Degree signal given threshold gives conclusion judgement.So as to cause practical maintenance discovery, the axis locking failure reported is mostly to arrange on inspection
False alarm caused by the shaft speed sensor failure of vehicle.The phenomenon has seriously affected the operation of train, reduces the operation of train
Efficiency.It is therefore desirable to be directed to the phenomenon, design corresponding for distinguishing the locking of train axis and shaft speed sensor failure
Fault detection and diagnosis algorithm.
Technology in the field, i.e. fault detection and diagnosis technology speed sensor fault diagnostic field application, greatly
It mostly still stays in fault simulations emulation and carries out detection-phase, be not yet related to the diagnostic work of physical fault type, for from speed
The angle of signal fault detection and diagnosis solves the problems, such as the alert technology of axis locking false, still belongs to blank.
Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide a kind of train speeds to sense
The detection of device abnormal signal and axis locking method for diagnosing faults.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of detection of train speed sensor abnormal signal and axis locking method for diagnosing faults, comprising:
Step S1: between centers residual computations two-by-two are carried out to 4 shaft speed sensor signals of train, extraction obtains 6 stability maintenance states
Characteristic quantity;
Step S2: obtaining the principal component model under health condition using the 6 stability maintenance state characteristic quantities training under non-faulting state,
And obtain statistic control limit;
Step S3: the 6 stability maintenance state characteristic quantities obtained based on actual measurement sensor signal are inputted to principal component model, calculates and surveys
The statistics magnitude of data is tried, and judges whether the statistics magnitude exceeds control limit, breaks down if YES then assert and executes step
Rapid S4 does not break down if it has not, then assert;
Step S4: fault type and abort situation are obtained according to actual measurement sensor signal.
The 6 stability maintenance state characteristic quantity are as follows:
xij=| Vi-Vj|, i, j=1,2,3,4, i < j
Wherein: Vi、VjThe sensor signal of the velocity sensor acquisition of respectively i-th and j-th axis.
Statistic control in the step S2 is limited to:
Wherein: ζ is statistic control limit, and g is coefficient,It is freedom degree for h, confidence level is the χ of α2Distribution, h are certainly
By spending, α is confidence level, and S is the covariance matrix of training data, φ is by test data conversion be overall target matrix, tr
() is the mark of matrix, and X is training data, and m is training data variable number.
Statistics magnitude in the step S3 specifically:
Fai=xTφx
Wherein: fai is statistics magnitude, and x is test data.
The step S4 is specifically included:
Step S41: calculating the reconstruct contribution margin of each variable, according to reconstruct contribution plot separation principle, extracts failure variable,
And according to isolation full terms, all failure collection X are obtainedf;
Step S42: after the variable for determining failure, carrying out regular quantization to comprehensive statistics amount fai value, and after recording quantization
The translation type of each state and the time of generation obtain state change matrix fait;
Step S43: application failure pattern analysis and impact analysis result bonding state translation type establish Fault Petri Net
Model determines network associate Matrix C and transition excitation matrix U;
Step S44: reading state transition matrix fait, obtain Petri network initial marking M0, application failure Petri network pushes away
Manage equation of stateObtain next mark M1;
Step S45: repeating step S44, until mark Mi+1Reach fault type library institute.
The reconstruct contribution margin specifically:
RBCi Index=x Tφξi(ξi Tφξi)-1ξi Tφx
Wherein: RBCi IndexFor the reconstruct contribution margin of i-th of variable,xFor former data, it is comprehensive that φ, which is by test data conversion,
Close the matrix of index, ξiFor fault direction matrix, ()-1For inverse matrix.
In the step S44, operatorCalculating process are as follows:
Then ci=∨ (ai,bi), 1≤i≤n
Wherein: ∨ () is two Matrix Calculating "or",
OperatorCalculation method are as follows:
Enable C ∈ Rn×m,U∈Rn×1, have
Wherein: C is the incidence matrix of Petri net model, and U is transition excitation matrix.
Compared with prior art, the invention has the following advantages:
1) it is directed to critical component-train shaft end velocity sensor, using specific effective technique algorithm, improves diagnosis
Accuracy, can be effectively reduced train axis locking false alarm rate, give maintenance personal and effectively prompt, reduce maintenance cost,
Improve the efficiency of operation and operational reliability of train.
2) in the case where determining failure axis, the base of failure mode and effect analysis is carried out in velocity sensor and axis locking
On plinth, improved Fault Petri Net pessimistic concurrency control is established, and common velocity sensor and axis armful are realized by measured data verifying
The differentiation of dead failure.
3) the axis locking false alarm problem often occurred for train, devise based on principal component analysis-reconstruct contribution plot with
And improved Fault Petri Net network method, site of deployment actual measurement and test platform data acquired are with demonstrating inventive method effective
Property, and solve the problems, such as that train axis locking false alarm is speed sensor fault on inspection.
Detailed description of the invention
Fig. 1 is the key step flow diagram of the method for the present invention;
Fig. 2 is general frame schematic diagram of the present invention;
Fig. 3 is Petri net model schematic diagram established by the present invention;
Fig. 4 is the algorithm flow schematic diagram of the method for the present invention application.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with the technology of the present invention side
Implemented premised on case, the detailed implementation method and specific operation process are given, but protection scope of the present invention is unlimited
In following embodiments.
A kind of detection of train speed sensor abnormal signal and axis locking method for diagnosing faults, as shown in Figure 1, comprising:
Step S1: between centers residual computations two-by-two are carried out to 4 shaft speed sensor signals of train, extraction obtains 6 stability maintenance states
Characteristic quantity;
Before carrying out above-mentioned work, failure mould is carried out to velocity sensor most common failure and axis locking failure etc. first
Formula and impact analysis, analysis result is as described in Figure 2, and different zones are corresponding with different fault-signal performance forms, as d class breaks
Failure and axis locking failure have its similar place in the form of expression of speed signal, i.e. speed signal can all be reduced to 0, this
It is also the reason of velocity sensor disconnection fault is often axis locking failure by false alarm.But they also have between the two it is certain
The performance of speed signal is not the case where being instantaneously reduced to 0 when difference, i.e. axis locking failure, and velocity sensor disconnection fault occurs
When, speed signal be one instantaneously be reduced to 0 state.And other several fault types, such as slide, interfere, with axis locking
Failure has obvious difference.
It is of the invention in realization step 1: before the detection of failure, it would be desirable to logical on the basis of understanding above-mentioned mechanism
The raw velocity signal for crossing 4 axis obtains the steady state characteristic amount x for capableing of characterization failure12、x13、x14、x23, x24, x34:
xij=| Vi-Vj|, i, j=1,2,3,4, i < j
Wherein Vi、VjIndicate the raw velocity of corresponding axis.
Step S2: obtaining the principal component model under health condition using the 6 stability maintenance state characteristic quantities training under non-faulting state,
And obtain statistic control limit;
Statistic control in step S2 is limited to:
Wherein: ζ is statistic control limit, and g is coefficient,It is freedom degree for h, confidence level is the χ of α2Distribution, h are certainly
By spending, α is confidence level, and S is the covariance matrix of training data, φ is by test data conversion be overall target matrix, tr
() is the mark of matrix, and X is training data, and m is training data variable number.
By xi under normal circumstancesj6 dimensional vectors are formed after Z-score standardization, drawing method is contributed according to accumulation
It is shown below and seeks corresponding principal component, m indicates variable number in formula, herein m=6.The cumulative variance tribute of current k pivot
When the rate of offering reaches 90%, taking pivot number is k, has just obtained k eigenvalue λ1,λ2,…,λk(λ1> λ2> ... λk), and it is corresponding
Feature vector p1,p2,…pk。
If between centers residual error steady state characteristic amount under normal circumstances is Xij, obtain the score matrix after principal component are as follows:
ti=XijPi
Foundation
Xij=t1p1 T+t2p2 T+…+tkpk T
=Xijp1p1 T+Xijp2p2 T+…+Xijpkpk T
E=Xij-Xij
Principal component model, and input test vector x are finally obtained, corresponding comprehensive statistics figureofmerit is obtained:
Q and T in comprehensive statistics amount fai (x)2It is the Statistic in Common in principal component analysis, QUCL and TUCL difference
For statistic Q and T2Corresponding control limit.T2Statistic measures sample vector in the variation of principal component space:
T2=xTP·Λ-1·PTx≤Tα 2
Wherein, Λ=diag { λ1,λ2,...λk, T α2The control limit for being α for confidence level.
Fk,n-k:αIt is with k and n-k freedom degree, confidence level is the F Distribution Value of α
(TUCL), wherein n is test sample quantity.
