CN103018058A - Similarity-based fault isolation method of train suspension system - Google Patents

Abstract

The invention relates to a similarity-based fault isolation method of a train suspension system. The method includes taking the train suspension system as a condition monitoring and fault diagnosis object, obtaining a system output residual through analyzing the motion state during operation of a train and Kalman filtering, obtaining fault signal frequency domain information through performing fast Fourier transform on the system output residual, and calculating a similarity score of a currently detected fault and a typical fault through the Eros algorithm to determine the type and position of the current fault. The method achieves intelligent isolation of train suspension system faults and plays important roles in guaranteeing train safe operation, guiding maintenance and overhauling of the train suspension system and improving the train operating efficiency.

Description

Train suspension fault separating method based on similarity

Technical field

The present invention relates to the train suspension fault separating method based on similarity.

Background technology

The train suspension between train body and the bogie and bogie and wheel between, consisted of by a large amount of different parts, comprise volute spring, damper, air spring etc.The suspension of rail vehicle usually be divided into and one be (wheel to and bogie between) with two be (between bogie and car body), also be divided into laterally and vertical system according to its impact on the train motion state simultaneously.On the one hand, suspension is supporting car body and bogie; On the other hand, hang system also play buffering by the caused wheel-rail force of track irregularity, This train is bound for XXX in control, keep the effects such as running comfort.

The outer comparatively ripe train fault detection system in transit of Present Domestic is mostly for the subsystems such as train dynamics system, auxiliary system, braking system, the i.e. fault detect of traction electric machine, inverter, air-conditioning system, door device, Pneumatic brake systems.In that traveling is in the technological system of condition monitoring for train, take positions such as wheel, bearing, bogie frames as main, almost do not have clear and definite with the System and method for of train suspension as condition monitoring and fault detect object yet.The suspension localization of fault is separated and is then still belonged to blank, therefore need to a kind ofly carry out the completely new approach that localization of fault is separated for the train suspension.

Summary of the invention

For above the deficiencies in the prior art, the invention provides a kind of train suspension fault separating method based on similarity.

Purpose of the present invention is achieved through the following technical solutions:

Based on the train suspension fault separating method of similarity, the method comprises following step:

The motion state information of each position when 1) utilizing acceleration transducer and gyroscope to obtain train operation;

2) acceleration signal that obtains in the step 1 is carried out anti-aliasing filter, high-pass filtering, quadratic integral processing, angular velocity signal is only carried out anti-aliasing filter and high-pass filtering process, export with the acquisition system; System output is that car body is vertical, transversal displacement and car body are nodded, shake the head, angle of roll position, and the nodding of vertical, the transversal displacement of trailing or leading bogie and trailing or leading bogie, shake the head, angle of roll position;

3) set up the vehicle suspension system kinetic model:

$\stackrel{\·}{x}=\mathrm{Ax}+B_{d}d$

y=Cx+D _dd

Wherein x is the state of motion of vehicle variable, and d is track excitation, and y is displacement signal, A, B, C _d, D _dBe the corresponding matrix of coefficients of state-space equation, design corresponding Kalman wave filter, output is processed to the described system of step 2, in the estimating step 2 vertical, the transversal displacement of car body and car body nod, shake the head, angle of roll position, and the nodding of vertical, the transversal displacement of trailing or leading bogie and trailing or leading bogie, shake the head, the running status such as angle of roll position, acquisition system output residual error;

4) utilize fault detection unit to carry out fault detect according to the output of the residual error in the step 3, after detecting fault, report to the police, and start the fault separable programming;

5) fault detection unit detect have fault in the system after, the system of Kalman wave filter output residual error in the acquisition step 3, data length is 10 seconds, and this segment data is done fast Fourier transform (FFT):

$R\left(k\right)=\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}r\left(n\right)e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00002599151500022.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}(k-1)\frac{n-1}{N}},1\&le;k\&le;N$

Obtain the fault-signal frequency domain information.Wherein r (n) represents the output residual error of Kalman wave filter in the step 3, and R (k) is the frequency domain information of fault-signal;

6) use Eros similarity matching algorithm to carry out the similarity coupling with what obtain in the step 5 when prior fault frequency domain information and the feature that is stored in typical fault in the knowledge base, finish fault and separate.Get the fault signature that is stored in the typical fault in the knowledge base, frequency domain information with the current fault-signal that from step 5, obtains, use the Eros algorithm to do one by one coupling, the fault that the calculating current detection arrives and the similarity score of typical fault, the fault of namely thinking current generation of score maximum.

Described motion state information comprises the car body vertical acceleration, values of lateral, and shake the head angular velocity and car body of the car body angular velocity of nodding, car body sidewinders angular velocity; And the vertical acceleration of trailing or leading bogie, transverse acceleration, the angular velocity of nodding shake the head angular velocity and angle of roll speed.

Described Eros similarity matching algorithm is as follows:

Ask for the covariance matrix when prior fault frequency domain information sequence

A=cov(R(k))

Wherein R (k) is fault residual signal frequency domain information sequence, and A is its covariance matrix;

Do the svd of covariance matrix:

A＝U∑V ^T

V _A=[a ₁,a ₂,...,a _m]

∑ _A=diag(λ _A1,λ _A2,...,λ _Am)

V wherein _ARight proper vector a serves as reasons _iThe right eigenmatrix that forms, ∑ _AFor by singular value λ _AiThe diagonal matrix that forms;

Determining Weights:

${\mathrm{\&omega;}}_{\mathrm{Ai}}={\mathrm{\&lambda;}}_{\mathrm{Ai}}/\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{\mathrm{Aj}}$

${\mathrm{\&omega;}}_{\mathrm{Bi}}={\mathrm{\&lambda;}}_{\mathrm{Bi}}/\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{\mathrm{Bj}}$

${\mathrm{\&omega;}}_i=({\mathrm{\&omega;}}_{\mathrm{Ai}}+{\mathrm{\&omega;}}_{\mathrm{Bi}})/(\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{\mathrm{Aj}}+\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{\mathrm{Bj}})$

ω wherein _iThe weight of corresponding i proper vector during for the similarity of compute matrix A and B;

Similarity measurement:

$\mathrm{Sim}(A,B)=\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_i|<a_i,b_i>|=\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_i\left|\cos {\mathrm{\&theta;}}_i\right|$

Sim (A, B) is required similarity, θ _iBe the corresponding right proper vector a with B of matrix A _iWith b _iAngle.

The invention has the advantages that:

Current international and domestic there is no for train suspension fault carried out the application technology that intelligent trouble separates, the localization of fault of suspension with separate the mode of generally taking the technician parking lot to verify maintenance and carry out.The present invention has realized that the intelligent trouble of train suspension fault separates, for ensureing safe train operation, instructing the maintenance of train suspension system and maintenance, raising train operation efficient all to have vital role.

Description of drawings

Figure 1 shows that fault diagnosis system structural framing figure of the present invention;

Figure 2 shows that sensor installation position synoptic diagram;

Fig. 3: signal conditioning circuit flow process;

Fig. 4: data pretreatment unit data process method;

Fig. 5: train suspension fault is separated basic procedure;

Fig. 6: based on the fault separation principle of Eros similarity matching algorithm.

Embodiment

Below in conjunction with accompanying drawing hardware configuration of the present invention, composition and fault separating method are described.

1. system structural framework

Be illustrated in figure 1 as fault diagnosis system structural framing figure of the present invention, described fault detection system comprises: sensor is used for obtaining train at the acceleration information of each position; Data acquisition unit, be responsible for connecting sensor and data pretreatment unit, the analog signal conversion that sensor is sent is the form that the data pretreatment unit can be identified, and sends to the data pretreatment unit with unified communication protocol, realizes each sensor measurement data acquisition and conversion reason; The data pretreatment unit, upper data management and the vehicle network management of being responsible for each signal gathering unit, and the data that the receive data collecting unit is sent here, data are carried out the work such as coordinate transform, high-pass filtering, quadratic integral computing, and then the pre-service result is passed to failure diagnosis unit by Ethernet; Described data acquisition unit and pretreatment unit have installation on each compartment, and train of fault diagnosis main frame is equipped with one, it is failure diagnosis unit, collect the information that gathers on each compartment by Ethernet, the data pre-service result who receives is judged, judge whether the train suspension breaks down and fault is separated.

2. sensor specification and layout scheme

Sensor type has three kinds, be respectively car body three-dimensional acceleration transducer, bogie two to acceleration transducer and angular-rate sensor, be respectively applied to obtain three acceleration signals of car body XYZ, bogie YZ binomial acceleration signal, and the nodding of car body and bogie, shake the head, angle of roll rate signal.Specification is as shown in the table:

Table one: sensor specification

Be illustrated in figure 2 as sensor installation position synoptic diagram, car body acceleration sensor and angular-rate sensor are laid in the central point of car bottom plate, to obtain car body sink-float, yaw, to nod, shake the head, sidewinder and the motion state information of 6 degree of freedom such as lengthwise movement.

Bogie acceleration transducer and angular-rate sensor are laid in cartridge position, center, bogie frame top, with the motion state information of 5 degree of freedom such as obtaining bogie sink-float, yaw, nod, shake the head, sidewinder.

3. data acquisition unit major function:

Data acquisition unit is responsible for connecting sensor and data pretreatment unit, the analog signal conversion that sensor is sent is the form that the data pretreatment unit can be identified, upwards send with unified communication protocol, realize each sensor measurement data acquisition and conversion reason.Data acquisition unit specifically will be finished following work: current signal is to the conversion of voltage signal, anti-aliasing filter, voltage transitions, and the work such as the A/D conversion of simulating signal, and by Ethernet image data is passed to the data pretreatment unit.

The signal gathering unit entity is signal regulating panel, mainly is to carry out isolation processing, analog to digital conversion and digital filtering for the simulating signal that sensor transmissions is come, and for providing data to composite node, its circuit flow process as shown in Figure 3.

● the signal regulating panel specification

Adopt 16 A/D convertor circuits, each passage capable of being reaches the 200KSPS sampling rate, adopts the anti-aliasing low-pass filter of second order, provides digital filter to realize the over-sampling function

Adopt 10M/100M/1000M Ethernet interface the transmission of data, support IEE1588 Network Synchronization agreement

CPU adopts Freescale MPC series high-performance processor, dominant frequency 400M, and internal memory is not less than 128MB, and plate carries 64M flash

Altera Cyclone EPC4 Series FPGA can be used for image data pre-service computing as coprocessor

The 24VDC power supply, board power consumption≤15W

All adopt the wide temperature device of high reliability technical grade

4. Signal Pretreatment Elementary Function:

Signal gathering unit and network node physically are responsible for connecting in the Signal Pretreatment unit, are responsible in logic data management and the vehicle network management of each signal gathering unit.The Signal Pretreatment unit receives and converges the data that signal gathering unit is sent here, data is carried out the work such as coordinate transform, filtering, integral operation, and then the pre-service result is passed to Diagnostic Service Host by Ethernet.

The Signal Pretreatment element entity is the data disposable plates, and major function is that data are carried out coordinate transform, digitized filtered, the work such as integral operation, and the integration function of finishing the isomeric data that accesses from different sensors.

Data disposable plates logical diagram is as shown in Figure 4:

● data disposable plates specification

CPU adopts Freescale MPC series high-performance processor, dominant frequency 400M

Internal memory is not less than 128MB

Plate carries 64M flash

Adopt 10M/100M/1000M Ethernet interface the transmission of data

The 5VDC power supply, board power consumption≤15W

All adopt the wide temperature device of high reliability technical grade

5. vehicle-mounted suspension fault diagnosis main frame:

Vehicle-mounted suspension fault diagnosis main frame is made of vehicle-mounted high-performance industrial computer and built-in bicycle fault detection algorithm, fault separation algorithm.The fault detect in transit of finishing the train suspension with separate.Particular content sees the diagnosis algorithm part for details.

Suspension fault separation algorithm step:

Train suspension fault based on similarity is separated basic procedure as shown in Figure 5, comprises following step:

The motion state information of each position when 1) utilizing acceleration transducer and gyroscope to obtain train operation.Comprise the car body vertical acceleration, values of lateral, shake the head angular velocity and car body of the car body angular velocity of nodding, car body sidewinders angular velocity; And the vertical acceleration of trailing or leading bogie, transverse acceleration, the angular velocity of nodding shake the head angular velocity and angle of roll speed.

2) acceleration signal that obtains in the step 1 is carried out the processing such as anti-aliasing filter, high-pass filtering, quadratic integral, angular velocity signal is only carried out anti-aliasing filter and high-pass filtering process, export with the acquisition system.System output is that car body is vertical, transversal displacement and car body are nodded, shake the head, angle of roll position, and the nodding of vertical, the transversal displacement of trailing or leading bogie and trailing or leading bogie, shake the head, angle of roll position.

3) set up the vehicle suspension system kinetic model:

$\stackrel{\&CenterDot;}{x}=\mathrm{Ax}+B_{d}d$

y=Cx+D _dd

Wherein x is the state of motion of vehicle variable, and d is track excitation, and y is displacement signal, A, B, C _d, D _dBe the corresponding matrix of coefficients of state-space equation.Design corresponding Kalman wave filter, output is processed to the described system of step 2, in the estimating step 2 vertical, the transversal displacement of car body and car body nod, shake the head, angle of roll position, and the nodding of vertical, the transversal displacement of trailing or leading bogie and trailing or leading bogie, shake the head, the running status such as angle of roll position, obtain residual error output.

4) utilize fault detection unit to carry out fault detect according to the output of the residual error in the step 3, after detecting fault, report to the police, and start the fault separable programming.The detection algorithm of fault detection unit is not this patent emphasis, does not repeat them here.

5) fault detection unit detect have fault in the system after, every residual error of Kalman wave filter output in the acquisition step 3, data length is 10 seconds.This segment data is done fast Fourier transform (FFT):

$R\left(k\right)=\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}r\left(n\right)e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00002599151500022.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}(k-1)\frac{n-1}{N}},1\&le;k\&le;N$

Obtain the fault-signal frequency domain information.Wherein r (n) represents the output residual error of Kalman wave filter in the step 3, and R (k) is the frequency domain information of fault-signal.

6) use Eros similarity matching algorithm to carry out the similarity coupling with what obtain in the step 5 when prior fault frequency domain information and the feature that is stored in typical fault in the knowledge base, finish fault and separate.The principle of similarity coupling is, get the fault signature that is stored in the typical fault in the knowledge base, frequency domain information with the current fault-signal that from step 5, obtains, use the Eros algorithm to do one by one coupling, the fault that the calculating current detection arrives and the similarity score of typical fault, the fault of namely thinking current generation of score maximum.Eros similarity matching algorithm is as follows.

Ask for the covariance matrix when prior fault frequency domain information sequence

A=cov(R(k))

Wherein R (k) is fault residual signal frequency domain information sequence, and A is its covariance matrix.

Do the svd of covariance matrix:

A＝U∑V ^T

V _A=[a ₁,a ₂,...,a _m]

∑ _A=diag(λ _A1,λ _A2,...,λ _Am)

V wherein _ARight proper vector a serves as reasons _iThe right eigenmatrix that forms, ∑ _AFor by singular value λ _AiThe diagonal matrix that forms.

Determining Weights:

${\mathrm{\&omega;}}_{\mathrm{Ai}}={\mathrm{\&lambda;}}_{\mathrm{Ai}}/\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{\mathrm{Aj}}$

${\mathrm{\&omega;}}_{\mathrm{Bi}}={\mathrm{\&lambda;}}_{\mathrm{Bi}}/\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{\mathrm{Bj}}$

${\mathrm{\&omega;}}_i=({\mathrm{\&omega;}}_{\mathrm{Ai}}+{\mathrm{\&omega;}}_{\mathrm{Bi}})/(\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{\mathrm{Aj}}+\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{\mathrm{Bj}})$

ω wherein _iThe weight of corresponding i proper vector during for the similarity of compute matrix A and B.

Similarity measurement:

$\mathrm{Sim}(A,B)=\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_i|<a_i,b_i>|=\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_i\left|\cos {\mathrm{\&theta;}}_i\right|$

Sim (A, B) is required similarity, θ _iBe the corresponding right proper vector a with B of matrix A _iWith b _iAngle.

As shown in Figure 6, use the Eros algorithm successively current detection to be carried out similarity measurement to the feature of fault with the fault signature that is stored in the knowledge base, the typical fault of similarity maximum namely think with current break down identical, thereby realize that fault separates.

Should be appreciated that the above detailed description of technical scheme of the present invention being carried out by preferred embodiment is illustrative and not restrictive.Those of ordinary skill in the art is reading on the basis of instructions of the present invention and can make amendment to the technical scheme that each embodiment puts down in writing, and perhaps part technical characterictic wherein is equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution break away from the spirit and scope of various embodiments of the present invention technical scheme.

Claims (3)

1. based on the train suspension fault separating method of similarity, it is characterized in that the method comprises following step:

The motion state information of each position when 1) utilizing acceleration transducer and gyroscope to obtain train operation;

2) acceleration signal that obtains in the step 1 is carried out anti-aliasing filter, high-pass filtering, quadratic integral processing, angular velocity signal is only carried out anti-aliasing filter and high-pass filtering process, export with the acquisition system; System output is that car body is vertical, transversal displacement and car body are nodded, shake the head, angle of roll position, and the nodding of vertical, the transversal displacement of trailing or leading bogie and trailing or leading bogie, shake the head, angle of roll position;

3) set up the vehicle suspension system kinetic model:

$\stackrel{\&CenterDot;}{x}=\mathrm{Ax}+B_{d}d$

y=Cx+D _dd

Wherein x is the state of motion of vehicle variable, and d is track excitation, and y is displacement signal, A, B, C _d, D _dBe the corresponding matrix of coefficients of state-space equation, design corresponding Kalman wave filter, output is processed to the described system of step 2, in the estimating step 2 vertical, the transversal displacement of car body and car body nod, shake the head, angle of roll position, and the nodding of vertical, the transversal displacement of trailing or leading bogie and trailing or leading bogie, shake the head, angle of roll position running status, acquisition system output residual error;

4) utilize fault detection unit to carry out fault detect according to the output of the residual error in the step 3, after detecting fault, report to the police, and start the fault separable programming;

5) fault detection unit detect have fault in the system after, the system of Kalman wave filter output residual error in the acquisition step 3, data length is 10 seconds, and this segment data is fast fourier transform FFT:

$R\left(k\right)=\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}r\left(n\right)e^{-j2\mathrm{\&pi;}(k-1)\frac{n-1}{N}},1\&le;k\&le;N$

Obtain the fault-signal frequency domain information.Wherein r (n) represents the output residual error of Kalman wave filter in the step 3, and R (k) is the frequency domain information of fault-signal;

6) use Eros similarity matching algorithm to carry out the similarity coupling with what obtain in the step 5 when prior fault frequency domain information and the feature that is stored in typical fault in the knowledge base, finish fault and separate.Get the fault signature that is stored in the typical fault in the knowledge base, frequency domain information with the current fault-signal that from step 5, obtains, use the Eros algorithm to do one by one coupling, the fault that the calculating current detection arrives and the similarity score of typical fault, the fault of namely thinking current generation of score maximum.

2. the train suspension fault separating method based on similarity according to claim 1, it is characterized in that described motion state information comprises car body vertical acceleration, values of lateral, shake the head angular velocity and car body of the car body angular velocity of nodding, car body sidewinders angular velocity; And the vertical acceleration of trailing or leading bogie, transverse acceleration, the angular velocity of nodding shake the head angular velocity and angle of roll speed.

3. the train suspension fault separating method based on similarity according to claim 1 is characterized in that, described Eros similarity matching algorithm is as follows:

Ask for the covariance matrix when prior fault frequency domain information sequence

A=cov(R(k))

Wherein R (k) is fault residual signal frequency domain information sequence, and A is its covariance matrix;

Do the svd of covariance matrix:

A＝U∑V ^T

V _A=[a ₁,a ₂,...,a _m]

∑ _A=diag(λ _A1,λ _A2,...,λ _Am)

V wherein _ARight proper vector a serves as reasons _iThe right eigenmatrix that forms, ∑ _AFor by singular value λ _AiThe diagonal matrix that forms;

Determining Weights:

${\mathrm{\&omega;}}_{\mathrm{Ai}}={\mathrm{\&lambda;}}_{\mathrm{Ai}}/\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{\mathrm{Aj}}$

${\mathrm{\&omega;}}_{\mathrm{Bi}}={\mathrm{\&lambda;}}_{\mathrm{Bi}}/\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{\mathrm{Bj}}$

${\mathrm{\&omega;}}_i=({\mathrm{\&omega;}}_{\mathrm{Ai}}+{\mathrm{\&omega;}}_{\mathrm{Bi}})/(\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{\mathrm{Aj}}+\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{\mathrm{Bj}})$

ω wherein _iThe weight of corresponding i proper vector during for the similarity of compute matrix A and B;

Similarity measurement:

$\mathrm{Sim}(A,B)=\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_i|<a_i,b_i>|=\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_i\left|\cos {\mathrm{\&theta;}}_i\right|$

Sim (A, B) is required similarity, θ _iBe the corresponding right proper vector a with B of matrix A _iWith b _iAngle.

Images