JPH0629711Y2 - Wheel flat detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention relates to a wheel flat detection device capable of detecting a flat generated on a wheel of a railway vehicle or the like during traveling.

[Conventional technology]

In railway vehicles and the like, when the braking force applied to the wheel during braking becomes larger than the adhesive force between the wheel and the rail, the wheel is locked and slides on the rail, so that the wear is concentrated on a specific part of the wheel outer periphery, A flat abrasion called a so-called flat is generated.

Once this flat occurs, it grows as the vehicle travels, and if it exceeds a certain size, a shock is generated each time the flat contacts the rail, which is harmful to the vehicle (such as falling out of parts or burning of axle bearings). Not only does it generate various vibrations, but it also causes noise pollution to residents along the line.

Therefore, conventionally, the occurrence of flats was discovered at an early stage,
Measures are taken to remove flats by grinding. Specifically, every time the vehicle is loaded, the presence or absence of flatness is inspected visually or by tentacles, and the wheels with flats are refurbished and repaired.

In addition, attempts have been made to quantitatively discover the flats even earlier than these methods and to perform wheel repair and repair at an appropriate initial stage.

As an inspection apparatus therefor, for example, Japanese Patent Laid-Open No. 52-13
There is one described in Japanese Patent No. 1302. This inspection device has an accelerometer installed on the rails or sleepers through which the vehicle passes, and measures the impact generated when the wheel flats roll on the rail surface as the vehicle passes, as the amplitude of acceleration, and measures it. The size of the flat is estimated from the size of the value.

Further, as another inspection device, there is one described in JP-A-58-48809. This inspection device installs a strain gauge on the bottom of the rail, detects the rail strain due to wheel load and the impact strain due to the flat at the same time, and separates the impact component due to the flat from the obtained strain to estimate the size of the flat. It is a thing.

Further, as another inspection device, Japanese Patent Application Laid-Open No. 51-13720.
There is one described in Japanese Patent No. This inspection device installs a strain gauge on the abdomen of the rail, simultaneously detects the shearing force of the rail due to the wheel load and the shock wave component due to the flat superimposed on it, and detects the shock wave amplitude value and wheel load component passing through the high pass filter The size of the flat is estimated from the ratio.

[Problems to be solved by the device]

However, in the case of the flat inspection device using the accelerometer described above, the circuit has a relatively simple structure and can be manufactured at low cost, but since the amplitude of the shock wave generated during traveling varies, sufficient measurement is possible. There is a drawback that accuracy cannot be obtained.

In the case of the flat detector using the above strain gauges, the strain gauges are installed within the existing sleeper intervals (about 40 cm) installed at 40 cm intervals, and the sensitivity range is small, so the effective detection range is about 25 cm. Therefore, in order to inspect one wheel circumference (the circumference is about 270 cm), it is necessary to install at least 11 measuring points to cover the entire circumference, which increases the number of wheel load detectors. At the same time as the cost increases, the detection signal for each point is input, which increases the amount of processing.

In addition, if you try to reduce the number of measurement points and arrange the measurement points across the sleepers, the sensitivity of the sleepers located at the center of the detection range will decrease, and the impact due to the flat will be separated by analog comparison. In this case, the central part of the sensitivity region becomes a dead zone, which causes a problem that sufficient measurement accuracy cannot be obtained.

The present invention has been made to solve these problems, and an object of the present invention is to provide a wheel foot detection device capable of performing a flat inspection with sufficient accuracy even if the number of measurement points is reduced. Especially.

[Means for Solving the Problems]

To achieve the above object, the present invention provides a wheel detector for detecting passage of a wheel, a wheel passage speed calculation means for detecting a passage speed of a wheel from a wheel passage timing detected by the wheel detector, and a rail. And a wheel weight value detector for detecting the wheel weight value of a passing wheel and a wheel weight value storage means for extracting the wheel weight value detected by the wheel weight value detector in a predetermined cycle and storing it in the memory. A moving average calculating means for calculating a moving average value by reading an appropriate number of wheel weight values in the memory according to the wheel speed, and a moving average of the wheel weight values obtained by the moving average calculating means and The difference between the wheel load detection value is calculated, and the flatness of the wheel is calculated by extracting the flat impact component emitted from the flat portion of the wheel and superimposed on the wheel load detection value from the ratio of the difference and the moving average. And the flat size calculation means Is obtained is characterized by comprising and an output means for outputting the magnitude value of the obtained flat by the flat magnitude calculation means by the print or the like.

[Action]

In the wheel flat detection device configured as described above, when the wheel detector detects the passage of a wheel, the wheel passage speed calculation means calculates the passage speed of the wheel from the detected wheel passage timing. The wheel weight value detector installed on the rail detects the wheel weight value of the passing wheel, and the wheel weight value storage means extracts the detected wheel weight value at a predetermined cycle and stores it in the memory.

The moving average calculating means reads a suitable number of wheel weight values stored in the memory according to the wheel speed and calculates a moving average.

The flat size calculating means obtains the difference between the moving average of the obtained wheel load values and the wheel load detection value at that time, and the wheel load detection value is emitted from the flat portion of the wheel from the ratio of the difference and the moving average. The flat impact component superimposed on is extracted and the flat size of the wheel is calculated.

The output means outputs the size of the foot thus obtained by printing or the like.

〔Example〕

Hereinafter, embodiments of the present invention will be specifically described with reference to the accompanying drawings.

FIG. 2 (a) is an explanatory view showing an arrangement example of the wheel detectors 1 and 2 and the shear force detectors 4a, 4b as wheel weight value detectors, and FIG. 2 (b) is a shear force detector. It is a wave form diagram which showed the detection sensitivity of 4a, 4b.

As shown in Fig. 2 (a), the rail 12 on which the wheels 11 supporting the vehicle roll is laid on the sleepers 13a to 13d. Wheel detectors 1 and 2 are disposed in front of the sleepers 13a and behind the sleepers 13d, and positions on the rail 12 corresponding to the positions between the sleepers 13a and 13b and between the sleepers 13c and 13d. The shear force detector 4
a and a shear force detector 4b are provided. The total length L of the measurement section is set to approximately four times the laying pitch of the sleepers 13. The wheel detectors 1 and 2 detect the passage of wheels by magnetism, infrared rays, or the like. The passing time and the passing speed of the vehicle can be calculated from the difference between the distance L between the wheel detectors 1 and 2 and the passage detection timing of the wheel detectors 1 and 2.

The shear force detectors 4a and 4b are, as shown in FIG. A on the neutral axis N.A. Two pairs of strain gauge elements g ₁ mounted orthogonal to each other at an angle of 45 ° and 135 ° with respect to A, respectively.
And consists g ₂ and g ₃ and g _4, by these strain gauge elements g ₁ to g ₄ constitute a Wheatstone bridge as shown in Figure 4 to apply the bridge voltage E, generated in the rail 12 Shear A voltage signal proportional to the force is obtained at the output. Since the sleeper 13 sinks due to the wheel load in a normal track, the shear force sensitivity region has a considerable spread.

The shearing force thus detected by the shearing force detector 4a increases or decreases depending on the position of the wheel 11, and as shown by the waveform e in FIG.
The value increases negatively as 1 approaches, and the shear force detector 4
The polarity is reversed at the position of a, the positive maximum value is obtained immediately after passing, and then gradually decreases, resulting in a symmetrical waveform. In general, when these shear force detectors detect the passing wheel load, it is explained that the shear force changes rapidly from negative to positive. In addition, when the wheel 11 and the rail 12 are in surface contact with each other and the shear force detector has a length, when the value of the shear force to be detected is reversed, as shown in FIG. Thus, the waveform is inclined.

Similarly, the shear force detector 4b can also detect the characteristic indicated by the waveform e. The waveform f'indicated by the solid line is an inversion of the waveform f, and in actual processing, the waveforms e and f'indicated by the solid line are used to perform the flat size detection process.

FIG. 1 is a block diagram showing a configuration of a wheel flat detecting device according to the present invention.

In the figure, reference numerals 1 and 2 denote wheel detectors, and when a wheel passes over the upper rail, a signal for detecting the wheel is output.

Reference numeral 3 denotes a truck signal generator, which is a signal indicating that the truck is passing from the wheel detection signals h and j output from the wheel detectors 1 and 2 as shown in FIG. 6 (hereinafter, referred to as "truck signal"). Output k.

Reference numeral 4 denotes a shear force detector as a wheel load value detector, which detects the wheel load applied to the rail as a shear force. A shock wave component generated when the wheel flat rolls on the rail surface is superimposed on this shear force.

An amplifier 5 amplifies the shear force detection signal detected by the shear force detector 4 at a predetermined magnification.

Reference numeral 6 denotes a sample / hold circuit, which samples and holds the analog shearing force detection signal amplified by the amplifier 5 at a predetermined interval under the control of a signal processing unit 8 described later.

Reference numeral 7 is an A / D converter, which similarly controls the amplified shearing force detection signal into a digital signal under the control of the signal processing unit 8.

Reference numeral 8 denotes a signal processing unit, which is specifically configured by a microcomputer, and when a trolley signal k indicating that the trolley is passing is input from the trolley signal generator 3, calculates the passing speed of the trolley. The sample / hold circuit 6 and the A / D converter 7 are operated to take in the signal detected by the shearing force detector 4 and perform the process of calculating the magnitude of the flat impact component.

Reference numeral 9 denotes a printer which prints out the calculated magnitude of the flat impact component, the bogie position and the bogie speed at which the flat wear occurred, as well as the passing date, hour and minute of the train, the number of vehicles, and the like.

FIG. 5 and FIG. 6 are explanatory views showing signal processing in the truck signal generator 3.

Consider a case in which a bogie consisting of wheels 11 and wheels 11 ′ having an interaxial distance of 1 passes through a measurement section having a total length L.

In the wheel detector 1, first the wheel 11 and then the wheel 1
The passage of 1'is detected, and a signal h composed of the pulses S1 and S2 at the respective passage timings is output.

Similarly, the wheel detector 2 detects the passage of the wheel 11 and then the wheel 11 'later than the wheel detector 1, and outputs a signal j composed of pulses S3 and S4 at the passage timings.

After the pulse S1 is detected in this way, the pulse S4
Since the period until the vehicle is detected is the actual passing time of the truck, the truck signal k having the H level only during the passage of the truck is synthesized from the signals h and j.

As a result, the wheel 1 detected by the shear force detectors 4a and 4b
The shear force detection signal m of 1 and the wheel 11 'is taken in as an effective value only while the truck signal k is at H level. That is, the signal processing unit 8 monitors only the truck signal k, and starts the processing after detecting the rising of the H level or the rising edge of the truck signal k. Further, the interval between the pulse S1 and the pulse S4 of the wheel detection signals h, j is measured by an internal timer or the like, and the time T, the measurement section length L and the wheels 11,1 are measured.
The passing speed V can be obtained from the inter-axis distance 1 of 1'by the following equation.

V = (L + 1) / T FIG. 7 is an explanatory diagram showing the processing contents in the signal processing unit 8.

In the present invention, in order to separate and evaluate the shock wave component generated due to the flatness, the shear force waveform on which the shock wave component is superimposed is sampled at a constant cycle while the truck is passing, and A / D conversion is performed. The data is temporarily stored in a memory in the signal processing unit 8 as detection data of a digital value.

That is, in addition to the shear strain generated on the rail 12 due to the wheel load, the flat shock wave component generated on the outer circumference of the wheel is superimposed on the waveform indicating the temporal change in the shear force detected by the shear force detectors 4a and 4b.

FIG. 7 (a) shows an example of a change in shear strain including a flat shock wave component input to the signal processing unit 8. The flat shock waves F1 and F2 caused by the two flats generated on the wheel outer periphery are Since the flat shock wave F1 is generated, the flat shock wave F1 comes into contact with the rail surface at a position on the left side in the figure, that is, at a position distant from the shear force detector 4, so that the shock component is detected to be relatively small. On the other hand, since the flat shock wave F2 comes into contact with the rail surface at the maximum amplitude part of the shear force waveform near the center, that is, at a position close to the shear force detector 4, the shock component is detected relatively large.

FIG. 7 (b) shows a flat shock wave component approximately by moving average from the temporal fluctuation of the shear force including the flat shock wave component detected by the shear force detector 4 shown in FIG. 7 (a). The fluctuation waveform of the shearing force when removed is shown.

When the moving average of the shearing force is sequentially calculated based on the sampling number N, the shock wave components of the flat shock waves F1 and F2 are leveled to slight levels as shown in the figure. Here, the two waveforms are superimposed and the non-matching portion is extracted.
That is, by calculating the difference between the two, the flat shock wave F
The impact components of 1 and F2 can be taken out.

Specifically, moving average processing is performed in order from the beginning of the stored data by the number N corresponding to the vehicle speed, the i-th data (amplitude) from the beginning of the stored data is set to Pi, and the moving average is calculated (i-N ) The numerical value (amplitude) is P _1-N , and the coefficient R is calculated by the following equation.

R = (Pi−P _1−N ) / P _{1−N Since} this coefficient R is obtained by dividing the difference between the sampled value and the moving average at the same timing by the moving average value, the shock wave component is superimposed. If not, it will be almost 0,
If the shock wave component is superposed, the value becomes proportional to the shock wave component. The process of obtaining the coefficient R is sequentially performed from the beginning of the data, and if a large value is obtained, it means that the wheel is flat. Since this coefficient R is the ratio of the shock wave amplitude to the wheel load, the same coefficient R is always obtained from the same flat regardless of the distance between the impact position of the wheel flat and the detector position, and within the shear force detection range. The detection sensitivity of the shock amplitude becomes uniform. Further, the obtained coefficient R can be subjected to processing such as correction based on the magnitude of the passing speed V to estimate the magnitude of flat wear.

According to experiments conducted by the inventors, the amplitude of the shock wave component generated by the flat wear increases in proportion to the vehicle speed until the vehicle speed becomes almost constant (about 40 km / h). By calculating V and separately multiplying the coefficient R by an empirically obtained speed correction coefficient, it is possible to more accurately estimate the magnitude of flat wear.

Similarly, if the sampling period of the sample and hold circuit 6 is constant, the data of the detection value of the shearing force taken in by the passing velocity V of the wheel will be coarse and fine, so that the target of the moving average is almost inversely proportional to the velocity V. The number N of data to be changed is changed.

The signal processing unit 8 performs processing such as wheel passing speed calculation, speed correction coefficient calculation, moving average processing, flat wear magnitude calculation according to the passing speed, and flat size discrimination evaluation, and the like. Is output from the printer 9.

The output items (printed items) from the printer 9 are the train passing date and time, the number of vehicles, the bogie position where flat wear has occurred, etc. in addition to the above processing results, and are printed out after the train has passed.

FIG. 8 is a flowchart explaining the operation of the signal processing unit 8. The operation of the signal processing unit 8 will be systematically described below based on this flowchart.

When a predetermined initial process is completed by turning on the power, the signal processing unit 8 monitors the truck signal k input from the truck signal generator 3 (step 801).

When the trolley signal k is inverted from the L level to the H level due to the passage of the trolley (Yes in step 801), the time measurement process of the internal timer is started (step 802).

At the same time, a start signal is sent to the sample and hold circuit 6 and the A / D converter 7 to start the operation (step 80).
3).

As a result, the value detected by the shear force detector 4 is input as a digital signal from the A / D converter 7 to the signal processing unit 8 at regular intervals.

The input shear force detection value is sequentially stored in the internal memory until the truck signal k is inverted to L level (steps 804 and 805).

Next, when it is detected that the truck signal has been inverted to the L level (Yes in step 805), the internal timer is stopped, the wheel passing speed V is calculated from the measured time T and the measured section length L up to that time, and the internal memory is stored. (Step 80)
6).

At the same time, the operations of the sample and hold circuit 6 and the A / D converter 7 are temporarily stopped (step 807).

Then, the optimum number N of shearing force detection values are sequentially read from the internal memory according to the speed V to calculate a moving average value (step 808).

Furthermore, the calculated moving average value is compared with the data used for obtaining the moving average to obtain the difference, and the ratio of this difference to the corresponding moving average value gives a coefficient that is approximately proportional to the size of the flat. R is calculated (step 809).

Further, the coefficient R is corrected according to the magnitude of the speed V, and the magnitude of flat wear is estimated (step 810).

Next, the signal processing unit 8 outputs the calculated speed V, the amount of flat wear, and other output data to the printer 9.
To output at an appropriate timing (step 81).
1).

With these processes, the inspection of the flat when one truck passes is completed, and the process returns to step 801 to wait for the next truck to pass.

As described above, according to this embodiment, it is possible to provide a relatively simple and inexpensive wheel flat detection device by reducing the number of detectors to be installed, thereby reducing the occurrence rate of failures and performing maintenance. It is possible to obtain an easy wheel flat detection device.

The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the invention.

For example, in the embodiment, two shear force detectors are installed, but if the detection range of one shear force detector can cover the entire outer circumference of the wheel, it is possible to use one shear force detector. Is.

[Effect of device]

As described above in detail, according to the present invention, the number of wheel weight value detectors to be installed is very small, the simple configuration has a low failure rate, and the inexpensive and highly accurate flat inspection is possible. A possible wheel flat detection device can be provided.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a wheel flat detecting device according to the present invention, FIG. 2 (a) is an explanatory view showing an arrangement example of a wheel detector and a shear force detector, and FIG. 2 (b). )
Is a waveform diagram showing the detection sensitivity of the shear force detector, FIG. 3 is an explanatory diagram showing a mounting example of the shear force detector, and FIG. 4 is a connection diagram of strain gauge elements constituting the shear force detector. 5 and 6 are schematic diagrams and timing charts for explaining the signal processing contents of the truck signal generator, and FIG. 7 is a waveform diagram for explaining the processing contents in the signal processing unit, and FIG. The figure is a flowchart for explaining the operation of the signal processing unit. 1, 2 ... Wheel detector, 3 ... Truck signal generator, 4, 4a, 4b ... Shear force detector, 5 ... Amplifier, 6 ... Sample and hold circuit, 7 ... A / D converter , 8 ... Signal processing unit, 9 ... Printer, 11, 11 '... Wheels, 12 ... Rail, 13a-13d ... Sleepers, h, j ... Vehicle detection signal, k ... Bogie signal, g ₁ ~ G ₄ ... Strain gauge element, m ... Shear force detection signal, F1, F2 ... Flat shock wave, V ... Wheel passing speed, R ... Coefficient.

Claims (1)

[Scope of utility model registration request] 1. A wheel detector for detecting passage of a wheel, wheel passage speed calculation means for calculating a passage speed of a wheel from a wheel passage timing detected by the wheel detector, and a wheel passage installed on a rail. A wheel load value detector for detecting a wheel load value, a wheel load value storage means for extracting the wheel load value detected by the wheel load value detector in a predetermined cycle, and storing it in a memory, and a wheel load value in the memory. An appropriate number of values are read according to the vehicle speed to calculate a moving average value, and a difference between the moving average of the wheel weight values obtained by the moving average calculation means and the wheel weight detection value at that time is calculated. Obtained from the ratio of the difference and the moving average, the flat size calculation means for calculating the flat size of the wheel by extracting the flat impact component emitted from the flat part of the wheel and superimposed on the wheel load detection value. Calculating the size of this flat Wheel flat detection apparatus characterized by comprising output means for outputting the like printed value of the flat size obtained by the means.