KR101247377B1 - Automatic Operation System for Wheel Tread Inspection of Rail Installation Type - Google Patents

Abstract

The present invention relates to an unmanned automated system for rail-mounted wheel face flaw detection, comprising: a rigid sensor structure consisting of a rail-shaped sensor mechanism; A vehicle recognition device for recognizing that the railway vehicle enters a measurement standby state on the rigid sensor structure; A wheel recognition device installed in the rigid sensor structure to detect and output only the corresponding information when the wheel step surface reaches the sensor position; A control device for switching, amplifying, filtering, converting an electrical signal output from a sensor of the sensor mechanism into a digital signal, and interfacing the converted digital signal to a computer; It includes software for processing with a built-in algorithm to extract and recognize only the defect information from the digital signal input to the computer via the interface unit.
According to the present invention, it is possible to diagnose and monitor cracks on the wheel face at a more improved speed after acquiring a magnetic image of the wheel face that enters a constant speed without stopping the high speed rail car, thereby ensuring safety such as derailment by the wheel of the high speed rail car. Effective in preventing accidents

Description

Automatic Operation System for Wheel Tread Inspection of Rail Installation Type}

The present invention relates to an unmanned automated system for rail-mounted wheel face flaw detection, and more particularly to a wheel face face provided with an array of sensors on the rail for high-speed diagnosis and monitoring of defects occurring on the wheel face of a high-speed railway vehicle. It relates to an unmanned automated system for flaw detection.

Wheels are the key structural components that support the weight of railroad cars and move in repetitive rolling contact with the rails. In particular, the damage to the wheels caused by high vertical stresses at the micro contact with the rails, thermal stress due to frictional heat generated by the wheels in contact with the braking surface, and non-metallic inclusions inevitably generated during the manufacturing process can be caused by high speed trains. Has a profound impact on driving safety. In addition, since the number of wheels is about 1000 per one high-speed railway vehicle, it takes a lot of time to measure the health of the wheels one by one. Therefore, when the railway vehicle enters the maintenance shop at the normal speed, a system for early identification of wheel tread cracks is required without pausing.

The present invention is to improve the defect detection capacity of the wheel surface cracks, when the high-speed railway vehicle enters the maintenance shop, diagnose the cracks on the wheel surface at high speed by the magnetic sensor array installed on the rail while passing through without stopping , To monitor.

In order to solve such a technical problem,

Rigid sensor structure consisting of a rail-shaped sensor mechanism; A vehicle recognition device for recognizing that the railway vehicle enters a measurement standby state on the rigid sensor structure; A wheel recognition device installed in the rigid sensor structure to detect and output only the corresponding information when the wheel step surface reaches the sensor position; A control device for switching, amplifying, filtering, converting an electrical signal output from a sensor of the sensor mechanism into a digital signal, and interfacing the converted digital signal to a computer; Provides a rail-mounted wheel tread test unmanned automation system, characterized in that it comprises a; software for processing with a built-in algorithm to extract and recognize only the defect information from the digital signal input to the computer via the interface unit.

Here, the sensor mechanism is equipped with a plurality of sensors for detecting the wheel tread surface, and configured to support half of the flange portion and the wheel tread surface, I-type that can inspect the inside of the vehicle tread surface from the half of the rail And an O-type capable of inspecting the outer side of the vehicle's undercarriage.

In addition, the vehicle recognition device irradiates the non-contact light to the mechanical proximity sensor or the entering railway vehicle to generate a trigger signal when the side of the entering railway vehicle is in contact with the vehicle to block the light when the railway vehicle passes through the trigger signal It is characterized in that it comprises a light sensor to generate.

In addition, the wheel recognition device is characterized in that it comprises a wheel recognition circuit for measuring the magnetic field distribution only in the region in contact with the wheel in the sensor arranged in two dimensions on the surface.

The sensor may be configured to include a magnetic field application and a high spatial resolution sensor that apply a magnetic field evenly to the wheel tread face in a limited space and extract the distribution of the magnetic field that varies depending on the presence, position, and size of the defect. It features.

In addition, the magnetic field application and high spatial resolution sensor is a parallel signal processing for which a plurality of magnetic sensors are arranged in a plane on a permanent magnet having a vertical magnetization direction, switching the power input unit of the magnetic sensor, and connected to the output unit as one Characterized in that it is made.

In addition, a spring is interposed between the lower end of the sensor mechanism for fixing the sensor and the surface of the rail to prevent the return of the sensor mechanism and damage of the sensor.

In addition, the control device is characterized in that it comprises a differential operational amplifier for amplifying the signal.

The circuits included in the wheel recognition device, the magnetic field application and the high spatial resolution sensor, and the control device for detecting the wheel tread of the railroad vehicle are characterized in that they are embedded in the rigid sensor structure as a module.

According to the present invention, after acquiring a magnetic image of the wheel face surface entering at a constant speed without stopping the high speed railway vehicle, it is possible to diagnose and monitor cracks in the wheel face surface at a more improved speed, such as derailment due to the wheel of the high speed railway vehicle. There is an advantage that can effectively prevent safety accidents.

In addition, by arranging the sensors in half of the rails in the longitudinal direction by crossing the sensors, the wheel surface information is acquired only in a structure that can withstand the load and speed of the vehicle sufficiently and the area where the wheels contact the sensor, thereby reducing the defects of the wheel surface. There is an effect that allows more efficient and precise inspection, such as efficient data processing by capacity.

1 is a block diagram of an unmanned automated system for rail-mounted wheel tread flaw detection according to the present invention.
2 is an installation state diagram of a sensor mechanism according to the present invention.
Figure 3 is a sensor mechanism according to the invention Figure 3a is I-type, Figure 3b is O-type.
Figure 4 shows a vehicle recognition device according to the invention, Figure 4a is an electric circuit diagram, Figure 4b is a photograph showing the installation state of the mechanical proximity sensor.
Figure 5 shows a vehicle recognition apparatus according to the present invention, Figures 5a and 5b is a photograph showing the installation state of the optical sensor.
6 is a view illustrating a concept for wheel recognition according to the present invention.
7 shows a wheel recognition apparatus according to the present invention. FIG. 7A is a wheel recognition circuit diagram and FIG. 7B is a graph showing data output from the wheel recognition circuit diagram.
8 is a photograph showing an arrangement of sensors for wheel recognition according to the present invention.
9 is a view showing the arrangement of the magnetic field application and high spatial resolution sensor according to the present invention, Figure 9a shows the arrangement of the magnetic sensor mounted on the PCB on the permanent magnet top plate, Figure 9b is a photograph showing the installation state. .
10 is a photograph showing an installation state of the sensor fixing mechanism according to the present invention.
11 is a view showing a structure of a sensor fixing mechanism according to the present invention.
12 is a circuit diagram showing the arrangement of the sensor according to the present invention.
Fig. 13 shows a preferred application of the modular rail-mounted wheel automatic flaw detector.

Hereinafter, with reference to the accompanying drawings, a rail-mounted wheel stepping surface unmanned automated system according to the present invention will be described in detail.

In the present invention, components such as a sensor, a magnetic field applying device, a switching circuit, an amplifier circuit, an A / D conversion circuit, and an interface are embedded in a rail.

First, according to FIG. 1, the unmanned automated system for rail-mounted wheel tread test according to the present invention is a rail-shaped sensor mechanism to prevent derailment of the wheel 20 while sufficiently enduring the speed and load of the railroad vehicle 10. Rigid sensor structure (1) consisting of, a vehicle recognition device (2) for recognizing that the railway vehicle 10 for unmanned detection enters the measurement standby state on the rigid sensor structure (1), and the rigid sensor structure ( 1) a wheel recognition device 3 which detects and outputs only the corresponding information when the wheel answering surface reaches the sensor position so as to minimize the amount of data that detects the entire wheel answering surface of the one wheel; A control device 5 for switching, amplifying, filtering, converting an electrical signal output from the sensor of 1) into a digital signal, and interfacing the converted digital signal to a computer 4, and a digital input to the computer 5; It consists of software 6 which processes by built-in algorithms so that only defect information can be extracted and recognized from the real signal.

At this time, the general rigid sensor structure (1) is an electronic circuit including a sensor in the form of a wafer, it does not bear a load of about 17 tons or more and is broken. Therefore, the rigid sensor structure 1 of the present invention is made of a sensor mechanism having a rail shape as shown in Figure 3 so as to withstand the load of the vehicle sufficiently to prevent the derailment of the wheel. The sensor mechanism is designed to always support half of the flange portion and the wheel step surface.

That is, the sensor mechanism in FIG. 3 is configured to inspect the wheel tread of the inside (I-type, see FIG. 3A) and the outside (O-type, see FIG. 3B) of the vehicle at half of the rail.

And, the vehicle recognition device 2 is to detect the vehicle enters the maintenance window for the recognition of the railway vehicle (10). In general, although the railroad vehicle 10 manually checks by the worker until entering the sensor area or the system area for measurement, in the present invention, all the inspection work is automatically performed after the operator turns on the system. To make it possible. Therefore, the vehicle recognition device 2 recognizes that the vehicle enters the measurement standby state for the unmanned detection.

At this time, the vehicle recognition device 2 is divided into contact and contactless. The contact type is a mechanical proximity sensor as shown in FIG. 4 so that a trigger signal is generated when the side of the vehicle contacts the spring-shaped cantilever. On the other hand, using these signals it was possible to use the prototype to use the bogie to operate the brake system of the bogie. This means that the brake system is automatically activated at the time of entering, because a large amount of risk must be taken when a hand-held truck stops traveling at a certain speed. However, there is a risk of damage to the vehicle, shortened life of mechanical proximity sensor and malfunction when applied in the field.

Therefore, in the present invention, an infrared light sensor is applied to non-contact optically recognize the vehicle as shown in FIG. In other words, by irradiating infrared light in the eye-safe area, the trigger signal is generated from the principle of blocking the light when the vehicle passes.

A total of 553,896 sensors are embedded in the rigid sensor structure 1 of the I-type and O-type in order to detect the whole of one wheel tread. In addition, when the vehicle's entry speed is calculated from 6 to 30 km / h, a data acquisition speed of approximately 76,929 to 384,405 sampling / sec (sampling / sec) is required, and as a result, the amount of data for inspecting one wheel 20 is When converted into 12 bits, it is approximately 3.4639 Gbytes.

Accordingly, the wheel recognition device 3 can acquire the corresponding signal only in the region where the wheel 20 is in contact with each other so as to minimize the amount of data as shown in FIG. 6. That is, in order to measure the magnetic field distribution only in an area in contact with the wheel in the sensor arranged in two dimensions on the plane in consideration of the point of contact between the rail and the line (or narrow surface contact) with the rail, As the wheel recognition circuit is applied, the magnetic field is applied from the rail to the wheel, and the change of the magnetic field and the output of the magnetic sensor change when the wheel is close, so the change amount is measured and the data is output only in the corresponding local area.

FIG. 8 shows an actual photograph to which the above principle and circuit are applied, and a portion indicated by a thick line (red color) is a wheel recognition sensor 30, and sensors arranged in a direction perpendicular to the same are used to display information on the answer surface in the corresponding area. It is a sensor to acquire. When the wheel enters the sensor area, the wheel recognition sensor 30 recognizes this and activates the sensors at the direction perpendicular to the wheel, that is, the wheel tread.

The magnetic field application and high spatial resolution sensor mounted on the sensor mechanism uses permanent magnets N35 and Nd-Fe-B as magnetic sources for the operation of the sensor and the resources with less power. As shown in FIG. 9, the magnetic sensor array 46 mounted on the PCB 44 on the upper plate of the permanent magnet 42 having the magnetization direction in the vertical direction was arranged in two dimensions.

And, the Hall sensor is applied as a magnetic sensor. The hall sensor uses a principle that inputs a current to a semiconductor and a magnetic field is applied in a direction perpendicular to the surface thereof, which causes distortion of an electron and a voltage difference caused by the electron. That is, when a high intensity magnetic field in the vertical direction is applied, the output of the hall sensor reaches saturation. Therefore, by arranging two Hall sensors and measuring the difference, saturation is prevented and the local magnetic field distribution due to the presence of a defect is measured.

In order to maintain a constant distance between the sensor and the resource, and to minimize the damage caused by friction with the rigid sensor structure (1), a separate sensor fixing mechanism (50) was manufactured and applied as shown in FIG. The sensor fixing mechanism 50 maintains a constant gap between the sensor and the resource and allows the electronic circuit to be embedded therein.

In addition, when a high load wheel approaches, attraction may be generated by the force of the magnet, but in FIG. 11, the spring 52 is fixed to the lower portion and designed and manufactured to prevent reversion and damage.

On the other hand, the Hall sensor is composed of two inputs and two outputs, and when arranged in two dimensions, the number of wiring increases exponentially in proportion to the number of sensors. This has a problem that not only signal processing but also very difficult results in wiring in a limited space.

Therefore, in the present invention, as shown in FIG. 12, the problem is solved by parallel signal processing of switching the power input unit and connecting the output unit as one, so that the sensor having a high spatial resolution can be arranged in two dimensions.

And the control apparatus 5 of this invention consists of an amplifier part, a filter part, an A / D conversion part, a switching part, and an interface part.

The amplifier may amplify the potential difference of the hall sensor using a differential operational amplifier (hereinafter referred to as OP-Amp).

In the present invention, an analog filter is applied to remove unwanted noise components. In other words, since the noise component has a high frequency band, a low-pass filter (hereinafter referred to as analog or LPF) is introduced to filter a defect signal that appears as an irregular waveform. After the differential amplification signal processing, noise was reduced through LPF, and then A / D (analog / digital) conversion was performed.

That is, the leakage magnetic flux distribution resulting from the defect is converted into the distribution of the electrical signal having a high S / N ratio by the arranged magnetic sensor, the differential amplifier circuit, and the analog LPF.

On the other hand, since the microprocessor takes longer to read the converted data than the A / D conversion time, it becomes a factor that makes high-speed signal processing difficult. Therefore, the MDA method can be applied directly to the Asynchronous First-In First-Out (FIFO) buffer memory in synchronization with the A / D conversion pulse signal. Can be.

When a high-speed rail car passes through I-type and O-type sensor structures to measure the entire wheel face, it passes over a sensor consisting of multiple wheel recognition sensors and multiple magnetic sensor arrays. When the wheel is close to the wheel recognition sensor, the change in the magnetic field and the output of the magnetic sensor change only in the region in contact with the wheel, so that the amount of change is measured, amplified and input to the switching unit. The signal of the amplified magnetic field is compared and calculated by a CPLD (Complex Programmable Logic Device), and outputs a control signal to the analog multiplexer. The analog multiplexer activates the high-speed switching MOFET to sequentially activate the sensors on the wheel face, perpendicular to the wheel recognition sensor.

On the other hand, a large number of data is generated per frame when continuously outputting data from the magnetic sensor array installed in the measuring section of the rail in order to measure the entire wheel face, but when using the switching unit per frame A very small number of data can be acquired, which is very useful for data processing and storage.

The interface unit stores the electrical signal converted by the A / D converter and the microprocessor in the computer through the USB interface.

Rigid sensor structure (1) to facilitate installation and maintenance while the wheel recognition device, magnetic field application and high spatial resolution sensor array, amplification and electronic circuits related to A / D conversion and interface are embedded in the rigid sensor structure Was modularized as a unit object. Fig. 13 shows a modular rail-mounted wheel auto flaw detector. An amplification circuit, a switching circuit, an A / D conversion circuit and an interface are integrated in the module.

For example, the module has a total length of 500 mm and a total of 23,079 sensors are arranged in a 471 * 49 matrix form. There are 4 resources (125 * 40 * 5Tmm) in total, and 3 wheels (157 channels) wheel recognition circuit, 3 switches (157 channels) and 21 circuits (32 channels) It is composed. In addition, six (32-channel compatible) A / D converters and one interface are incorporated, and these six modules can be combined to measure one-half of one wheel. Therefore, 224 modules are required to detect the entire left and right wheels.

While the present invention has been described in connection with what is presently considered to be preferred embodiments thereof, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

1: rigid sensor structure 2: vehicle recognition device
3: wheel recognition device 4: computer
5: controller 6: software
10: railroad car 20: wheel
30: sensor for wheel recognition 50: mechanism for fixing the sensor

Claims (9)

Rigid sensor structure consisting of a rail-shaped sensor mechanism;
A vehicle recognition device for recognizing that the railway vehicle enters a measurement standby state on the rigid sensor structure;
A wheel recognition device installed in the rigid sensor structure to detect and output only the corresponding information when the wheel step surface reaches the sensor position;
A control device for switching, amplifying, filtering, converting an electrical signal output from a sensor of the sensor mechanism into a digital signal, and interfacing the converted digital signal to a computer; And
And a software for processing with an embedded algorithm to extract and recognize only defect information from the digital signal input to the computer.
The sensor mechanism is equipped with a plurality of sensors for detecting the wheel tread, and configured to support the flange portion and the half of the wheel tread, and the I-type and the vehicle that can inspect the inside of the tread of the vehicle on half of the rail Unmanned automation system for rail-mounted wheel tread test on the road, characterized in that it comprises an O-type that can inspect the outside of the tread.
delete According to claim 1, The vehicle recognition device is a mechanical proximity sensor that generates a trigger signal when the side of the incoming railway vehicle touches or irradiates the incoming railway vehicle in a non-contact manner to block the light when the railway vehicle passes through Unmanned automation system for rail-mounted wheels, stepping, flaw detection, comprising an optical sensor for generating a trigger signal.
According to claim 1, wherein the wheel recognition device is a rail-mounted wheel response surface unmanned for inspection, characterized in that it comprises a wheel recognition circuit for measuring the magnetic field distribution only in the region in contact with the wheel in the sensor arranged in two dimensions on the surface Automation system.
5. The sensor of claim 4, wherein the sensor includes a magnetic field applying and high spatial resolution sensor that applies the magnetic field evenly to the wheel step surface in a limited space and extracts the distribution of the magnetic field that varies depending on the presence, position, and size of the defect. Unmanned automated system for rail-mounted wheels, flaw detection, characterized in that configured.
The magnetic field applying and high spatial resolution sensor of claim 5, wherein a plurality of magnetic sensors are arranged in a plane on a permanent magnet having a vertical magnetization direction, switching power inputs of the magnetic sensors, and connecting the outputs as one. Unmanned automation system for rail-mounted wheel tread flaw detection, characterized in that the signal processing is performed.
The unmanned automated system for rail-mounted wheel tread test according to claim 6, wherein a spring is interposed between the lower end of the sensor mechanism for fixing the sensor and the surface of the rail to prevent the return of the sensor mechanism and the damage of the sensor. .
2. The unmanned automated system for rail-mounted wheel tread test according to claim 1, further comprising a differential operational amplifier for amplifying a signal in the control device.
7. The rail mounting type wheel answering surface as claimed in claim 6, wherein the wheel recognition device, the magnetic field application and the high spatial resolution sensor, and the circuit included in the control device are used as a module in the rigid sensor structure. Unmanned automated system for inspection.

Images