JP2008120258A - Detection system of wheel tread state - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To specify a position of an abnormal wheel for each axle, in a system detecting the wheel tread state of a railroad train based on a traveling sound. <P>SOLUTION: Main components of a detection system of a wheel tread state include: a sound collecting device disposed immediately under an elevated bridge; two beam sensors arranged along a rail tracks; an IC tag on the train and an ID tag reader along a railway line; and a signal processing means. The signal processing means takes out an AC waveform signal from sound signals collected by the sound collecting device, and passes it through a band pass filter of a predetermined frequency band to be an extracted waveform. The extracted waveform is collated with a train speed calculated from outputs of the beam sensors and axle positions determined from the unit number of the ID tag. The right or wrong of the wheel tread state can be detected for each axle, and of the detection data are transmitted to a management PC such as a vehicle site. The axle position is easily and correctly specified when the temporal rolling and abrasion of the abnormal axle is performed. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a system for determining the quality of a wheel tread state in a railroad train based on traveling sound, and more specifically, to provide a system that can grasp the quality of a wheel tread state individually for each axle. And

  If surface wear such as flat wear or unevenness is present on the wheel treads of a railroad train, traveling noise such as rolling noise and vibration noise increases, so the environment along the railway line is deteriorated. In addition, there is a problem that riding comfort is lowered because noise in the vehicle and body vibration increase. These problems are items that require countermeasures especially for high-speed trains such as the Shinkansen. In view of this, the present applicant has proposed in Japanese Patent Application Laid-Open No. 2004-228787 a system that can measure the traveling sound of a train at an appropriate location along the line and detect the wheel tread state from the sound pressure level of the traveling train sound.

In the system described in Patent Document 1, a sound collecting device (microphone) is installed at a position of about 1 m immediately below the viaduct, and a sound signal collected by the microphone is subjected to predetermined processing to change a sound pressure level over time. It is converted to a direct current waveform signal (DC waveform), and the peak value appearing in the obtained DC waveform is compared with a reference value to detect the presence or absence of an abnormal wheel.
JP 2006-155536 A

  In the system of Patent Document 1, each peak (mountain) of the DC waveform is formed by front and rear wheel groups (two-axis four-wheel or four-axis eight-wheel) sandwiching the vehicle connecting portion, and the bottom (valley) is almost at the center of the vehicle. Due to the nature of the correspondence, two adjacent vehicles are the minimum unit for position identification, and it is impossible to separate each vehicle. Therefore, when an abnormal wheel is found, the position of the vehicle to which it belongs can only be roughly specified, so it takes time to accurately specify the abnormal wheel position. Further, since the DC waveform is created from the entire sound collected by the microphone, it may include foreign sound data other than train running sound, which may cause erroneous determination of the wheel tread state. Furthermore, when a clear mountain-valley shape does not appear in the waveform, it becomes difficult to specify the vehicle position.

  The present invention is a wheel that can improve the prior art and can grasp the wheel tread state in units of axles, and can extract only the wheel rolling sound by eliminating the influence of foreign sounds other than the train running sound. It aims at providing the detection system of a tread state. A feature of the detection system according to the present invention is that, as described in claim 1, a sound collecting device that is installed directly below an elevated structure on which a track is laid and collects a traveling sound of a train, and a sound collecting device Signal processing means for processing the output audio signal, and speed detection means for detecting the speed of the train, in the signal processing means, the AC waveform signal is extracted from the audio signal, the obtained AC waveform signal, Passing through a bandpass filter of a predetermined bandwidth set based on the train speed and structure characteristics obtained from the speed detection means, and regarding the extracted waveform signal after passing through the bandpass filter, the train speed data and the train organization By collating the train formation data, the sound pressure level is individually measured for each axle.

  In the above, the elevated structure is typically a viaduct. A microphone is used for the sound collector. It is desirable that the sound collecting device be installed at a slight distance (for example, about 1 m) from the lower surface rather than being in close contact with the lower surface of the elevated structure. This is because it is known that when directly attached to the lower surface of the elevated structure, the sensitivity changes depending on the contact state, and the reproducibility of the detection data for the same train may decrease. The distance from the bottom surface of the structure is set so that the sound can be collected at a sound pressure level as high as possible and the waveform of the collected audio data can be easily distinguished from peaks and valleys. . Usually, the sound waveform forms a peak at the axle position of the carriage and forms a bottom between the carriages. If it is too close to the structure, the distance attenuation of the sound is reduced, and the difference between the peak and the bottom becomes unclear. However, if the distance from the structure is too long, there is a possibility that external factors are mixed due to the influence of external noise. Specifically, it is appropriate that there is an interval of 200 ms or more between the peak and the bottom, and that there is a level difference of about 10 to 15 dB. In order to clearly discriminate noise based on the abnormal state of the wheel tread surface, it is desirable to select a location that is operated at a speed as close to the maximum speed as possible as the installation location of the sound collector.

  The audio signal processing means mainly includes a waveform processing unit that extracts an AC waveform signal from the audio signal, and a calculation unit that performs a predetermined calculation process on the AC waveform signal, and these may be integrated or separate. The signal processing means may be installed in the vicinity of the sound collecting device and used as a local device. However, if appropriate communication means is provided between the sound collecting device and the sound collecting device, the signal processing means is installed in a remote place away from the sound collecting position. It is also possible. The waveform processing unit of the signal processing means has a function of extracting an AC waveform signal from the train running sound, but may have a function of performing general processing such as amplification and audibility correction as necessary. An electronic computer such as a personal computer (PC) can be used for the calculation unit, and a function for processing the extracted AC waveform signal through a bandpass filter (BPF), an extracted waveform signal after passing through the BPF, and a train It has a function to collate speed data and train organization data, but also has a function to output the collation result to a display or printer, a function to judge the presence or absence of an abnormal state of the wheel tread based on the collation result, and a collation result A function of transmitting the determination result to a computer installed in another place may be appropriately provided.

  The frequency bandwidth of the bandpass filter applied to the waveform extraction process is set based on the train speed and the structure characteristics. Underpass noise caused by abnormalities on the tread surface of a wheel appears when vibration due to wheel rotation propagates through the structure. Therefore, the frequency of underpass noise is mainly determined by the wheel rotation frequency and the natural frequency of the structure. The wheel rotation frequency can be obtained by dividing the train speed by the circumference of the wheel tread. (To give a specific example, the driving speed V of the Shinkansen is 300 km / h≈833333.3 mm / s, and the wheel diameter D of the Shinkansen is 860 mm, so the wheel tread circumference E = D × π (circumference) ≈2701.8 mm Therefore, the wheel rotation frequency is V ÷ E≈30.8 Hz.) However, since the natural frequency of the structure is often not easily obtained, the frequency that affects the overhead noise can be specified by testing. That's fine. For example, at a specific viaduct, when the Shinkansen travels at a speed of 300 km / h (when the wheel rotation frequency is 30.8 Hz), it is known that an abnormality in the state of the wheel tread affects the frequency band around 200 Hz. Therefore, the frequency of BPF applies a band including 200, that is, 180 to 250 Hz.

  According to a second aspect of the present invention, the speed detection means includes two beam sensors that are arranged at a predetermined interval along the track direction and radiate detection light to the side surface of the vehicle body at a height position of the front end portion of the vehicle body of the train It is good also as what consists of. In this case, it is possible to calculate the train speed from the rise time difference of each beam sensor. Visible red light or infrared light can be used as the detection light of the beam sensor, but the wavelength used can be appropriately selected according to the situation.

  According to a third aspect of the present invention, the vehicle body and inter-vehicle position of the train may be detected by the beam sensor.

  According to a fourth aspect of the present invention, one of the beam sensors and the sound collecting device are arranged in the same cross section perpendicular to the track so that the sound collecting device operates when the beam sensor detects a train. May be set.

  Further, as described in claim 5, an ID tag having train formation data recorded is arranged on the side of the vehicle, acquired from the ID tag by an ID tag reader arranged in the vicinity of the track, and the train formation data is output to the signal processing means. You may make it do. In addition, it does not prevent that train organization data is previously acquired from PRC (operation management device) or CTC (central control device) and stored in the signal processing means.

In the detection system according to the first aspect of the present invention, since the sound collecting device is installed directly under the elevated structure, the installation work / maintenance work becomes work outside the track, and is excellent in safety. In addition, the position just below the elevated structure is less susceptible to aerodynamic noise generated from the train's car body and other external sounds, the sensitivity of the sound collector is stable, and data reproducibility. It has advantages such as being good.
Since an AC waveform signal extracted from an audio signal output from the sound collector is used, frequency analysis and BPF processing are possible. By passing the AC waveform signal through a BPF having a predetermined bandwidth, an extracted waveform signal specialized for wheel rolling sound can be obtained. By collating this extracted waveform signal with train speed data and train organization data, the time axis of the extracted waveform indicating the temporal change of the sound pressure level can be replaced with vehicle position information. Therefore, since the axle position can be specified on the extracted waveform, the sound pressure level can be measured for each axle.
Due to the above characteristics, it becomes easy to identify a wheel having an abnormality in the tread surface state, so that the efficiency of the temporary correction work can be improved. When a wheel abnormality suddenly occurs, this can be detected at an early stage, so that it can be quickly reflected by reflecting it in the inspection plan. Since the data is collected every time the train passes through the installation location of the detection system according to the present invention, the wheel tread state is managed in time series in units of axles, carts, or trains by accumulating the detection data and creating a database. As a result, it is possible to abolish the complementary wheel inspection that has been performed in the past, and to reduce the work manpower and expenses associated therewith.

  According to a second aspect of the present invention, if the speed detecting means is constituted by two beam sensors arranged at a predetermined interval along the line direction, speed data with high reliability can be obtained at low cost.

  According to the third aspect of the present invention, if the vehicle body and inter-vehicle position of the train are detected by the beam sensor, the obtained vehicle body and inter-vehicle position data and the extracted waveform are collated, and the wheel position accuracy is verified. Can be increased.

  When one of the beam sensors and the sound collecting device are arranged in the same cross section perpendicular to the track, and the sound collecting device is set to operate when the beam sensor detects a train, Thus, it is easy to associate the obtained audio signal waveform with the train position, and the problem of erroneously recording the external noise generated when the train does not pass can be avoided. Moreover, since data collection is executed only when the train passes, the capacity of the storage device necessary for data recording can be reduced.

  According to a fifth aspect of the present invention, an ID tag having train formation data recorded is arranged on the side of the vehicle, and train formation data is acquired from the ID tag by an ID tag reader arranged in the vicinity of the track, and this is output to the signal processing means If this is done, the accuracy of data relating to train organization can be maintained. For example, train formation data can be acquired from a PRC (operation management device) or CTC (central control device) and stored in a signal processing means. In this case, the train operation diagram is changed after data acquisition. If a train is different from the data stored in the signal processing means, there is a possibility of using wrong train formation data. However, as in this example, if the train organization data is acquired from the ID tag arranged for each train, the above problem can be avoided.

  A case where the wheel tread surface detection system according to the present invention (hereinafter referred to as the present system) is applied to a Shinkansen will be described. 1 to 4 conceptually show a schematic configuration of the system of the present invention, and a sound collecting device (microphone) (FIG. 2) disposed directly under an elevated structure (bypass) on which a train travels, a train body A beam sensor (FIG. 3) that emits detection light (visible red light) toward the side, an ID tag placed on the side of the train, an ID tag reader installed along the railway (FIG. 4), and audio data and beams output from the microphone The signal processing means for executing a predetermined process based on the position data output from the sensor and the train organization data acquired from the ID tag is a main component. Further, a management PC for transmitting the processing result of the signal processing means through a communication line such as the Internet is provided.

  As shown in FIG. 2, the microphone which is a sound collector is arrange | positioned in the position of about 1 m from the viaduct lower surface. As a means for installing the microphone, a method using a support attached to a bridge pier, a method using a column provided on a facility roof under a viaduct, or the like can be considered.

  As shown in FIG. 3, the beam sensor is arranged using a soundproof wall on the side of the railway, for example. The type of beam sensor is a reflection type optical sensor, and the wavelength used is visible red light or infrared light. The height position of the beam sensor is the height position of the train head (nose), and the two beam sensors are arranged at regular intervals along the track direction.

  The microphone on the lower viaduct and the one beam sensor on the upper viaduct are, as a rule, arranged so as to be located in the same vertical section in the direction orthogonal to the line direction.

  As shown in FIG. 4A, the ID tag is provided at an appropriate position of the vehicle, and the attachment position may be either the vehicle body part or the carriage part. Usually, only one ID tag is required per train train. The data recorded in the ID tag relates to train formation and is a formation number unique to the vehicle. Based on this organization number, necessary data such as train length and axle position can be acquired from a separately prepared database. An ID tag reader for acquiring data from an ID tag is installed using a soundproof wall on the side of the railway. The holder for holding the ID tag reader may be provided with a hinge or the like so that the posture can be adjusted, and if necessary, the holder can be tilted by an appropriate angle θ so that it can face the ID tag.

  By the way, the distance L between the ID tag and the ID tag reader is determined by the installation position of the both, and when the distance L is determined, the communicable range of the ID tag reader is determined. It is necessary to have a length capable of securing a time for which a large amount of communication is possible. For example, assuming that the communicable area of the ID tag reader is a range as shown on the left side of FIG. 4B, and the width of the communicable range when the distance between the ID tag and the ID tag reader is L is 900 mm / hour. The time required for the Shinkansen traveling at 300 km to pass 900 mm is 10.8 ms. On the other hand, it is assumed that the time required for the ID tag reader to acquire 6-byte data from the ID tag is 11 ms. Under such conditions, accurate data acquisition may fail, so the train speed must be reduced or the amount of communication data must be reduced. Therefore, as shown on the right side of FIG. 4B, the antenna angle (radiation angle) of the ID tag reader is slightly inclined with respect to the train traveling direction. By tilting 15 degrees as shown in the figure, the width of the communication range of the ID tag reader can be expanded to 990 mm. Thereby, it is possible to secure a communication time of 11 ms between the ID tag and the ID tag reader until the train speed is up to 324 km / h.

  The system of the present invention collects traveling sound when a train passes over a microphone and outputs voice data to the signal processing means. Note that the sound collection operation by the microphone may be automatically started when the front end of the train is detected by the upper beam sensor, and automatically stopped when a predetermined time has elapsed after passing through the rear end of the train. The audio signal collected by the microphone is processed by a waveform processing unit as shown in FIG. 5, for example. A waveform processing unit shown in the figure amplifies an audio signal with a preamplifier, and after performing an A characteristic correction with an auditory sensation correction unit, takes out and outputs an AC waveform signal (AC waveform). Further, the waveform processing unit includes a dynamic characteristic circuit including a peak detection circuit and a logarithmic operation circuit, and also has a function of outputting an AC signal as a DC waveform signal (DC waveform) obtained by converting the AC signal in dB. The DC waveform signal can be used as an index indicating the audible noise level.

  The beam sensor arranged on the side of the railway line detects the vehicle body of the train and outputs a detection signal. Therefore, the beam sensor is always in an ON state except that the detection signal is interrupted at the connection part (inter-vehicle part) between the vehicles. Therefore, the output signal waveform of the upper beam sensor is a vehicle body and inter-vehicle detection waveform, and basically the ON state continues and the OFF state appears at regular intervals. The position in the OFF state is the inter-vehicle position. FIG. 6A shows the vehicle body and inter-vehicle detection waveforms. On the other hand, train organization data can be acquired based on the vehicle organization number read from the ID tag, and the axle position can be known from this train organization data. The axle position is displayed in FIG. This graph represents two axles provided on one carriage by three lines. The wheel position can be specified for each axle from the output line indicating the inter-vehicle distance and the axle position display line. The output signal and axle position data from the beam sensor are sent to the signal processing means and used for signal processing to be described later.

  The signal processing means is based on the AC waveform signal obtained from the audio signal, the vehicle body and inter-vehicle distance detection data obtained from the beam sensor, the axle position information, and the train organization data obtained from the ID tag as follows. And signal processing. First, an AC waveform signal is extracted from the audio signal collected by the microphone. The extracted AC waveform and the DC waveform signal shown for reference are as shown in the graphs of FIGS. In each graph, the left side is the front of the traveling direction and the right side is the rear of the traveling direction. In these graphs, axle position display lines are superimposed. Although the DC waveform reflects an increase or decrease in the audibility of the train running sound, it cannot be easily felt whether there is an abnormality in the wheel tread state. Moreover, it is difficult to specify the position of the abnormal wheel. On the other hand, with the AC waveform as it is, it is difficult to determine the presence and position of an abnormal wheel.

  In the output waveform signal of the beam sensor, the continuous range represents the train detection time. Train length data is given from an ID tag attached to the train. The signal processing means calculates the train speed from the rise time difference and the installation interval of two beam sensors arranged at an appropriate interval in the track direction.

  The AC waveform signal is passed through a bandpass filter BPF in a frequency band set based on the test. In the case of a Shinkansen that travels at a speed of 300 km / h, it was appropriate to set the frequency band of the BPF to 180-250 Hz on a certain viaduct. An extracted waveform obtained by the BPF passage process is shown in FIG. Since the extracted waveform is a waveform specialized for wheel rolling sound, it is possible to easily determine whether there is an abnormality in the wheel tread state from the magnitude of the sound pressure level shown in this graph. According to the graph of FIG. 7C, it can be seen that there is an abnormality in the wheel of the front axle in the 16th vehicle from the front. In addition, for example, the crest factor (CF = peak value / running value) is used as a criterion for determining the degree of abnormality. If CF> 6.0, it is determined that a sudden tread abnormality has occurred suddenly and an immediate rolling request is made. However, when 6.0 ≧ CF> 3.0, it is determined that the tread surface is in a state where attention is required, and the examination of planned rolling is requested, and when 3.0 ≧ CF, it is determined that it is normal. It is done. The above is an example of the abnormality determination method, and the present invention is not limited to this.

  The wheel tread surface detection data thus obtained is transmitted to a management PC installed in a vehicle place or the like via a communication line as appropriate (see FIG. 1). Then, if a train that is determined to require turning is routed to the vehicle station, temporary rolling of the wheel having an abnormality is executed based on the detection data. At that time, the position of the abnormal wheel can be easily and accurately specified, so that the rolling operation becomes extremely efficient.

  In addition, the detection system of this example collects detection data by fixed point monitoring and is excellent in data reproducibility. Therefore, by accumulating the detection data in the management PC, a highly reliable database regarding the wheel tread condition can be obtained. Can be built. As a result, the wheel condition can be managed in time series in units of axles for each train formation, so that an efficient rolling plan can be made.

  Note that it is also possible to transmit data or commands from a management PC installed in a vehicle place or the like to a signal processing means on site where the detection system of the present invention is constructed. For example, there may be a case where a command for changing the collection start condition of train traveling sound is output from a vehicle place.

BRIEF DESCRIPTION OF THE DRAWINGS It is related with one Embodiment of this invention, Comprising: It is drawing which shows notionally the schematic structure of a detection system. It is a front view explaining the installation condition of a microphone (sound collector) regarding one embodiment of the detection system of the present invention. It is a front view explaining the installation condition of a beam sensor regarding one embodiment of the detection system of the present invention. Fig. (A) is a drawing for explaining the installation status of ID tags and ID tag readers in the detection system of the present invention, and Fig. (B) is a drawing for explaining means for expanding the communicable area in the ID tag reader. It is a block diagram which shows an example of the waveform process part in a signal processing means regarding one Embodiment of this invention detection system. Regarding an embodiment of the detection system of the present invention, FIG. (A) is a graph showing vehicle body and inter-vehicle part detection waveforms obtained by a beam sensor, FIG. (B) is a graph showing axle position display lines, and FIG. It is a graph which superimposes and shows A) and (B). Regarding an embodiment of the detection system of the present invention, FIG. (A) is a graph showing a DC waveform obtained from a collected audio signal, FIG. (B) is a graph showing an AC waveform extracted from the audio signal, and FIG. It is a graph which shows the extraction waveform obtained by giving the pass process of a band pass filter to a waveform.

Claims (5)

  A sound collector that is installed directly under the elevated structure where the track is laid and collects the running sound of the train, a signal processing means that processes the audio signal output from the sound collector, and a speed detector that detects the speed of the train Means for extracting an AC waveform signal from the audio signal and setting the obtained AC waveform signal on the basis of the train speed and the structure characteristics obtained from the speed detection means. The sound pressure level is measured individually for each axle by passing through the band-pass filter of the bandwidth and collating the extracted waveform signal after passing through the band-pass filter with the train speed data and train formation data relating to train formation. This is a wheel tread state detection system.   The said speed detection means is arrange | positioned at predetermined intervals along a track direction, and consists of two beam sensors which irradiate a detection light to the vehicle body side surface in the height position of the vehicle body front-end | tip part of a train. Wheel tread condition detection system.   The wheel tread state detection system according to claim 2, wherein the vehicle body and inter-vehicle position of the train are detected by the beam sensor.   4. The wheel according to claim 3, wherein one of the beam sensors and the sound collecting device are arranged in the same cross section perpendicular to the track, and the sound collecting device is set to operate when the beam sensor detects a train. Tread surface detection system.   The wheel tread state according to claim 1, wherein an ID tag that records train formation data is arranged on a side of the vehicle, and train formation data acquired from the ID tag by an ID tag reader arranged near a track is output to the signal processing means. Detection system.