Q index measures sample vector in the variation of the projection in residual error space:
Q=| | (I-PPT)·x||2≤δα 2
Wherein, δα 2(QUCL) indicate that the control that confidence level is α limits.
δα 2Common calculation formula is as follows:
Wherein,λj iFor the characteristic value of the covariance matrix of X, CαFor mark
Threshold value of the quasi normal distribution in the case where confidence level is α.
Step S3: the 6 stability maintenance state characteristic quantities obtained based on actual measurement sensor signal are inputted to principal component model, calculate test
The statistics magnitude of data, and judge whether the statistics magnitude exceeds control limit, it breaks down if YES then assert and executes step
S4 does not break down if it has not, then assert;
Statistics magnitude in step S3 specifically:
Fai=xTφx
Wherein: fai is statistics magnitude, and x is test data.
Through overall target fai compared with control limit ζ, think to break down beyond limit value, otherwise normally.Control
Limit the calculation formula of ζ are as follows:
Wherein: ζ is statistic control limit, and g is coefficient,It is freedom degree for h, confidence level is the χ of α2Distribution, h are certainly
By spending, α is confidence level, and S is the covariance matrix of training data, φ is by test data conversion be overall target matrix, tr
() is the mark of matrix, and X is training data, and m is training data variable number.
When a failure occurs, the theory of reference reconstruct contribution plot, isolates corresponding failure variable.
Step S4: fault type and abort situation are obtained according to actual measurement sensor signal, specifically included:
Step S41: calculating the reconstruct contribution margin of each variable, according to reconstruct contribution plot separation principle, extracts failure variable,
And according to isolation full terms, all failure collection X are obtainedf;
Step S42: after the variable for determining failure, carrying out regular quantization to comprehensive statistics amount fai value, and after recording quantization
The translation type of each state and the time of generation obtain state change matrix fait;
Step S43: application failure pattern analysis and impact analysis result bonding state translation type establish Fault Petri Net
Model determines network associate Matrix C and transition excitation matrix U;
Step S44: reading state transition matrix fait, obtain Petri network initial marking M0, application failure Petri network reasoning
Equation of stateObtain next mark M1;
Step S45: repeating step S44, until mark Mi+1Reach fault type library institute.
Reconstruct contribution margin specifically:
RBCi Index=x Tφξi(ξi Tφξi)-1ξi Tφx
Wherein: RBCi IndexFor the reconstruct contribution margin of i-th of variable, x is former data, and it is comprehensive that φ, which is by test data conversion,
Close the matrix of index, ξiFor fault direction matrix, ()-1For inverse matrix.
In step S44, operatorCalculating process are as follows:
Then ci=∨ (ai,bi), 1≤i≤n
Wherein: ∨ () is two Matrix Calculating "or",
OperatorCalculation method are as follows:
Enable C ∈ Rn×m,U∈Rn×1, have
Wherein: C is the incidence matrix of Petri net model, and U is transition excitation matrix.
In addition, reconstruct contribution plot failure variable is reconstructed it is as follows:
zi=x- ξifi
Wherein x is former data, ξiIt is fault direction matrix, fiIt is corresponding failure amplitude matrix.
Reconstruct contribution plot theory thinks variable xiTo the reconstruct contribution rate (reconstruction- of fault detection statistic
based contribution RBC)RBCi IndexMaximum is corresponding failure variable, RBCi IndexIs defined as:
RBCi Index=| | ξifi||2φ=| | ξi(ξi Tφξi)-1ξi Tφx||2M=xTφξi(ξi Tφξi)-1ξi Tφx
Complete two conditions of Fault Isolation have been determined:
(1) when selection variable belongs to the variable of responsible failure, the RBC of Testing index fai should be greater than its control limit.
(2) when the fault direction of isolation is accurate, reconstruction value ZiMonitoring index Index (zi) meet:
Index(zi)=(x- ξifi)Tφ(x-ξifi)≤ζ
Wherein ZiIt is the value to the reconstruct of failure variable, ξiIndicate fault direction matrix.ζ is the control limit of overall target fai
Value.
The present invention realizes fault detection in application principal component analytical method, and the comprehensive statistics figureofmerit of test data is fai,
Failure variables collection X is isolated using reconstruct contribution drawing methodf, and determination is out of order after axis, according to fault mode and influence
Analysis result carries out sign to the comprehensive statistics figureofmerit fai of test data and quantifies to obtain Q_fai, and records the variation of Q_fai
Obtain fai state of value transformation matrices fait。
According to faitIt is as shown in Figure 3 that value establishes corresponding Fault Petri Net:
Several basic elements of Petri network include library institute, transition and incidence matrix.
The institute of library shown in Fig. 3 of the present invention p1,p2,p3,p4,p5,p6,p7For library related with fai state of value transformation matrices
Institute, and library institute p8,p9,p10,p11,p12It is to indicate different fault types, including failure classes such as velocity sensor broken string, axis locking
Type.
Of the invention specific detection, isolation are as shown in Figure 4 with diagnosis algorithm process.
Invention using principal component analysis realize failure detection, using reconstruct contribution plot method realize failure variable every
From finally by the diagnosis for establishing Fault Petri Net model realization failure.In diagnosis of partial, after Fault Petri Net is established,
The present invention obtains initial marking M by reading fai state of value transformation matrices0, and pass through equation of state:
Obtain next mark M after transition excitation1, wherein C is the incidence matrix of Petri net model, and U is and reality
Fault category determines related transition excitation matrix.
The present invention is that Petri network dynamic variation characteristic related to time is applied among online fault diagnosis,
On the basis of method detects that faulty and isolation is out of order, Petri network is obtained by constantly reading transition excitation matrix state
Identify Mi, until MiReach the library institute p of characterization failure type8,p9,p10,p11,p12, diagnosis is to terminate.
Claims (7)
1. a kind of train speed sensor abnormal signal detection and axis locking method for diagnosing faults characterized by comprising
Step S1: between centers residual computations two-by-two are carried out to 4 shaft speed sensor signals of train, extraction obtains 6 dimension steady state characteristics
Amount;
Step S2: the principal component model under health condition is obtained using the 6 stability maintenance state characteristic quantities training under non-faulting state, and is obtained
It controls and limits to statistic;
Step S3: the 6 stability maintenance state characteristic quantities obtained based on actual measurement sensor signal are inputted to principal component model, calculate test data
Statistics magnitude, and judge the statistics magnitude whether exceed control limit, if YES then assert break down and execute step S4, if
Be it is no, then assert and do not break down;
Step S4: fault type and abort situation are obtained according to actual measurement sensor signal.
2. a kind of train speed sensor abnormal signal detection according to claim 1 and axis locking method for diagnosing faults,
It is characterized in that, the 6 stability maintenance state characteristic quantity are as follows:
xij=| Vi-Vj|, i, j=1,2,3,4, i < j
Wherein: xijFor the between centers residual error of the sensor signal of the velocity sensor acquisition of i-th and j-th axis, Vi、VjRespectively
The sensor signal of the velocity sensor acquisition of i-th and j-th axis.
3. a kind of train speed sensor abnormal signal detection according to claim 1 and axis locking method for diagnosing faults,
It is characterized in that, the statistic control in the step S2 is limited to:
Wherein: ζ is statistic control limit, and g is coefficient, χh 2 ,αIt is freedom degree for h, confidence level is the χ of α2Distribution, h is freedom degree, α
For confidence level, S is the covariance matrix of training data, φ is by test data conversion be overall target matrix, tr () is
The mark of matrix, X are training data, and m is training data variable number.
4. a kind of train speed sensor abnormal signal detection according to claim 3 and axis locking method for diagnosing faults,
It is characterized in that, the statistics magnitude in the step S3 specifically:
Fai=xTφx
Wherein: fai is statistics magnitude, and x is test data.
5. a kind of train speed sensor abnormal signal detection according to claim 4 and axis locking method for diagnosing faults,
It is characterized in that, the step S4 is specifically included:
Step S41: calculating the reconstruct contribution margin of each variable, according to reconstruct contribution plot separation principle, extracts failure variable, and root
According to isolation full terms, all failure collection X are obtainedf;
Step S42: after the variable for determining failure, carrying out regular quantization to comprehensive statistics amount fai value, and records each after quantization
The translation type of state and the time of generation obtain state change matrix fait;
Step S43: application failure pattern analysis and impact analysis result bonding state translation type establish Fault Petri Net mould
Type determines network associate Matrix C and transition excitation matrix U;
Step S44: reading state transition matrix fait, obtain Petri network initial marking M0, application failure Petri network reasoning state
EquationObtain next mark M1;
Step S45: repeating step S44, until mark Mi+1Reach fault type library institute.
6. a kind of train speed sensor abnormal signal detection according to claim 5 and axis locking method for diagnosing faults,
It is characterized in that, the reconstruct contribution margin specifically:
RBCi Index=xTφξi(ξi Tφξi)-1ξi Tφx
Wherein: RBCi IndexFor the reconstruct contribution margin of i-th of variable, x is former data, and it is that synthesis refers to that φ, which is by test data conversion,
Target matrix, ξiFor fault direction matrix, ()-1For inverse matrix.
7. a kind of train speed sensor abnormal signal detection according to claim 6 and axis locking method for diagnosing faults,
It is characterized in that, in the step S44, operatorCalculating process are as follows:
Then ci=∨ (ai,bi), 1≤i≤n
Wherein: ∨ () be two matrixes or operation,
OperatorCalculation method are as follows:
Enable C ∈ Rn×m,U∈Rn×1, have
Wherein: C is the incidence matrix of Petri net model, and U is transition excitation matrix.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810856916.5A CN109187060B (en) | 2018-07-31 | 2018-07-31 | The detection of train speed sensor abnormal signal and axis locking method for diagnosing faults |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810856916.5A CN109187060B (en) | 2018-07-31 | 2018-07-31 | The detection of train speed sensor abnormal signal and axis locking method for diagnosing faults |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109187060A true CN109187060A (en) | 2019-01-11 |
CN109187060B CN109187060B (en) | 2019-10-18 |
Family
ID=64937499
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810856916.5A Active CN109187060B (en) | 2018-07-31 | 2018-07-31 | The detection of train speed sensor abnormal signal and axis locking method for diagnosing faults |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109187060B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109855855A (en) * | 2019-03-13 | 2019-06-07 | 山东科技大学 | Bullet train closed loop brake system intermittent fault detection method |
CN111038474A (en) * | 2020-01-02 | 2020-04-21 | 中车青岛四方机车车辆股份有限公司 | Method and device for judging locking fault of rail vehicle |
CN111380708A (en) * | 2020-05-11 | 2020-07-07 | 中车青岛四方机车车辆股份有限公司 | Train axle locking fault diagnosis method and train axle locking fault diagnosis system |
CN112861071A (en) * | 2021-02-05 | 2021-05-28 | 哈尔滨工程大学 | High-speed rail traction system anomaly detection method based on deep self-coding |
CN113670360A (en) * | 2021-08-31 | 2021-11-19 | 三一专用汽车有限责任公司 | Monitoring method, system, device, vehicle, medium and product |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6359346B1 (en) * | 2000-08-23 | 2002-03-19 | General Electric Company | Processor and method for accommodating failed speed sensors in a locomotive |
CN1918475A (en) * | 2003-12-26 | 2007-02-21 | 博世株式会社 | Failure diagnosis device for vehicle body acceleration sensor and anti-lock brake system |
CN102175460A (en) * | 2011-01-28 | 2011-09-07 | 中国人民解放军海军工程大学 | Gas path fault diagnostic method for marine gas turbine based on fuzzy Petri net (FPN) |
CN102680817A (en) * | 2012-04-28 | 2012-09-19 | 辽宁省电力有限公司朝阳供电公司 | Transformer fault diagnosis method based on fuzzy Petri net |
CN103713628A (en) * | 2013-12-31 | 2014-04-09 | 上海交通大学 | Fault diagnosis method based on signed directed graph and data constitution |
CN104596780A (en) * | 2015-02-12 | 2015-05-06 | 清华大学 | Diagnosis method for sensor faults of motor train unit braking system |
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 |
CN107255792A (en) * | 2017-06-20 | 2017-10-17 | 华中科技大学 | A kind of electronic type voltage transformer error on-line monitoring method and system |
CN107782917A (en) * | 2016-08-24 | 2018-03-09 | 联合汽车电子有限公司 | The fault diagnosis and processing method of vehicle speed sensor |
-
2018
- 2018-07-31 CN CN201810856916.5A patent/CN109187060B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6359346B1 (en) * | 2000-08-23 | 2002-03-19 | General Electric Company | Processor and method for accommodating failed speed sensors in a locomotive |
CN1918475A (en) * | 2003-12-26 | 2007-02-21 | 博世株式会社 | Failure diagnosis device for vehicle body acceleration sensor and anti-lock brake system |
CN102175460A (en) * | 2011-01-28 | 2011-09-07 | 中国人民解放军海军工程大学 | Gas path fault diagnostic method for marine gas turbine based on fuzzy Petri net (FPN) |
CN102680817A (en) * | 2012-04-28 | 2012-09-19 | 辽宁省电力有限公司朝阳供电公司 | Transformer fault diagnosis method based on fuzzy Petri net |
CN103713628A (en) * | 2013-12-31 | 2014-04-09 | 上海交通大学 | Fault diagnosis method based on signed directed graph and data constitution |
CN104596780A (en) * | 2015-02-12 | 2015-05-06 | 清华大学 | Diagnosis method for sensor faults of motor train unit braking system |
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 |
CN107782917A (en) * | 2016-08-24 | 2018-03-09 | 联合汽车电子有限公司 | The fault diagnosis and processing method of vehicle speed sensor |
CN107255792A (en) * | 2017-06-20 | 2017-10-17 | 华中科技大学 | A kind of electronic type voltage transformer error on-line monitoring method and system |
Non-Patent Citations (1)
Title |
---|
张明伟: "列车齿轮速度传感器性能及故障诊断检测技术", 《机械制造与自动化》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109855855A (en) * | 2019-03-13 | 2019-06-07 | 山东科技大学 | Bullet train closed loop brake system intermittent fault detection method |
CN111038474A (en) * | 2020-01-02 | 2020-04-21 | 中车青岛四方机车车辆股份有限公司 | Method and device for judging locking fault of rail vehicle |
CN111380708A (en) * | 2020-05-11 | 2020-07-07 | 中车青岛四方机车车辆股份有限公司 | Train axle locking fault diagnosis method and train axle locking fault diagnosis system |
CN112861071A (en) * | 2021-02-05 | 2021-05-28 | 哈尔滨工程大学 | High-speed rail traction system anomaly detection method based on deep self-coding |
CN113670360A (en) * | 2021-08-31 | 2021-11-19 | 三一专用汽车有限责任公司 | Monitoring method, system, device, vehicle, medium and product |
Also Published As
Publication number | Publication date |
---|---|
CN109187060B (en) | 2019-10-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109187060B (en) | The detection of train speed sensor abnormal signal and axis locking method for diagnosing faults | |
CN110704801B (en) | Bridge cluster structure operation safety intelligent monitoring and rapid detection complete method | |
CN109186813A (en) | A kind of temperature sensor self-checking unit and method | |
CN111737909B (en) | Structural health monitoring data anomaly identification method based on space-time graph convolutional network | |
CN109710983B (en) | Diesel engine cylinder layered fault diagnosis method based on key performance indexes | |
CN110704911B (en) | Bridge damage cross positioning method based on cluster structure similarity | |
CN102829967A (en) | Time-domain fault identifying method based on coefficient variation of regression model | |
Li et al. | Fault identification in PCA method during sensor condition monitoring in a nuclear power plant | |
CN106647650A (en) | Distributed industrial process monitoring method based variable weighting PCA (Principal Component Analysis) model | |
Ma et al. | Probabilistic reconstruction for spatiotemporal sensor data integrated with Gaussian process regression | |
CN110399986A (en) | A kind of generation method of pumping plant unit fault diagnosis system | |
CN112632845B (en) | Data-based mini-reactor online fault diagnosis method, medium and equipment | |
Ma et al. | A prediction method based on stepwise regression analysis for train axle temperature | |
CN115048998B (en) | Cable-stayed bridge group cable force abnormity identification and positioning method based on monitoring data | |
Ji | Optimization-based incipient fault isolation for the high-speed train air brake system | |
CN116292367A (en) | Power plant fan system abnormal condition detection method based on one-dimensional convolution | |
Gao et al. | Detection of multi-type data anomaly for structural health monitoring using pattern recognition neural network | |
CN107330264A (en) | A kind of verification method of bridge monitoring data reliability | |
CN110705129B (en) | Fault diagnosis method based on fully-correlated dynamic KPLS | |
CN101968379B (en) | Method for extracting characteristic information of operating condition vibration signal of aircraft engine rotor system | |
Yang et al. | Bridge Cable Anomaly Detection Based on Local Variability in Feature Vector of Monitoring Group Cable Forces | |
Li et al. | Methods for the rapid detection of boundary condition variations in structural systems | |
Li et al. | Sensor fault localization with accumulated residual contribution rate for bridge SHM | |
Gao et al. | Research and application of pattern recognition LSTM based bridge data anomaly detection | |
Uhrig | Potential Use Of Neural Networks In Nuclear Power Plants |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |