CN108334908B - Method and device for detecting railway rail damage - Google Patents
Method and device for detecting railway rail damage Download PDFInfo
- Publication number
- CN108334908B CN108334908B CN201810186643.8A CN201810186643A CN108334908B CN 108334908 B CN108334908 B CN 108334908B CN 201810186643 A CN201810186643 A CN 201810186643A CN 108334908 B CN108334908 B CN 108334908B
- Authority
- CN
- China
- Prior art keywords
- rail
- detection
- damage
- detection probability
- probability
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
Abstract
The invention provides a method and a device for detecting rail damage of a railway, which are used for calculating the detection capability of a flaw detection vehicle on various damages, wherein the method for detecting the rail damage of the railway comprises the following steps: acquiring original data of rail damage of a railway; establishing detection probability models of different damage types according to a logistic regression method and a machine learning method; determining fitting parameters in the detection probability model by using an MLE maximum probability estimation method according to the original data of the railway steel rail damage; and determining the detection probability of different damage types or the speed range of the steel rail flaw detection vehicle according to the determined detection probability models of different damage types and the current speed of the steel rail flaw detection vehicle or the set detection probability of the damage types.
Description
Technical Field
The invention relates to the field of steel rail flaw detection, in particular to a method and a device for detecting railway steel rail flaws.
Background
In the steel rail flaw detection system of China, two modes of steel rail flaw detection vehicle detection and steel rail flaw detector detection are provided. The rail flaw detection vehicle has high flaw detection speed and strong adaptability, but the flaw detection operation result needs to be reviewed in a manual field to determine the existence and the size of the flaw. The steel rail flaw detector has high flaw detection sensitivity, but low speed and poor efficiency. With the rapid development of railway transportation, the rail flaw detection vehicle takes more and more flaw detection tasks, and particularly on high-speed rail lines and plateau lines, the rail flaw detection vehicle is mainly adopted for rail flaw detection due to the difficulty of manual flaw detection operation caused by long distance in sections, harsh environment and the like.
According to the regulations of the operation management method of the steel rail flaw detection vehicle, the steel rail flaw detection vehicle mainly utilizes a B-type diagram to judge the flaw, and the ultrasonic detection suspicious flaw alarm of the flaw detection vehicle is divided into three stages: the first-level alarm, namely the B-type diagram does not form obvious damage form trend; secondary alarm, namely a B-type graph forms a more obvious damage trend; the three-level alarm, type B, shows severe injury.
At present, the relationship between the rail damage detection capability of a flaw detection vehicle and indexes such as detection speed, damage size, damage angle and the like cannot be quantitatively analyzed. This presents problems and challenges to the planning, implementation and supervisory control of rail inspection work.
Disclosure of Invention
In order to solve the technical problem that the relationship between the rail damage detection capability of a flaw detection vehicle and indexes such as detection speed, damage size, damage angle and the like cannot be quantitatively analyzed in the prior art, the invention provides a method and a device for detecting rail damage of a railway.
The invention provides a method for detecting rail damage of a railway, which is used for calculating the detection capability of a rail flaw detection vehicle on various damages, and comprises the following steps: acquiring original data of rail damage of a railway; the original data of the railway steel rail damage comprises the following steps: the characteristic parameters of rail damage, the speed information of the flaw detection vehicle, wherein the characteristic parameters of rail damage include: the length of a screw hole crack, the detection angle, the aperture of a rail web through hole, the length of a rail head cross hole and the aperture of the rail head cross hole; establishing detection probability models of different damage types according to a logistic regression method and a machine learning method; according to the original data of the rail damage of the railway, fitting and determining model coefficients of detection probability models of different damage types by using an MLE maximum probability estimation method; the detection probability models of different damage types comprise: the method comprises the following steps of (1) detecting a probability model of screw hole cracks when the damage type is the screw hole cracks, detecting a probability model of rail surface flat-bottom holes when the damage type is rail surface flat-bottom holes, detecting a probability model of rail head cross holes when the damage type is rail head cross holes, and detecting a probability model of rail web through holes when the damage type is rail web through holes; and determining the detection probability of different damage types/the vehicle speed range of the steel rail flaw detection vehicle according to the determined detection probability models of different damage types and the current vehicle speed of the steel rail flaw detection vehicle/the set detection probability of the damage types.
Further, the detection probability model of the screw hole crack is as follows:
wherein q is the detection probability of screw hole cracks; v is the current speed of the rail flaw detection vehicle; a is a detection angle; and b is the crack length of the screw hole.
Further, the detection probability model of the rail surface flat bottom hole is as follows:
wherein q is the detection probability of the rail surface flat bottom hole; v is the current speed of the rail flaw detection vehicle; and b is the aperture of the rail surface flat bottom hole.
Further, the detection probability model of the railhead cross hole is as follows:
wherein q is the detection probability of the railhead cross hole; v is the current speed of the rail flaw detection vehicle; a is the aperture of the rail head cross hole; and b is the length of the rail head cross hole.
Further, the detection probability model of the rail web through hole is as follows:
wherein q is the detection probability of the rail web through hole; v is the current speed of the rail flaw detection vehicle; and b is the aperture of the rail web through hole.
In order to achieve the above object, the present invention correspondingly provides a rail damage detection device for calculating the detection capability of a rail flaw detection vehicle for various damages, comprising: the data acquisition module is used for acquiring original data of the railway steel rail damage; the original data of the railway steel rail damage comprises the following steps: the characteristic parameters of rail damage, the speed information of the flaw detection vehicle, wherein the characteristic parameters of rail damage include: the length of a screw hole crack, the detection angle, the aperture of a rail web through hole, the length of a rail head cross hole and the aperture of the rail head cross hole; the model establishing module is used for establishing detection probability models of different damage types according to a logistic regression method and a machine learning method; the model determining module is used for fitting and determining model coefficients of detection probability models of different damage types by utilizing an MLE maximum probability estimation method according to original data of the railway steel rail damage; the detection probability models of different damage types comprise: the method comprises the following steps of (1) detecting a probability model of screw hole cracks when the damage type is the screw hole cracks, detecting a probability model of rail surface flat-bottom holes when the damage type is rail surface flat-bottom holes, detecting a probability model of rail head cross holes when the damage type is rail head cross holes, and detecting a probability model of rail web through holes when the damage type is rail web through holes; and the detection probability/detection vehicle speed determining module is used for determining the detection probability of different damage types/the vehicle speed range of the rail flaw detection vehicle according to the determined detection probability models of different damage types and the vehicle speed of the current rail flaw detection vehicle/the set detection probability of the damage types.
The method and the device for detecting the rail damage of the railway have the advantages that the relationship between the detection capability of the rail damage and indexes such as the detection speed, the size of the damage, the angle of the damage and the like can be quantitatively determined, the detection capability of the flaw detection vehicle on various damages is calculated by combining a detection probability model under the given detection speed, and further, the detection speed required to be set when the flaw detection vehicle detects various damages is calculated under the condition that the detection probability requirement is given (such as not lower than 80%). The detection probability and the detection speed of the flaw detection vehicle can be set according to the expected flaw situation. And the method can also obtain visual display of the relation between related variables according to the establishment of a detection probability model so as to promote the decision of damage detection business and provide support for planning, implementation and supervision control of steel rail flaw detection work.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a flowchart illustrating steps of a method for detecting a rail defect according to an embodiment of the present invention.
Fig. 2 is a velocity-detection probability relationship diagram (detection angle 15 °) of a screw hole crack according to an embodiment of the present invention.
Fig. 3 is a velocity-detection probability relationship diagram (detection angle 25 °) of a screw hole crack according to an embodiment of the present invention.
Fig. 4 is a velocity-detection probability relationship diagram (detection angle 37 °) of a screw hole crack according to an embodiment of the present invention.
Fig. 5 is a velocity-detection probability relationship diagram (detection angle 45 °) of a screw hole crack according to an embodiment of the present invention.
FIG. 6 is a graph of velocity versus probability of detection for a rail face flat bottom hole in accordance with an embodiment of the present invention.
Fig. 7 is a graph of the velocity-detection probability relationship for a railhead cross bore (length of railhead cross bore 35) in accordance with an embodiment of the present invention.
Fig. 8 is a graph of the velocity-detection probability relationship for a railhead crosshole (railhead crosshole length of 25) in accordance with an embodiment of the present invention.
Fig. 9 is a graph of velocity-detection probability relationship of the rail web through hole according to an embodiment of the present invention.
FIG. 10 is a schematic diagram of a lesion detection rate for a lesion according to an embodiment of the present invention
Fig. 11 is a schematic structural view of a railway rail flaw detection apparatus according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by persons skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
The invention provides a method and a device for detecting rail damage of a railway, aiming at efficiently and accurately quantitatively analyzing the relationship between the rail damage detection capability of a flaw detection vehicle and indexes such as detection speed, damage size, damage angle and the like by using methods in the fields of data mining and machine learning and providing support for planning, implementing and supervising control of rail flaw detection work.
Fig. 1 is a flowchart illustrating steps of a method for detecting a rail damage according to an embodiment of the present invention, and as shown in fig. 1, the method for detecting a rail damage according to an embodiment of the present invention includes: s100, acquiring original data of rail damage of a railway; s200, establishing detection probability models of different damage types according to a logistic regression method and a machine learning method; s300, according to the original data of the rail damage of the railway, fitting and determining model coefficients of detection probability models of different damage types by utilizing an MLE maximum probability estimation method; and S400, determining the detection probability of different damage types/the vehicle speed range of the steel rail flaw detection vehicle according to the determined detection probability models of different damage types and the current vehicle speed of the steel rail flaw detection vehicle/the set detection probability of the damage types.
In step S100, raw data of rail damage is acquired. In the implementation process, the original data of the railway steel rail damage comprises the following steps: the characteristic parameter of rail damage, the speed of a motor vehicle information of detection car, wherein the characteristic parameter of rail damage includes: screw hole crack length, detection angle, aperture of rail web through hole, length of railhead cross hole and aperture of railhead cross hole. The original data of the railway rail damage can be obtained through the counted historical detection data, or a plurality of sections of test tracks (a plurality of different types of rail damages are set on the test tracks) are set, and the damage information of the plurality of sections of test tracks is collected by using a detection vehicle so as to obtain the original data of the railway rail damage.
After the original data of the rail damage is obtained, in step S200, detection probability models of different damage types are established according to a logistic regression method and a machine learning method. In a specific implementation process, a technician can construct a plurality of logistic regression models as detection probability models of different damage types according to the original data of the railway steel rail damage, wherein the logistic regression models use Sigmoid functions, and the Sigmoid functions are P (y is 1| x; theta) exp (theta)Tx)/(1+exp(θTx)), where θ)TAnd x is a model parameter. The data of various injuries can be realized by using R language or Java programming. Then, establishing the logistic regression according to the speed of the flaw detection vehicle and the characteristic parameters of the rail flawModel parameters in a model, wherein the model parameters are the sum of the speed of the flaw detection vehicle, the product of the characteristic parameters of a plurality of rail flaws and the corresponding regression coefficients, and the characteristic parameters of the rail flaws include: screw hole crack length, detection angle, aperture of rail web through hole, length of railhead cross hole and aperture of railhead cross hole. I.e. thetaTx＝w0+w1v+w2a+w3b + …, wherein v, a and b … are characteristic parameters of rail damage, w0、w1、w2… are the corresponding regression coefficients.
In step S300, fitting and determining model coefficients of the detection probability models of different damage types by using an MLE maximum probability estimation method according to the original data of the rail damage of the railway. In a specific implementation process, technicians can establish target variables of the logistic regression model according to detection results of different damage types in original data of the railway steel rail damage, wherein the target variables are detection probabilities, the undetected mark is 0, and the detected mark is 1; according to the original data of the rail damage of the railway, fitting and determining regression coefficients of the multiple logistic regression models by using an MLE maximum probability estimation method, namely determining detection probability models of different damage types; establishing a mathematical relationship between the predicted variable and the target variable according to a Logistic Regression (Logistic Regression) method, establishing detection probability models of different damage types through machine learning, determining parameters in the models by using an MLE maximum probability estimation method, and determining mathematical relationship expressions of the detection probability models of different damage types, namely determining the detection probability models of different damage types.
In the specific implementation process, because different lines or mileage sections have the condition of inconsistent external environments, model training can be performed according to the damage type, the line, the category, the mileage section and the time interval as limiting conditions during data training, and after the model training is completed, the limiting conditions and the calculation time are used as indexes, and the models under different conditions are stored. Based on the damage data of the test line of the general Chinese railway company, the following four relatively common damages are detected, and the determined detection probability model is specifically as follows:
when the damage type is the screw crack, the detection probability model of the screw crack is as follows:
wherein q is the detection probability of screw hole cracks; v is the current speed of the rail flaw detection vehicle; a is a detection angle; and b is the crack length of the screw hole. The relationship between the relevant variables can be obtained according to the detection probability model of the screw hole crack, and fig. 2 to 5 are graphs showing the relationship between the speed and the detection probability of the screw hole crack according to the embodiment of the invention, which are graphs showing the detection probability and the vehicle speed of the screw hole crack with the lengths of 3mm, 5mm and 8mm when the detection angles are 15 °, 25 °, 37 ° and 45 °, respectively. The skilled person can assist the detection work according to fig. 2 to 5. In a specific implementation process, in one case, the detection probabilities of different damage types are determined according to the determined detection probability models of different damage types and the current speed of the rail flaw detection vehicle, a detector can take the current speed of the rail flaw detection vehicle as an independent variable and bring the independent variable into the detection probability model of the screw hole cracks, the detection probabilities of the screw hole cracks with different lengths are calculated, and a measurement report is compiled; in another case, the speed range of the rail flaw detection vehicle is determined according to the determined detection probability models of different flaw types and the set detection probability of the flaw types, and a detector can take the set detection probability of a certain screw hole crack as an independent variable according to the detection probability requirement of the current section and bring the independent variable into the detection probability model of the screw hole crack to calculate the allowable range of the speed of the rail flaw detection vehicle, so that the rail can be conveniently detected at a proper speed within the allowable probability range.
When the damage type is a rail surface flat-bottom hole, the detection probability model of the rail surface flat-bottom hole is as follows:
wherein q is the detection outline of the rail surface flat bottom holeRate; v is the current speed of the rail flaw detection vehicle; and b is the aperture of the rail surface flat bottom hole. FIG. 6 is a velocity-detection probability relationship diagram of a rail plane flat-bottom hole according to an embodiment of the present invention, in which the diameters of the rail plane flat-bottom hole are respectively
In the process, the detection probability of the rail surface flat bottom hole and the vehicle speed are plotted, and the skilled person can assist the detection work according to the graph in fig. 6. In a specific implementation process, in one case, the detection probabilities of different damage types are determined according to the determined detection probability models of different damage types and the current speed of the rail flaw detection vehicle, a detector can take the speed of the current rail flaw detection vehicle as an independent variable and bring the speed into the detection probability model of the rail plane flat-bottom hole, the detection probabilities of the rail plane flat-bottom holes with different apertures are calculated, and a measurement report is compiled; the other situation is that the speed range of the rail flaw detection vehicle is determined according to the determined detection probability models of different flaw types and the set detection probability of the flaw types, a detection person can take the set detection probability of a certain rail plane flat-bottom hole as an independent variable according to the requirement of the detection probability of the current section, the detection person brings the detection probability into the detection probability model of the rail plane flat-bottom hole, the allowable range of the speed of the rail flaw detection vehicle is calculated, and the rail is conveniently detected at a proper speed in the allowable probability range.
When the damage type is the railhead cross bore, the detection probability model of the railhead cross bore is as follows:
wherein q is the detection probability of the railhead cross hole; v is the current speed of the rail flaw detection vehicle; a is the aperture of the rail head cross hole; and b is the length of the rail head cross hole. The relation between related variables can be obtained according to a detection probability model of the rail head cross hole, and the method is shown in figure 7,FIG. 8 is a graph of velocity-probability of cross hole of railhead, wherein the diameter of the cross hole is 35 and 25 when the cross hole is long, respectively
Is plotted against the vehicle speed. The detection work can be assisted by those skilled in the art according to fig. 7 and 8. In a specific implementation process, in one case, the detection probabilities of different damage types are determined according to the determined detection probability models of different damage types and the current speed of the rail flaw detection vehicle, and a detector can take the speed of the current rail flaw detection vehicle as an independent variable and bring the speed into the detection probability model of the rail head cross hole, calculate the detection probability of the rail head cross hole with different aperture diameters and compile a measurement report; in another case, the speed range of the rail flaw detection vehicle is determined according to the determined detection probability models of different flaw types and the set detection probability of the flaw types, and a detection person can take the set detection probability of a certain rail head cross hole as an independent variable according to the detection probability requirement of the current section and bring the independent variable into the detection probability model of the rail head cross hole to calculate the allowable range of the speed of the rail flaw detection vehicle, so that the rail can be conveniently detected at a proper speed within the allowable probability range.
When the damage type is the rail waist through hole, the detection probability model of the rail waist through hole is as follows:
wherein q is the detection probability of the rail web through hole; v is the current speed of the rail flaw detection vehicle; and b is the aperture of the rail web through hole. FIG. 9 is a graph of the velocity-detection probability relationship of the rail web through holes of the rail web holes of the present invention with the diameters of the rail web through holes being respectively
And (4) a relation graph of the detection probability of the rail web through hole and the vehicle speed. The detection work can be assisted by a person skilled in the art according to fig. 9. In one embodiment, the method is based onThe detection probability models of different damage types and the current speed of the rail flaw detection vehicle are determined, the detection probabilities of different damage types are determined, a detector can take the current speed of the rail flaw detection vehicle as an independent variable and bring the independent variable into the detection probability model of the rail waist through hole, the detection probabilities of the rail waist through holes with different apertures are calculated, and a measurement report is compiled; in another case, the speed range of the rail flaw detection vehicle is determined according to the determined detection probability models of different flaw types and the set detection probability of the flaw types, and a detection person can take the set detection probability of a certain rail waist through hole as an independent variable according to the requirement of the detection probability of the current section and bring the detection probability into the detection probability model of the rail waist through hole to calculate the allowable range of the speed of the rail flaw detection vehicle, so that the rail can be conveniently detected at a proper speed within the allowable probability range.
FIG. 10 is a schematic diagram of a lesion detection rate for viewing a certain lesion according to an embodiment of the present invention. As shown in fig. 10, when looking up a certain injury, a person skilled in the art inputs a specific feature and selects a model according to a requirement, and then the system enumerates and displays as many cases as possible; the user inputs a plurality of values in a certain item (such as an angle), and the corresponding damage detection rates under different conditions (such as angles of 30, 45 and 60) are enumerated. The embodiment is helpful for a user to simultaneously grasp various conditions and analyze and compare the influence of different conditions on the detection rate.
After the method for detecting a rail flaw according to the embodiment of the present invention is described, a rail flaw detection apparatus according to the embodiment of the present invention will be described. The implementation of the device can be referred to the implementation of the method, and repeated details are not repeated. The terms "module", "unit", and the like, as used hereinafter, may be software and/or hardware that implements a predetermined function.
Fig. 11 is a schematic structural view of a railway rail flaw detection device according to an embodiment of the present invention. As shown in fig. 11, the apparatus for detecting a rail damage according to an embodiment of the present invention is used for calculating the capability of a flaw detection vehicle to detect various types of damage, and includes: the data acquisition module 100 is used for acquiring original data of rail damage of a railway; the model establishing module 200 is used for establishing detection probability models of different damage types according to a logistic regression method and a machine learning method; the model determining module 300 is used for fitting and determining model coefficients of detection probability models of different damage types by using an MLE maximum probability estimation method according to original data of the railway steel rail damage; and the detection probability/detection vehicle speed determining module 400 is used for determining the detection probability of different damage types/the vehicle speed range of the rail flaw detection vehicle according to the determined detection probability models of different damage types and the vehicle speed of the current rail flaw detection vehicle/the set detection probability of the damage type.
In the detection probability/detection vehicle speed determination module 400, when the damage type is a screw hole crack, the detection probability model of the screw hole crack is as follows:
wherein q is the detection probability of screw hole cracks; v is the current speed of the rail flaw detection vehicle; a is a detection angle; and b is the crack length of the screw hole.
In the detection probability/detection vehicle speed determination module 400, when the damage type is a rail surface flat bottom hole, the detection probability model of the rail surface flat bottom hole is as follows:
wherein q is the detection probability of the rail surface flat bottom hole; v is the current speed of the rail flaw detection vehicle; and b is the aperture of the rail surface flat bottom hole.
In the detection probability/detection vehicle speed determination module 400, when the damage type is a railhead cross hole, the detection probability model of the railhead cross hole is as follows:
wherein q is the detection probability of the railhead cross hole; v is the current speed of the rail flaw detection vehicle; a is the aperture of the rail head cross hole; and b is the length of the rail head cross hole.
In the detection probability/detection vehicle speed determination module 400, when the damage type is a rail web through hole, the detection probability model of the rail web through hole is as follows:
wherein q is the detection probability of the rail web through hole; v is the current speed of the rail flaw detection vehicle; and b is the aperture of the rail web through hole.
The method and the device for detecting the rail damage of the railway have the advantages that the relationship between the detection capability of the rail damage and indexes such as the detection speed, the size of the damage, the angle of the damage and the like can be quantitatively determined, the detection capability of the flaw detection vehicle on various damages is calculated by combining a detection probability model under the given detection speed, and further, the detection speed required to be set when the flaw detection vehicle detects various damages is calculated under the condition that the detection probability requirement is given (such as not lower than 80%). The detection probability and the detection speed of the flaw detection vehicle can be set according to the expected flaw condition. And the method can also obtain visual display of the relation between related variables according to the establishment of a detection probability model so as to promote the decision of damage detection business and provide support for planning, implementation and supervision control of steel rail flaw detection work.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A rail damage detection method for a railway is used for calculating the detection capability of a rail flaw detection vehicle on various damages, and is characterized by comprising the following steps:
acquiring original data of rail damage of a railway; the original data of the railway steel rail damage comprises the following steps: the characteristic parameters of rail damage, the speed information of the flaw detection vehicle, wherein the characteristic parameters of rail damage include: the length of a screw hole crack, the detection angle, the aperture of a rail web through hole, the length of a rail head cross hole and the aperture of the rail head cross hole;
establishing detection probability models of different damage types according to a logistic regression method and a machine learning method;
according to the original data of the rail damage of the railway, fitting and determining model coefficients of detection probability models of different damage types by using an MLE maximum probability estimation method; the detection probability models of different damage types comprise: the method comprises the following steps of (1) detecting a probability model of screw hole cracks when the damage type is the screw hole cracks, detecting a probability model of rail surface flat-bottom holes when the damage type is rail surface flat-bottom holes, detecting a probability model of rail head cross holes when the damage type is rail head cross holes, and detecting a probability model of rail web through holes when the damage type is rail web through holes;
and determining the detection probability of different damage types/the vehicle speed range of the steel rail flaw detection vehicle according to the determined detection probability models of different damage types and the current vehicle speed of the steel rail flaw detection vehicle/the set detection probability of the damage types.
2. The method for detecting a rail flaw in a railway according to claim 1, wherein in the step of determining the detection probability of the different flaw types/the vehicle speed range of the rail flaw detection vehicle based on the determined detection probability models of the different flaw types and the current vehicle speed of the rail flaw detection vehicle/the set detection probability of the flaw type, the detection probability model of the screw hole crack is:
wherein q is the detection probability of screw hole cracks;
v is the current speed of the rail flaw detection vehicle;
a is a detection angle;
and b is the crack length of the screw hole.
3. The method for detecting a rail defect according to claim 1, wherein in the step of determining the detection probability of different defect types/the vehicle speed range of the rail flaw detector according to the determined detection probability models of different defect types and the current vehicle speed of the rail flaw detector/the set detection probability of the defect type, the detection probability model of the rail plane flat-bottom hole is as follows:
wherein q is the detection probability of the rail surface flat bottom hole;
v is the current speed of the rail flaw detection vehicle;
and b is the aperture of the rail surface flat bottom hole.
4. The method for detecting a rail defect according to claim 1, wherein in the step of determining the detection probability of different defect types/the vehicle speed range of the rail flaw detector vehicle based on the determined detection probability models of different defect types and the current vehicle speed of the rail flaw detector vehicle/the set detection probability of the defect type, the detection probability model of the rail head cross bore is as follows:
wherein q is the detection probability of the railhead cross hole;
v is the current speed of the rail flaw detection vehicle;
a is the aperture of the rail head cross hole;
and b is the length of the rail head cross hole.
5. The method for detecting rail damage of railway according to claim 1, wherein in the step of determining the detection probability of different types of damage/the vehicle speed range of the rail flaw detection vehicle according to the determined detection probability models of different types of damage and the current vehicle speed of the rail flaw detection vehicle/the set detection probability of the type of damage, the detection probability model of the rail web through hole is as follows:
wherein q is the detection probability of the rail web through hole;
v is the current speed of the rail flaw detection vehicle;
and b is the aperture of the rail web through hole.
6. A rail damage detection device for calculating the detection capability of a rail detection vehicle to various damages is characterized by comprising:
the data acquisition module is used for acquiring original data of the railway steel rail damage; the original data of the railway steel rail damage comprises the following steps: the characteristic parameters of rail damage, the speed information of the flaw detection vehicle, wherein the characteristic parameters of rail damage include: the length of a screw hole crack, the detection angle, the aperture of a rail web through hole, the length of a rail head cross hole and the aperture of the rail head cross hole;
the model establishing module is used for establishing detection probability models of different damage types according to a logistic regression method and a machine learning method;
the model determining module is used for fitting and determining model coefficients of the detection probability models of different damage types by utilizing an MLE maximum probability estimation method according to the original data of the railway steel rail damage; the detection probability models of different damage types comprise: the method comprises the following steps of (1) detecting a probability model of screw hole cracks when the damage type is the screw hole cracks, detecting a probability model of rail surface flat-bottom holes when the damage type is rail surface flat-bottom holes, detecting a probability model of rail head cross holes when the damage type is rail head cross holes, and detecting a probability model of rail web through holes when the damage type is rail web through holes;
and the detection probability/detection vehicle speed determining module is used for determining the detection probability of different damage types/the vehicle speed range of the rail flaw detection vehicle according to the determined detection probability models of different damage types and the vehicle speed of the current rail flaw detection vehicle/the set detection probability of the damage types.
7. The apparatus according to claim 6, wherein in the detection probability/detection vehicle speed determination module, the detection probability model of the screw hole crack is:
wherein q is the detection probability of the screw hole crack;
v is the current speed of the rail flaw detection vehicle;
a is a detection angle;
and b is the crack length of the screw hole.
8. The apparatus according to claim 6, wherein in the detection probability/detection vehicle speed determination module, the detection probability model of the rail plane flat bottom hole is:
wherein q is the detection probability of the rail surface flat-bottom hole;
v is the current speed of the rail flaw detection vehicle;
and b is the aperture of the rail surface flat bottom hole.
9. The apparatus according to claim 6, wherein in the detection probability/detection vehicle speed determination module, the detection probability model of the railhead cross bore is:
wherein q is the detection probability of the railhead cross hole;
v is the current speed of the rail flaw detection vehicle;
a is the aperture of the rail head cross hole;
and b is the length of the rail head cross hole.
10. The apparatus according to claim 6, wherein in the detection probability/detection vehicle speed determination module, the detection probability model of the rail web through hole is:
wherein q is the detection probability of the rail web through hole;
v is the current speed of the rail flaw detection vehicle;
and b is the aperture of the rail web through hole.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810186643.8A CN108334908B (en) | 2018-03-07 | 2018-03-07 | Method and device for detecting railway rail damage |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810186643.8A CN108334908B (en) | 2018-03-07 | 2018-03-07 | Method and device for detecting railway rail damage |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108334908A CN108334908A (en) | 2018-07-27 |
| CN108334908B true CN108334908B (en) | 2022-06-24 |
Family
ID=62928968
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810186643.8A Active CN108334908B (en) | 2018-03-07 | 2018-03-07 | Method and device for detecting railway rail damage |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108334908B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109612806A (en) * | 2019-02-25 | 2019-04-12 | 北京航空航天大学 | A kind of efficient test material preparation and test method suitable for the test of surface crack defect detection probability |
| CN110503209B (en) * | 2019-07-24 | 2022-02-01 | 山东麦港数据系统有限公司 | Steel rail analysis early warning model construction method and system based on big data |
| CN112147221B (en) * | 2020-09-22 | 2023-02-03 | 济南大学 | Steel rail screw hole crack identification method and system based on ultrasonic flaw detector data |
| CN113447572B (en) * | 2021-09-01 | 2021-11-30 | 深圳市信润富联数字科技有限公司 | Steel rail flaw detection method, electronic device, steel rail flaw detection vehicle and readable storage medium |
| CN116039713B (en) * | 2022-12-30 | 2023-08-01 | 合肥德泰科通测控技术有限公司 | Intelligent damage identification management system for steel rail |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1236634A1 (en) * | 2001-02-20 | 2002-09-04 | Digital Image Research Limited | Method and apparatus for determining track condition |
| JP2008502538A (en) * | 2004-06-11 | 2008-01-31 | ストラテック システムズ リミテッド | Railway track scanning system and method |
| CN103954696A (en) * | 2014-05-22 | 2014-07-30 | 株洲时代电子技术有限公司 | Flaw detection simulated test method for steel rail |
| CN106864477A (en) * | 2017-03-15 | 2017-06-20 | 华南理工大学 | A kind of rail flaw ultrasonic detection intelligent detecting method based on time-frequency do-nothing function |
| CN107255675A (en) * | 2017-06-21 | 2017-10-17 | 株洲时代电子技术有限公司 | A kind of rail examination gate parameter setting and synchronous display system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202421133U (en) * | 2011-12-07 | 2012-09-05 | 暨南大学 | Railway track damage detection device based on magnetostriction and longitudinal ultrasonic guided waves |
| CN104751169B (en) * | 2015-01-10 | 2018-02-02 | 哈尔滨工业大学（威海） | High ferro rail defects and failures sorting technique |
| CN104850748B (en) * | 2015-05-26 | 2017-09-15 | 北京交通大学 | A kind of railway track fractures accident analysis method for early warning and system |
| CN107196844A (en) * | 2016-11-28 | 2017-09-22 | 北京神州泰岳信息安全技术有限公司 | Exception mail recognition methods and device |
-
2018
- 2018-03-07 CN CN201810186643.8A patent/CN108334908B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1236634A1 (en) * | 2001-02-20 | 2002-09-04 | Digital Image Research Limited | Method and apparatus for determining track condition |
| JP2008502538A (en) * | 2004-06-11 | 2008-01-31 | ストラテック システムズ リミテッド | Railway track scanning system and method |
| CN103954696A (en) * | 2014-05-22 | 2014-07-30 | 株洲时代电子技术有限公司 | Flaw detection simulated test method for steel rail |
| CN106864477A (en) * | 2017-03-15 | 2017-06-20 | 华南理工大学 | A kind of rail flaw ultrasonic detection intelligent detecting method based on time-frequency do-nothing function |
| CN107255675A (en) * | 2017-06-21 | 2017-10-17 | 株洲时代电子技术有限公司 | A kind of rail examination gate parameter setting and synchronous display system |
Non-Patent Citations (5)
| Title |
|---|
| Assessment of Rail Flaw Inspection Data;Brandon D. Jeffrey;《AIP Conference Proceedings》;20001116;第1-29页 * |
| 基于三维漏磁检测技术的高速铁路钢轨斜裂纹特征识别研究;赵飒;<中国优秀博硕士学位论文全文数据库(硕士) 基础科学辑>;20120315(第03期);第A004-27页 * |
| 基于可靠性的裂纹检出概率曲线测定方法;冯振宇等;《无损检测》;20101231;第32卷(第4期);第249-252页 * |
| 船体结构检测概率模型对裂纹分布更新的影响;李典庆等;《上海交通大学学报》;20041130;第38卷(第11期);第1865-1869页 * |
| 钢轨疲劳斜裂纹垂向磁化检测仿真;高运来等;《无损检测》;20141231;第36卷(第11期);第18-23页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108334908A (en) | 2018-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108334908B (en) | Method and device for detecting railway rail damage | |
| WO2020052147A1 (en) | Monitoring device fault detection method and apparatus | |
| CN109737309B (en) | Risk source and leakage tracing method and system based on risk identification | |
| CN111024728B (en) | Railway detection method and system based on computer vision and ultrasonic flaw detection | |
| US20140200831A1 (en) | Pipe Damage Interpretation System | |
| CN115643159B (en) | Equipment abnormity early warning method and system based on edge calculation | |
| CN113732558B (en) | Online detection method and device for typical quality problems of robot welding | |
| KR102306269B1 (en) | Method and apparatus for real-time arc-welding defect detection and classification | |
| CN113192037B (en) | Belt conveyor monitoring method, system, medium and electronic terminal | |
| CN106274979A (en) | Rail wear automatic detection device | |
| CN105866247A (en) | Device and method for detecting sticking compactness of steel plate | |
| CN116611574A (en) | Remanufacturing evaluation method and system for steel structure of pipeline welding robot | |
| CN112819262A (en) | Memory, process pipeline inspection and maintenance decision method, device and equipment | |
| CN103493081B (en) | The coating thickness of vehicle calculates system, computational methods | |
| CN115564141A (en) | Welding spot quality monitoring method and device and medium thereof | |
| KR20230072164A (en) | Product quality management system for zero defect rate | |
| CN115205228A (en) | Monitoring and analyzing system for production defect distribution diagram of liquid crystal display panel | |
| Papacharalampopoulos et al. | Experimental Investigation of rubber extrusion process through vibrational testing | |
| Iwata et al. | AI-aided Hammering Test System to Automatically Generate Anomaly Maps. | |
| US11868112B2 (en) | Multi-point measurement system and method thereof | |
| KR102309113B1 (en) | System for checking equipment and material quality | |
| CN117056686B (en) | Alarming method and system for detecting surface defects of pressure container | |
| CN114001939B (en) | Line connection device monitoring method, device, equipment and storage medium | |
| KR20200064185A (en) | Apparatus for detecting electric supply equipment | |
| CN115144259B (en) | Method and system for detecting deformation resistance of steel |
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 | ||
| CB02 | Change of applicant information |
Address after: No. 2, Da Liu Shu Road, Haidian District, Beijing Applicant after: China Railway Scientific Research Institute Group Co., Ltd. Applicant after: Institute of infrastructure inspection, China Railway Research Institute Group Co., Ltd. Applicant after: Beijing IMAP Technology Co., Ltd. Address before: No. 2, Da Liu Shu Road, Haidian District, Beijing Applicant before: China Academy of Railway Sciences Applicant before: Infrastructure Testing Institute of China Academy of Railway Sciences Applicant before: Beijing IMAP Technology Co., Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |