CN107203746A - A kind of switch breakdown recognition methods - Google Patents
A kind of switch breakdown recognition methods Download PDFInfo
- Publication number
- CN107203746A CN107203746A CN201710341307.1A CN201710341307A CN107203746A CN 107203746 A CN107203746 A CN 107203746A CN 201710341307 A CN201710341307 A CN 201710341307A CN 107203746 A CN107203746 A CN 107203746A
- Authority
- CN
- China
- Prior art keywords
- curve
- similarity
- identified
- track switch
- point
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Train Traffic Observation, Control, And Security (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a kind of switch breakdown recognition methods, wherein, this method includes：Gather each operation curve of track switch；Acquired track switch operation curve is divided into normalized curve and damage curve；For normalized curve and damage curve, the most representational curve of one feature of selection is the representative curve of such curve；The similarity 2 of curve is represented using similarity 1, curve to be identified and the failure of similarity algorithm calculating curve to be identified and normal representation curve；Compare calculating gained similarity, if similarity 1 is more than similarity 2, the curve is normalized curve, if similarity 1 is less than similarity 2, the curve is damage curve.Solved in the prior art by the present invention, switch breakdown type is judged by artificial experience, the problem of causing to fail to report and report by mistake, it is achieved thereby that track switch automatic identification fault category, improves overhaul efficiency and system reliability, it is ensured that traffic safety.
Description
Technical field
The present invention relates to field of track traffic, and in particular to a kind of switch breakdown recognition methods.
Background technology
Track switch is that train is transferred to or crossed line facility essential during another strand of track from one track, is rail
One important component in road, is also fault rate highest equipment.Once track switch is broken down, it is impossible to complete compulsory exercise,
Light then temporary parking a few hours, it is delayed the time of a large amount of passengers；Heavy then compartment derails, and causes casualties.
At present, China is mainly monitored using microcomputer detecting system to switch status, the practice situation from scene
See, the Switch current curve mainly gathered by manual observation microcomputer detecting system recognizes fault type, it is necessary to substantial amounts of
Staff；Whether the working experience that its identification accuracy depends on related personnel is enriched, and this to fail to report and reported by mistake existing
As occuring now and then；Staff needs time series analysis curvilinear characteristic and failure judgement type, it is impossible to which real-time online recognizes, efficiency compared with
It is low.In addition, railway department in order to further prevent the generation of accident, arrange related personnel inspect periodically and repair track switch, it is necessary to
Substantial amounts of human and material resources.The diagnostic mode of this switch breakdown has been not suitable with the fast-developing requirement of high-speed railway, such as
What fast and accurately judges that switch breakdown type is the important measure for ensureing traffic safety and passenger survival safety.
In the prior art, can not also automatic identification switch breakdown type, fast and effectively solution is not proposed yet.
The content of the invention
The invention provides a kind of switch breakdown recognition methods, at least to solve by artificial experience to judge in the prior art
Switch breakdown type, the problem of causing to fail to report and report by mistake.
A kind of switch breakdown recognition methods proposed by the present invention, comprises the following steps：
(1)：Gather each track switch operation curve of track switch；
(2)：Acquired each track switch operation curve is divided into normalized curve and damage curve；
(3)：For normalized curve and each class damage curve, the most representational curve of a feature is selected respectively for just
Chang represents curve and represents curve with each class failure；
(4)：Using similarity algorithm calculate the similarity 1 of curve to be identified and normal representation curve, curve to be identified with
Failure represents the similarity 2 of curve；Similarity algorithm is dynamic time warping algorithm or the algorithm based on Fu Leixie distances；
The similarity algorithm is dynamic time warping algorithm, is specially：
(4a1)：Curve to be identified is represented by T={ T (1), T (2) ... ..., T (n) ... ..., T (N) }, and n is time sequence
The sequential label of row, n=1 is time series starting point, and n=N is time series terminal, and T (n) is the value of the time series；
(4b1)：Normal representation curve and failure represent curve be represented by R=R (1), R (2) ... ..., R (m) ... ...,
R (M) }, m is the sequential label of time series, and m=1 is time series starting point, and m=M is time series terminal, and R (m) is described
The value of time series；
(4c1)：Each sequential label n of plot against time sequence to be identified is marked in transverse axis, is marked in the longitudinal axis and represents curve
Each sequential label m of time series, a net can be formed by drawing some co-ordinations by the rounded coordinate of these sequential labels
Network, all lattice points are followed successively by (1,1) ... ..., (n, m) ... ..., (N, M), and the optimal path of (N, M) is arrived in search (1,1)；
(4d1)：After path is by (n, m), next lattice point passed through can only be (n, m+1), (n+1, m), (n+1, m+
1), selection (n, m) is optimal path to the minimum range of next lattice point, calculates the accumulation minimum range that (1,1) arrives (N, M)；
(4e1)：Calculate plot against time sequence T to be identified and represent the Euclidean distance between plot against time sequence R；
(4f1)：Starting point (1,1) arrives the accumulation of terminal (N, M) to total accumulation distance of terminal (N, M) for starting point (1,1)
Minimum range, plot against time sequence T to be identified and represent the Euclidean distance sum between plot against time sequence R；
(4g1)：Total accumulation distance is negated, curve to be identified and normal representation curve or curve to be identified is represented
The similarity of curve is represented with failure；
Or：The similarity algorithm is the algorithm based on Fu Leixie distances, is specially：
(4a2)：Curve L to be identified1It is represented by P={ P (1), P (2) ... ..., P (n) ... ..., P (N) }, P (n)=
(xn,yn), n is curve L1On sampled point sequence number, n=1 be starting sample point, n=N be end sampled point, xnFor n-th
The abscissa of sampled point, xnFor the ordinate of n-th of sampled point；
(4b2)：Normal representation curve or failure represent curve L2It is represented by P '={ P ' (1), P ' (2) ... ..., P '
(m) ... ..., P ' (M) }, P ' (m)=(x '0,y′m), m is curve L2On sampled point sequence number, m=1 be starting sample point, m
=M is end sampled point, x 'mFor the abscissa of m-th of sampled point, y 'mFor the ordinate of m-th of sampled point；
(4c2)：Calculate L2Each sampled point is gone up to L2On the distance between each sampled point, obtain Distance matrix D as follows：
1≤m≤M, 1≤n≤N
In above formulaRepresent curve L2On m-th of sampled point to curve L2On
N-th of sampled point distance；
(4d2)：Select the ultimate range d in Distance matrix Dmax=max (D) and minimum range dmin=min (D), just
Beginningization target range f=dmin, and intercycle is set
(4e2)：Element in Distance matrix D less than or equal to f is set to 1, the element more than f is set to 0, so that
It is as follows to two values matrix D '：
1≤m≤M, 1≤n≤N
(4f2)：The path for meeting following condition in the middle search of two values matrix D ' one：The starting point in path is d '11, path
Terminal is d 'MN, path is passing through point d 'mnAfterwards, its it is next by point be only d 'm+1,n、d’m,n+1、d’m+1,n+1In one,
In path value a little be all necessary for 1；
(4g2)：If not finding the path of the condition of satisfaction in step (4f2), target range f=f+res is set, afterwards
Repeat step (4e2) and (4f2), if finding path or the target range f=d of the condition of satisfaction in step (4f2)max, then enter
Enter next step；
(4h2)：Curve and normal representation curve to be identified or failure represent Fu Leixie between curve apart from Frechet=
F,Represent that curve to be identified represents the similarity of curve with normal representation curve or failure；
(5)：Compare similarity obtained by calculation procedure (4), if similarity 1 is more than similarity 2, the curve is normal bent
Line, if similarity 1 is less than similarity 2, the curve is damage curve.
The track switch for gathering each operation curve of track switch in the present invention, described in step (1) to generate in microcomputer detecting system
Operation curve data or image, or be the track switch operation curve data or image in paper document.
It is track switch action current curve data or figure that each operation curve of track switch is gathered in the present invention, described in step (1)
Picture；Or be track switch operating power curve data or image.
In the present invention, step (2) is described to be divided into normalized curve and damage curve by acquired track switch operation curve, described
Damage curve is specifically divided into：The unexpected curve that stops operating, track switch accompany foreign matter song after start-up circuit broken string curve, switch starting
Line, goat stator rotor swinging cross curve, automatic lid actuator act dumb curve, goat start-up study curve, locking electricity
Flow exceeded curve and track switch action current is serrated curve；For each class damage curve, a feature is chosen most respectively
Representational curve as such damage curve representative curve；And calculate curve to be identified and each class failure respectively and represent song
The similarity of line；That a class ourve classification of similarity highest is the fault category of curve to be identified.
In the present invention, acquired track switch operation curve is pre-processed before step (3), comprised the following steps：
(1)：The average in colored track switch operation curve image between the R of each pixel, G, B component is taken as the pixel
Gray value, colored track switch operation curve image is transformed to gray level image；
(2)：The pixel value for setting a threshold value make it that gray value is more than the threshold value is 1, and gray value is less than the threshold value
Pixel value be 0, gray level image is transformed to bianry image；
(3)：The target area that reference axis is surrounded is found out, the isolated pixel that dissociates in target area is removed, and to song
The edge of line is smoothed, and removes noise；
(4)：Make each moment one value of correspondence, its pixel value is 0, be 0 to showing multiple pixels in the presence of one
Situation carry out micronization processes；
(5)：By functional transformation, each point coordinates on curve is extracted；
(6)：By each point coordinates bi-directional scaling, make each point transverse and longitudinal coordinate in same scope.
In the present invention, in step (3), for normalized curve, any one curve of selection is normal representation curve；For every
Any one curve is the representative curve of such damage curve in class damage curve, the such damage curve of selection.
It is to sum up shown, the beneficial effects of the present invention are：
(1) each operation curve of track switch is gathered in microcomputer detecting system, just be can recognize that without additionally installing other devices
Switch breakdown, economy is convenient, and practicality is stronger.
(2) acquired track switch operation curve is pre-processed, can not only eliminates the interference such as grid, noise, improve track switch
Fault Identification accuracy；The track switch from different microcomputer detecting systems, different Railway Bureaus, different weather can also be acted
Curve carries out Fault Identification so that the inventive method has a wide range of application, and is not limited to some small ranges, strong adaptability.
(3) dynamic time warping algorithm and based on Fu Leixie distance algorithms is used, it is not necessary to substantial amounts of historical data and specially
Family's knowledge base, only needs any selection to represent curve, so that it may recognize switch breakdown type, and reduction identification difficulty is reduced to correlation
The demand of professional.
(4) realize automatic identification switch breakdown, solve by artificial experience judge switch breakdown type belt come it is low
Efficiency and unreliability, have saved a large amount of manpower and materials, have improved judgment accuracy.
Brief description of the drawings
, below will be to specific in order to illustrate more clearly of the specific embodiment of the invention or technical scheme of the prior art
The accompanying drawing used required in embodiment or description of the prior art is briefly described, it should be apparent that, in describing below
Accompanying drawing is some embodiments of the present invention, for those of ordinary skill in the art, before creative work is not paid
Put, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is switch breakdown recognition methods flow chart according to embodiments of the present invention；
Fig. 2 is according to embodiments of the present invention 1 switch breakdown recognition methods flow chart；
Fig. 3 is a curve to be identified and each similarity histogram for representing curve according to embodiments of the present invention 1；
Fig. 4 is according to embodiments of the present invention 2 switch breakdown recognition methods flow chart；
Fig. 5 is selected just after being pre-processed to acquired track switch action current curve according to embodiments of the present invention 2
Chang represents curve and 8 kinds of failures represent curve image；Wherein：(a) it is normalized curve, (b) is start-up circuit broken string curve, (c)
For the curve that stopped operating suddenly after switch starting, (d) is that track switch accompanies foreign matter curve, and (e) is that the swinging cross of goat stator rotor is bent
Line, (f) is that automatic lid actuator acts dumb curve, and (g) is goat start-up study curve, and (h) is the exceeded song of locking electric current
Line, (i) is that track switch action current is serrated curve；
Fig. 6 is a curve to be identified and each similarity histogram for representing curve according to embodiments of the present invention 2.
Embodiment
Technical scheme is clearly and completely described below in conjunction with accompanying drawing, it is clear that described implementation
Example is a part of embodiment of the invention, rather than whole embodiments.Based on the embodiment in the present invention, ordinary skill
The every other embodiment that personnel are obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.
In the description of the invention, it is necessary to which explanation, term " first ", " second ", " the 3rd " are only used for describing purpose,
And it is not intended that indicating or implying relative importance.
As long as in addition, technical characteristic involved in invention described below different embodiments non-structure each other
It can just be combined with each other into conflict.
Embodiment 1
A kind of switch breakdown recognition methods is provided in the present embodiment, and Fig. 1 is track switch event according to embodiments of the present invention
Hinder recognition methods flow chart, as shown in figure 1, the flow chart comprises the following steps：
Step S11：Gather each operation curve of track switch；
Step S12：Acquired track switch operation curve is divided into normalized curve and damage curve；
Step S13：For normalized curve and damage curve, the most representational curve of one feature of selection is such curve
Representative curve；
Step S14：Similarity 1, the song to be identified of curve to be identified and normal representation curve are calculated using similarity algorithm
Line represents the similarity 2 of curve with failure；
Step S15：Compare calculating gained similarity, if similarity 1 is more than similarity 2, the curve is normalized curve,
If similarity 1 is less than similarity 2, the curve is damage curve.
By above-mentioned steps, the track switch operation curve fault category that automatic identification is collected, compared in the prior art,
By artificial experience judge switch breakdown type belt come poor efficiency and unreliability, above-mentioned steps solve in the prior art,
Switch breakdown type is judged by artificial experience, the problem of causing to fail to report and report by mistake, it is achieved thereby that track switch automatic identification failure
Classification, improves overhaul efficiency and system reliability, it is ensured that traffic safety.
Fig. 2 is according to embodiments of the present invention 1 switch breakdown recognition methods flow chart, as shown in Fig. 2 this method is included such as
Lower step：The each operating power curve of track switch is gathered in microcomputer detecting system；By acquired track switch operating power curve point
Class, can be divided into normalized curve and damage curve, and the damage curve can be further divided into：Start-up circuit broken string curve, track switch are opened
Stopped operating suddenly after dynamic curve, track switch accompanies that foreign matter curve, automatic lid actuator act dumb curve and locking electric current is exceeded
Curve；For each class track switch operating power curve, any one such track switch operating power curve of selection is that such power is bent
The representative curve of line；Using dynamic time warping algorithm, power curve to be identified and normal representation curve and all kinds of failures are calculated
Represent the distance of power curve；Negated to calculating gained distance, represent curve to be identified and normal representation curve and all kinds of failures
Represent the similarity of curve；Compare the similarity that curve to be identified represents curve with normal representation curve and all kinds of failures, it is similar
That a class ourve classification of degree highest is the classification of curve to be identified.
Illustrated with reference to a specific alternative embodiment.
(1)：The each operating power curve of track switch is gathered in microcomputer detecting system；
(2)：By acquired track switch operating power curve classification, normalized curve and damage curve, the failure can be divided into
Curve can be further divided into：The unexpected curve that stops operating, track switch accompany foreign matter song after start-up circuit broken string curve, switch starting
Line, automatic lid actuator act dumb curve and the exceeded curve of locking electric current；
(3)：For each class track switch operating power curve, any one such track switch operating power curve of selection is this
The representative curve of class power curve；
(4)：Using dynamic time warping algorithm, power curve to be identified and normal representation curve and all kinds of failures are calculated
The distance of power curve is represented, step is as follows：
(1) curve to be identified is taken, T={ T (1), T (2) ... ..., T (219) } is represented by, T (1)=0, T (2)=
1.84684 ... ..., T (219)=2.29189；
(2) normal representation curve is represented by R={ R (1), R (2) ... ..., R (144) }, and R (1)=0, R (2)=
5.435897 ... ..., R (144)=0.090598；
(3) each sequential label 219 of plot against time sequence to be identified is marked in transverse axis, when the longitudinal axis marks and represents curve
Between sequence each sequential label 144, a net can be formed by drawing some co-ordinations by the rounded coordinate of these sequential labels
Network, all lattice points are followed successively by (1,1) ... ..., (219,144), and the optimal path of (219,144) is arrived in search (1,1)；
(4)：After path is by (1,1), next lattice point passed through can only be (1,2), (2,1), (2,2), can be calculated
(1,1) the accumulation minimum range to (219,144) is 85.78232；
(5)：It can be calculated the Euclidean distance between plot against time sequence T to be identified and normal representation plot against time sequence R
For 0.28344；
(6)：Total accumulation distance of starting point (1,1) to terminal (219,144) is 86.06576；
It is same as above with distance method that all kinds of failures represent power curve to calculate power curve to be identified, and gained is to be identified
Power curve accompanies foreign matter curve with the curve that stopped operating suddenly after start-up circuit broken string curve, switch starting, track switch, opened automatically
Close that device acts dumb curve and the distance of the exceeded curve of locking electric current is respectively：886.71484、87.18578、1.00232、
103.44763、140.06902。
(5)：The distance point of power curve is represented with all kinds of failures to power curve to be identified and normal representation power curve
Do not negate as 0.01162,0.00112,0.01147,0.99769,0.00967,0.00714, represent curve to be identified with it is normal
The curve that stops operating suddenly after curve, start-up circuit broken string curve, switch starting, track switch is represented to accompany foreign matter curve, open automatically
Close device act dumb curve and the exceeded curve of locking electric current similarity be respectively 0.01162,0.00112,0.01147,
0.99769、0.00967、0.00714；
(6)：Compare the similarity that curve to be identified represents curve with normal representation curve and all kinds of failures, must can wait to know
Other curve accompanies foreign matter curve similarity highest with track switch, then curve category to be identified is that track switch accompanies foreign matter curve.
Fig. 3 is a curve to be identified and each similarity histogram for representing curve according to embodiments of the present invention 1, from figure
As can be seen that utilizing dynamic time warping algorithm in 3, the similar of curve to be identified and normalized curve and 5 kinds of damage curves is calculated
Degree, curve to be identified accompanies the similarity highest of foreign matter failure with track switch, therefore judges the fault type of curve to be identified for road
Trouble accompanies foreign matter.Empirical tests, judged result is correct.
Embodiment 2
A kind of switch breakdown recognition methods is also provided in the present embodiment.
Fig. 4 is according to embodiments of the present invention 2 switch breakdown recognition methods flow chart, as shown in figure 4, this method is included such as
Lower step：The each action current curve of track switch is gathered in microcomputer detecting system；By acquired track switch action current curve point
Class, can be divided into normalized curve and damage curve, and the damage curve can be further divided into：Start-up circuit broken string curve, track switch are opened
Stopped operating suddenly after dynamic curve, track switch accompanies foreign matter curve, goat stator rotor swinging cross curve, automatic lid actuator action not
Flexible curve, goat start-up study curve, the exceeded curve of locking electric current and track switch action current are serrated curve；Take colour
Average in track switch action current curve image between the R of each pixel, G, B component as the pixel gray value, by coloured silk
Color channel trouble action current curve image is transformed to gray level image；A threshold value is set to cause gray value to be more than the pixel of the threshold value
Value is 1, and the pixel value that gray value is less than the threshold value is 0, and gray level image is transformed into bianry image；Find out reference axis institute
The target area surrounded, removes the isolated pixel that dissociates in target area, and the edge of curve is smoothed, and removes
Noise；Make each moment one value of correspondence, its pixel value is 0, to there is a situation where that show multiple pixels enters for 0
Row micronization processes；By functional transformation, each point coordinates on curve is extracted；By each point coordinates bi-directional scaling, sit each point transverse and longitudinal
It is marked in same scope；For each class track switch action current curve, arbitrarily one such track switch action current curve of selection is
The representative curve of such current curve；It is bent with normal representation using current curve to be identified based on Fu Leixie distance algorithms, is calculated
Line and all kinds of failures represent the distance of current curve；Negated to calculating gained distance, represent that curve to be identified is bent with normal representation
Line and all kinds of failures represent the similarity of curve；Compare curve to be identified and represent curve with normal representation curve and all kinds of failures
Similarity, that a class ourve classification of similarity highest is the classification of curve to be identified.
Illustrated with reference to another specific alternative embodiment.
(1)：The each action current curve of track switch is gathered in microcomputer detecting system；
(2)：By acquired track switch action current curve classification, normalized curve and damage curve, the failure can be divided into
Curve can be further divided into：The unexpected curve that stops operating, track switch accompany foreign matter song after start-up circuit broken string curve, switch starting
Line, goat stator rotor swinging cross curve, automatic lid actuator act dumb curve, goat start-up study curve, locking electricity
Flow exceeded curve and track switch action current is serrated curve；
(3)：The average in colored track switch action current curve image between the R of each pixel, G, B component is taken to be used as this
The gray value of pixel, gray level image is transformed to by colored track switch action current curve image；
(4)：The pixel value for setting a threshold value make it that gray value is more than the threshold value is 1, and gray value is less than the threshold value
Pixel value be 0, gray level image is transformed to bianry image；
(5)：The target area that reference axis is surrounded is found out, the isolated pixel that dissociates in target area is removed, and to song
The edge of line is smoothed, and removes noise；
(6)：Make each moment one value of correspondence, its pixel value is 0, be 0 to showing multiple pixels in the presence of one
Situation carry out micronization processes；
(7)：By functional transformation, each point coordinates on curve is extracted；
(8)：By each point coordinates bi-directional scaling, make each point transverse and longitudinal coordinate in same scope
(9)：For each class track switch operating power curve, any one such track switch action current curve of selection is this
The representative curve of class current curve；
(10)：Using based on Fu Leixie distance algorithms, calculating current curve to be identified and normal representation curve and all kinds of events
Barrier represents the distance of current curve；
(1) fault type is taken to start the curve stopped operating suddenly as curve L to be identified1It is represented by P=
{ P (1), P (2) ... ..., P (97) }, P (1)=(0,0), P (1)=(0.0104,1) ... ..., P (1)=(1,0.0061), P
(1) it is starting sample point, P (97) is end sampled point；
(2) using normalized curve as representing curve L2It is represented by P '={ P ' (1), P ' (2) ... ..., P ' (M) }, P ' (1)
=(0,0), P ' (2)=(0.0015,0.1372) ... ..., P ' (654)=(1,0.0020), P ' (1) is starting sample point, P '
(654) it is end sampled point,；
(3) L is calculated1Each sampled point is gone up to L2On the distance between each sampled point, obtain Distance matrix D as follows：
1≤m≤654,1≤n≤97
In above formulaRepresent curve L2On m-th of sampled point arrive
Curve L1On n-th of sampled point distance.
(4) the ultimate range d in Distance matrix D is found outmax=max (D)=1.4054 and minimum range dmin=min
(D)=0, initialized target is apart from f=dmin=0, and intercycle is set
(5) element in Distance matrix D less than or equal to f is set to 1, the element more than f is set to 0, so as to obtain
Two values matrix D ' is as follows：
1≤m≤654,1≤n≤97
(6) path for meeting following condition in the middle search of two values matrix D ' one：The starting point in path is d '11, the end in path
Point is d 'MN；Path is passing through point d 'mnAfterwards, its it is next by point be only d 'M+1, n、d′M, n+1、d′M+1, n+1In one；Road
In footpath value a little be all necessary for 1.
(7) if not finding the path of the condition of satisfaction in step 8f, target range f=f+res=f+0.0141 is set,
Repeat step 8e and 8f afterwards；If found in step 8f the condition of satisfaction path or target away from f=dm xFrom then into next
Step.
By (5), (6), (7) three steps cycle calculations, in f=0.5059 condition, in 8f, this step finds one for we
Bar meets the path of condition, then jumps out this and is recycled into step (8).
(8) calculated more than, we can obtain the discrete Fu Leixie distances between curve to be identified and normalized curve
Frechet=f=0.5059, then curve and normalized curve to be identified similarity
By identical method, we can obtain the unexpected curve that stops operating after curve to be identified and switch starting, turn
Rut machine rotor confusion curve, track switch are added with foreign matter curve, start-up circuit broken string curve, the exceeded curve of locking electric current, goat
Track switch action current is serrated curve, goat start delay curve, the Fu Leixie distances of the dumb curve of automatic lid actuator
Respectively 0.1266,0.8278,0.9777,0.8406,0.4908,0.4907,0.5030,0.8416.Curve and road to be identified
The chaotic curve of the unexpected curve that stops operating, goat rotor, track switch are bent added with foreign matter curve, start-up circuit broken string after trouble startup
Line, the exceeded curve of locking electric current, goat track switch action current be serrated curve, goat start delay curve, open automatically
The similarity for closing the dumb curve of device is respectively 7.8963,1.2079,1.02277,1.1896,2.0372,2.0378,
1.9878、1.1881.Therefore, the similarity highest of curve to be identified and the curve that stopped operating suddenly after switch starting, be
7.8963, the i.e. fault type of the curve to be identified be switch starting after stop operating suddenly.
(11)：Negated to calculating gained distance, represent that curve to be identified is represented with normal representation curve and all kinds of failures
The similarity of curve；
(12)：Compare the similarity that curve to be identified represents curve with normal representation curve and all kinds of failures, similarity
That a class ourve classification of highest is the classification of curve to be identified.
Fig. 5 is selected just after being pre-processed to acquired track switch action current curve according to embodiments of the present invention 2
Chang represents curve and 8 kinds of failures represent curve image.
Fig. 6 is a curve to be identified and each similarity histogram for representing curve according to embodiments of the present invention 2, from figure
As can be seen that using Fu Leixie distance algorithms are based on, calculating the phase of curve to be identified and normalized curve and 8 kinds of damage curves in 6
Like degree, curve to be identified and the similarity highest stopped operating suddenly after switch starting, therefore judge the failure of curve to be identified
Type be switch starting after stop operating suddenly.Empirical tests, judged result is correct.
It should be understood by those skilled in the art that, embodiments of the invention can be provided as method, system or computer program
Product.Therefore, the present invention can be using the reality in terms of complete hardware embodiment, complete software embodiment or combination software and hardware
Apply the form of example.Moreover, the present invention can be used in one or more computers for wherein including computer usable program code
The computer program production that usable storage medium is implemented on (including but is not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of product.
The present invention is the flow with reference to method according to embodiments of the present invention, equipment (system) and computer program product
Figure and/or block diagram are described.It should be understood that can be by every first-class in computer program instructions implementation process figure and/or block diagram
Journey and/or the flow in square frame and flow chart and/or block diagram and/or the combination of square frame.These computer programs can be provided
The processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that produced by the instruction of computer or the computing device of other programmable data processing devices for real
The device for the function of being specified in present one flow of flow chart or one square frame of multiple flows and/or block diagram or multiple square frames.
These computer program instructions, which may be alternatively stored in, can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that the instruction being stored in the computer-readable memory, which is produced, to be included referring to
Make the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one square frame of block diagram or
The function of being specified in multiple square frames.
These computer program instructions can be also loaded into computer or other programmable data processing devices,
So that performing series of operation steps on computer or other programmable devices to produce computer implemented place
Reason, so that the instruction performed on computer or other programmable devices is provided for realizing in one flow or multiple of flow chart
The step of function of being specified in one square frame of flow and/or block diagram or multiple square frames.
Obviously, above-described embodiment is only intended to clearly illustrate example, and the not restriction to embodiment.It is right
For those of ordinary skill in the art, can also make on the basis of the above description it is other it is various forms of change or
Change.There is no necessity and possibility to exhaust all the enbodiments.And the obvious change thus extended out or
Among changing still in the protection domain of the invention.
Claims (6)
1. a kind of switch breakdown recognition methods, it is characterised in that comprise the following steps：
(1)：Gather each operation curve of track switch；
(2)：Acquired each track switch operation curve is divided into normalized curve and damage curve；
(3)：For each class normalized curve and damage curve, it is such song that the most representational curve of a feature is selected respectively
The normal representation curve of line represents curve with failure；
(4)：Similarity 1, curve to be identified and the failure of curve to be identified and normal representation curve are calculated using similarity algorithm
Represent the similarity 2 of curve；Similarity algorithm is dynamic time warping algorithm or the algorithm based on Fu Leixie distances；
The similarity algorithm is dynamic time warping algorithm, is specially：
(4a1)：Curve to be identified is represented by T={ T (1), T (2) ... ..., T (n) ... ..., T (N) }, and n is time series
Sequential label, n=1 is time series starting point, and n=N is time series terminal, and T (n) is the value of the time series；
(4b1)：Normal representation curve represents curve with failure and is represented by R={ R (1), R (2) ... ..., R (m) ... ..., R
(M) }, m is the sequential label of time series, and m=1 is time series starting point, and m=M is time series terminal, when R (m) is described
Between sequence value；
(4c1)：Each sequential label n of plot against time sequence to be identified is marked in transverse axis, is marked in the longitudinal axis and represents plot against time
Each sequential label m of sequence, a network, institute can be formed by drawing some co-ordinations by the rounded coordinate of these sequential labels
There is lattice point to be followed successively by (1,1) ... ..., the optimal path of (N, M) is arrived in (n, m) ... ..., (N, M), search (1,1)；
(4d1)：After path is by (n, m), next lattice point passed through can only be (n, m+1), (n+1, m), (n+1, m+1), choosing
The minimum range for selecting (n, m) to next lattice point is optimal path, calculates the accumulation minimum range that (1,1) arrives (N, M)；
(4e1)：Calculate plot against time sequence T to be identified and represent the Euclidean distance between plot against time sequence R；
(4f1)：Total accumulation distance of starting point (1,1) to terminal (N, M) is minimum for the accumulation of starting point (1,1) to terminal (N, M)
Distance, plot against time sequence T to be identified and represent the Euclidean distance sum between plot against time sequence R；
(4g1)：Total accumulation distance is negated, curve to be identified and normal representation curve or curve to be identified and event is represented
Barrier represents the similarity of curve；
Or：The similarity algorithm is the algorithm based on Fu Leixie distances, is specially：
(4a2)：Curve L to be identified1It is represented by P={ P (1), P (2) ... ..., P (n) ... ..., P (N) }, P (n)=(xn,
yn), n is curve L1On sampled point sequence number, n=1 be starting sample point, n=N be end sampled point, xnFor n-th of sampling
The abscissa of point, xnFor the ordinate of n-th of sampled point；
(4b2)：Normal representation curve or failure represent curve L2It is represented by P '={ P ' (1), P ' (2) ... ..., P '
(m) ... ..., P ' (M) }, P ' (m)=(x 'm,y′m), m is curve L2On sampled point sequence number, m=1 be starting sample point, m
=M is end sampled point, x 'mFor the abscissa of m-th of sampled point, y 'mFor the ordinate of m-th of sampled point；
(4c2)：Calculate L2Each sampled point is gone up to L2On the distance between each sampled point, obtain Distance matrix D as follows：
1≤m≤M, 1≤n≤N
In above formulaRepresent curve L2On m-th of sampled point to curve L2On
The distance of n sampled point；
(4d2)：Select the ultimate range d in Distance matrix Dmax=max (D) and minimum range dmin=min (D), initialization
Target range f=dmin, and intercycle is set
(4e2)：Element in Distance matrix D less than or equal to f is set to 1, the element more than f is set to 0, so as to obtain two
Value matrix D ' is as follows：
1≤m≤M, 1≤n≤N
(4f2)：The path for meeting following condition in the middle search of two values matrix D ' one：The starting point in path is d '11, the terminal in path
For d 'MN, path is passing through point d 'mnAfterwards, its it is next by point be only d 'm+1,n、d’m,n+1、d’m+1,n+1In one, path
Middle value a little is all necessary for 1；
(4g2)：If not finding the path of the condition of satisfaction in step (4f2), target range f=f+res is set, repeated afterwards
Step (4e2) and (4f2), if finding path or the target range f=d of the condition of satisfaction in step (4f2)max, then under entering
One step；
(4h2)：Curve and normal representation curve to be identified or failure represent Fu Leixie between curve apart from Frechet=f,Represent that curve to be identified represents the similarity of curve with normal representation curve or failure；
(5)：Compare similarity obtained by calculation procedure (4), if similarity 1 is more than similarity 2, the curve is normalized curve,
If similarity 1 is less than similarity 2, the curve is damage curve.
2. switch breakdown recognition methods according to claim 1, it is characterised in that track switch is gathered described in step (1) every
Secondary operation curve is the track switch operation curve data that generate or image in microcomputer detecting system, or moved for the track switch in paper document
Make curve data or image.
3. switch breakdown recognition methods according to claim 1, it is characterised in that track switch is gathered described in step (1) every
Secondary operation curve is track switch action current curve data or image；Or be track switch operating power curve data or image.
4. switch breakdown recognition methods according to claim 1, it is characterised in that step (2) is described by acquired road
Trouble operation curve is divided into normalized curve and damage curve, and the damage curve is specifically divided into：Start-up circuit broken string curve, track switch are opened
Stopped operating suddenly after dynamic curve, track switch accompanies foreign matter curve, goat stator rotor swinging cross curve, automatic lid actuator action not
Flexible curve, goat start-up study curve, the exceeded curve of locking electric current and track switch action current are serrated curve；For every
One class damage curve, chooses the most representational curve of a feature as the representative curve of such damage curve respectively；And point
Curve to be identified is not calculated represents the similarity of curve with each class failure；That a class ourve classification of similarity highest is to treat
Recognize the fault category of curve.
5. switch breakdown recognition methods according to claim 1, it is characterised in that to acquired road before step (3)
Trouble operation curve is pre-processed, and is comprised the following steps：
(1)：The average in colored track switch operation curve image between the R of each pixel, G, B component is taken as the ash of the pixel
Angle value, gray level image is transformed to by colored track switch operation curve image；
(2)：The pixel value for setting a threshold value make it that gray value is more than the threshold value is 1, and gray value is less than the picture of the threshold value
Vegetarian refreshments value is 0, and gray level image is transformed into bianry image；
(3)：The target area that reference axis is surrounded is found out, the isolated pixel that dissociates in target area is removed, and to curve
Edge is smoothed, and removes noise；
(4)：Make each moment one value of correspondence, its pixel value is 0, to showing the feelings that multiple pixels are 0 in the presence of one
Condition carries out micronization processes；
(5)：By functional transformation, each point coordinates on curve is extracted；
(6)：By each point coordinates bi-directional scaling, make each point transverse and longitudinal coordinate in same scope.
6. switch breakdown recognition methods according to claim 1, it is characterised in that in step (3), for normalized curve,
Any one curve of selection is normal representation curve；For every class damage curve, any one song in such damage curve is selected
Line is the representative curve of such damage curve.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710341307.1A CN107203746B (en) | 2017-05-12 | 2017-05-12 | Turnout fault identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710341307.1A CN107203746B (en) | 2017-05-12 | 2017-05-12 | Turnout fault identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107203746A true CN107203746A (en) | 2017-09-26 |
| CN107203746B CN107203746B (en) | 2020-06-19 |
Family
ID=59905102
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710341307.1A Active CN107203746B (en) | 2017-05-12 | 2017-05-12 | Turnout fault identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107203746B (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108182386A (en) * | 2017-12-13 | 2018-06-19 | 河南辉煌科技股份有限公司 | Track switch standard curve automatic generation method |
| CN108238066A (en) * | 2017-12-22 | 2018-07-03 | 同济大学 | Track switch operation curve template choosing method and its application |
| CN108256738A (en) * | 2017-12-22 | 2018-07-06 | 同济大学 | Track switch action reference curve choosing method and its application |
| CN108287273A (en) * | 2017-12-13 | 2018-07-17 | 河南辉煌科技股份有限公司 | Acclerating road switch jamming failure judgment method based on current curve |
| CN108491855A (en) * | 2018-02-08 | 2018-09-04 | 同济大学 | A kind of signal trouble recognition methods |
| CN108985380A (en) * | 2018-07-25 | 2018-12-11 | 西南交通大学 | A kind of goat fault recognition method based on clustering ensemble |
| CN109532949A (en) * | 2018-12-17 | 2019-03-29 | 深圳科安达电子科技股份有限公司 | Railway switch conversion process analyzes assessment system |
| CN109677448A (en) * | 2018-12-29 | 2019-04-26 | 广州铁科智控有限公司 | A kind of switch breakdown analysis method based on goat power |
| CN110321583A (en) * | 2018-03-28 | 2019-10-11 | 西门子股份公司 | The method and apparatus of failure identification for technological system |
| CN110406560A (en) * | 2018-04-30 | 2019-11-05 | 西门子股份公司 | Method and apparatus for the fault detection in railway points system |
| CN110441629A (en) * | 2019-07-29 | 2019-11-12 | 交控科技股份有限公司 | Method for diagnosing faults and device based on point machine action current curve |
| CN110728300A (en) * | 2019-09-09 | 2020-01-24 | 交控科技股份有限公司 | Method and system for identifying fault type based on turnout action current curve |
| CN110749785A (en) * | 2019-09-10 | 2020-02-04 | 交控科技股份有限公司 | Turnout rotation time sequence analysis method and system |
| CN110751853A (en) * | 2019-10-25 | 2020-02-04 | 百度在线网络技术（北京）有限公司 | Parking space data validity identification method and device |
| WO2020057402A1 (en) * | 2018-09-20 | 2020-03-26 | 西门子（中国）有限公司 | System-state monitoring method and device and storage medium |
| CN110954734A (en) * | 2018-09-26 | 2020-04-03 | 北京国铁盛阳技术有限公司 | Fault diagnosis method, device, equipment and storage medium |
| CN111105147A (en) * | 2019-12-02 | 2020-05-05 | 北京交通大学 | Turnout health state assessment method based on dynamic time warping |
| CN111915061A (en) * | 2020-07-06 | 2020-11-10 | 同济大学 | Switch action current curve prediction method and fault discrimination method thereof |
| CN112213579A (en) * | 2020-09-24 | 2021-01-12 | 交控科技股份有限公司 | Method and device for identifying faults of turnout switch machine |
| CN112345941A (en) * | 2020-11-05 | 2021-02-09 | 惠州市蓝微新源技术有限公司 | Background thermal runaway early warning method based on big data and variable quantity curve |
| CN114118485A (en) * | 2022-01-25 | 2022-03-01 | 中国空气动力研究与发展中心设备设计与测试技术研究所 | Health monitoring and preventive maintenance method for tail boom system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DD238219A1 (en) * | 1985-08-17 | 1986-08-13 | Zentralstelle Sicherungs Und F | DIAGNOSTIC CIRCUIT ARRANGEMENT FOR MONITORING POINT SOFT LEADING MEANS |
| CN102381340A (en) * | 2010-08-27 | 2012-03-21 | 卡斯柯信号有限公司 | Power curve-based speed-raising turnout fault information processing method and device |
| CN103129584A (en) * | 2011-11-30 | 2013-06-05 | 国际商业机器公司 | Turnout monitoring method and system, railway line management method and system |
| CN104811991A (en) * | 2015-04-17 | 2015-07-29 | 合肥工业大学 | Wireless link quality predicting method based on dynamic time warping algorithm |
| CN105260595A (en) * | 2015-04-02 | 2016-01-20 | 北京交通大学 | Feature extraction method for switch action current curve and switch fault diagnosis method |
| CN105787511A (en) * | 2016-02-26 | 2016-07-20 | 清华大学 | Track switch fault diagnosis method and system based on support vector machine |
| CN106124885A (en) * | 2016-06-13 | 2016-11-16 | 四川网达科技有限公司 | Switch breakdown detection apparatus and method |
-
2017
- 2017-05-12 CN CN201710341307.1A patent/CN107203746B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DD238219A1 (en) * | 1985-08-17 | 1986-08-13 | Zentralstelle Sicherungs Und F | DIAGNOSTIC CIRCUIT ARRANGEMENT FOR MONITORING POINT SOFT LEADING MEANS |
| CN102381340A (en) * | 2010-08-27 | 2012-03-21 | 卡斯柯信号有限公司 | Power curve-based speed-raising turnout fault information processing method and device |
| CN103129584A (en) * | 2011-11-30 | 2013-06-05 | 国际商业机器公司 | Turnout monitoring method and system, railway line management method and system |
| CN105260595A (en) * | 2015-04-02 | 2016-01-20 | 北京交通大学 | Feature extraction method for switch action current curve and switch fault diagnosis method |
| CN104811991A (en) * | 2015-04-17 | 2015-07-29 | 合肥工业大学 | Wireless link quality predicting method based on dynamic time warping algorithm |
| CN105787511A (en) * | 2016-02-26 | 2016-07-20 | 清华大学 | Track switch fault diagnosis method and system based on support vector machine |
| CN106124885A (en) * | 2016-06-13 | 2016-11-16 | 四川网达科技有限公司 | Switch breakdown detection apparatus and method |
Non-Patent Citations (5)
| Title |
|---|
| FENFANG ZHOU ET AL: "Fault diagnosis of high-speed railway turnout based on support vector machine", 《 2016 IEEE INTERNATIONAL CONFERENCE ON INDUSTRIAL TECHNOLOGY (ICIT)》 * |
| KAI ZHANG ET AL: "The railway turnout fault diagnosis algorithm based on BP neural network", 《2014 IEEE INTERNATIONAL CONFERENCE ON CONTROL SCIENCE AND SYSTEMS ENGINEERING》 * |
| V. ATAMURADOV ET AL: "Failure diagnostics for railway point machines using expert systems", 《 2009 IEEE INTERNATIONAL SYMPOSIUM ON DIAGNOSTICS FOR ELECTRIC MACHINES, POWER ELECTRONICS AND DRIVES》 * |
| 张星: "基于相似度的道岔故障电流曲线识别方法研究与实现", 《万方数据》 * |
| 张星等: "基于支持向量机的道岔故障动作电流识别", 《电脑知识与技术》 * |
Cited By (34)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108182386A (en) * | 2017-12-13 | 2018-06-19 | 河南辉煌科技股份有限公司 | Track switch standard curve automatic generation method |
| CN108287273A (en) * | 2017-12-13 | 2018-07-17 | 河南辉煌科技股份有限公司 | Acclerating road switch jamming failure judgment method based on current curve |
| CN108182386B (en) * | 2017-12-13 | 2021-12-17 | 河南辉煌科技股份有限公司 | Automatic generation method of turnout standard curve |
| CN108287273B (en) * | 2017-12-13 | 2020-07-24 | 河南辉煌科技股份有限公司 | Current curve-based speed-up turnout jamming fault judgment method |
| CN108238066A (en) * | 2017-12-22 | 2018-07-03 | 同济大学 | Track switch operation curve template choosing method and its application |
| CN108256738A (en) * | 2017-12-22 | 2018-07-06 | 同济大学 | Track switch action reference curve choosing method and its application |
| CN108256738B (en) * | 2017-12-22 | 2021-12-14 | 同济大学 | Turnout action reference curve selection method and application thereof |
| CN108491855A (en) * | 2018-02-08 | 2018-09-04 | 同济大学 | A kind of signal trouble recognition methods |
| CN110321583A (en) * | 2018-03-28 | 2019-10-11 | 西门子股份公司 | The method and apparatus of failure identification for technological system |
| CN110406560A (en) * | 2018-04-30 | 2019-11-05 | 西门子股份公司 | Method and apparatus for the fault detection in railway points system |
| CN108985380B (en) * | 2018-07-25 | 2021-08-03 | 西南交通大学 | Point switch fault identification method based on cluster integration |
| CN108985380A (en) * | 2018-07-25 | 2018-12-11 | 西南交通大学 | A kind of goat fault recognition method based on clustering ensemble |
| US11847619B2 (en) | 2018-09-20 | 2023-12-19 | Siemens Ltd., China | System-state monitoring method and device and storage medium |
| WO2020057402A1 (en) * | 2018-09-20 | 2020-03-26 | 西门子（中国）有限公司 | System-state monitoring method and device and storage medium |
| CN110928741A (en) * | 2018-09-20 | 2020-03-27 | 西门子（中国）有限公司 | System state monitoring method, device and storage medium |
| CN110928741B (en) * | 2018-09-20 | 2021-09-24 | 西门子（中国）有限公司 | System state monitoring method, device and storage medium |
| CN110954734B (en) * | 2018-09-26 | 2022-03-25 | 北京国铁盛阳技术有限公司 | Fault diagnosis method, device, equipment and storage medium |
| CN110954734A (en) * | 2018-09-26 | 2020-04-03 | 北京国铁盛阳技术有限公司 | Fault diagnosis method, device, equipment and storage medium |
| CN109532949B (en) * | 2018-12-17 | 2024-05-28 | 深圳科安达电子科技股份有限公司 | Railway switch conversion process analysis and evaluation system |
| CN109532949A (en) * | 2018-12-17 | 2019-03-29 | 深圳科安达电子科技股份有限公司 | Railway switch conversion process analyzes assessment system |
| CN109677448A (en) * | 2018-12-29 | 2019-04-26 | 广州铁科智控有限公司 | A kind of switch breakdown analysis method based on goat power |
| CN110441629B (en) * | 2019-07-29 | 2022-02-15 | 交控科技股份有限公司 | Fault diagnosis method and device based on turnout switch machine operating current curve |
| CN110441629A (en) * | 2019-07-29 | 2019-11-12 | 交控科技股份有限公司 | Method for diagnosing faults and device based on point machine action current curve |
| CN110728300A (en) * | 2019-09-09 | 2020-01-24 | 交控科技股份有限公司 | Method and system for identifying fault type based on turnout action current curve |
| CN110749785A (en) * | 2019-09-10 | 2020-02-04 | 交控科技股份有限公司 | Turnout rotation time sequence analysis method and system |
| CN110751853A (en) * | 2019-10-25 | 2020-02-04 | 百度在线网络技术（北京）有限公司 | Parking space data validity identification method and device |
| CN111105147A (en) * | 2019-12-02 | 2020-05-05 | 北京交通大学 | Turnout health state assessment method based on dynamic time warping |
| CN111105147B (en) * | 2019-12-02 | 2021-12-21 | 北京交通大学 | Turnout health state assessment method based on dynamic time warping |
| CN111915061A (en) * | 2020-07-06 | 2020-11-10 | 同济大学 | Switch action current curve prediction method and fault discrimination method thereof |
| CN111915061B (en) * | 2020-07-06 | 2022-06-07 | 同济大学 | Switch action current curve prediction method and fault discrimination method thereof |
| CN112213579A (en) * | 2020-09-24 | 2021-01-12 | 交控科技股份有限公司 | Method and device for identifying faults of turnout switch machine |
| CN112345941A (en) * | 2020-11-05 | 2021-02-09 | 惠州市蓝微新源技术有限公司 | Background thermal runaway early warning method based on big data and variable quantity curve |
| CN114118485A (en) * | 2022-01-25 | 2022-03-01 | 中国空气动力研究与发展中心设备设计与测试技术研究所 | Health monitoring and preventive maintenance method for tail boom system |
| CN114118485B (en) * | 2022-01-25 | 2022-04-08 | 中国空气动力研究与发展中心设备设计与测试技术研究所 | Health monitoring and preventive maintenance method for tail boom system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107203746B (en) | 2020-06-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107203746A (en) | A kind of switch breakdown recognition methods | |
| CN110441629B (en) | Fault diagnosis method and device based on turnout switch machine operating current curve | |
| CN105426905B (en) | Robot barrier object recognition methods based on histogram of gradients and support vector machines | |
| CN106919978A (en) | A kind of high ferro contact net support meanss parts recognition detection method | |
| CN105719305B (en) | Component falls off defect identification method and system in contact net | |
| CN109446920B (en) | Method for detecting passenger crowding degree of urban rail transit based on convolutional neural network | |
| CN109670404A (en) | A kind of road ponding image detection method for early warning based on mixed model | |
| CN105404867B (en) | A kind of substation isolating-switch state identification method of view-based access control model | |
| CN106198551A (en) | Method and device for detecting defects of power transmission line | |
| CN111695506B (en) | Early warning method and system for wind-induced foreign matter short-circuit fault of transmission line | |
| CN110111581A (en) | Target identification method, device, computer equipment and storage medium | |
| CN108898085A (en) | Intelligent road disease detection method based on mobile phone video | |
| CN113077005B (en) | Anomaly detection system and method based on LSTM self-encoder and normal signal data | |
| CN107194396A (en) | Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system | |
| CN102679935B (en) | System and method for calculating icing thickness of power transmission line | |
| CN103400379B (en) | The detection of a kind of icing based on image and thickness method of estimation | |
| CN102332094B (en) | Semi-supervised online study face detection method | |
| CN111046583A (en) | Switch machine fault diagnosis method based on DTW algorithm and ResNet network | |
| CN108256738A (en) | Track switch action reference curve choosing method and its application | |
| CN105354831A (en) | Multi-defect detection method based on image block variance-weighting eigenvalues | |
| CN108846331A (en) | The video frequency identifying method whether a kind of EMU chassis screw fastener falls off | |
| WO2012005461A2 (en) | Method for automatically calculating information on clouds | |
| CN108804539A (en) | A kind of track method for detecting abnormality under time and space double-visual angle | |
| CN113641486B (en) | Intelligent turnout fault diagnosis method based on edge computing network architecture | |
| CN103824092A (en) | Image classification method for monitoring state of electric transmission and transformation equipment on line |
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 |