Wheel uneven wear detection system for railroad vehicle and wheel uneven wear detection method for railroad vehicle

Abstract

To provide a technique capable of properly making a truing plan for wheel with improved determination accuracy for a wheel uneven wear detection system for railroad vehicle that detects wheel uneven wear based upon vibrations during a vehicle travel.SOLUTION: A wheel uneven wear detection system 1 for railroad vehicle samples, by a sampling part 221, a measured value of an acceleration sensor 21 at an interval of a fixed distance while a railroad vehicle 41 travels in a predetermined section. An arithmetic part 222 calculates a vibration processing value due to wheel uneven wear from measured values sampled by the sampling part 221. A management device 23 accumulates and stores vibration processing values for the predetermined travel section. A statistical processing part 122 statistically processes the vibration processing values for the predetermined travel section that the management device 23 accumulates and stores. A determination part 123 compares the statistical processing values that the statistical processing part 122 calculates with a threshold value TH to determine a state of wheel uneven wear.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a railway vehicle wheel uneven wear detection system and a railway vehicle wheel uneven wear detection method for detecting uneven wheel wear of railway vehicle wheels. [Background technology]

[0002] Railway vehicle wheels (hereinafter referred to as "wheels") wear as the vehicle travels. When uneven wheel wear, in which the tread wears in a circumferentially wavy pattern, occurs on the wheel, it increases vibrations while the vehicle is traveling, potentially leading to a deterioration in ride comfort and damage to bogie components. This problem can occur even when the depth of uneven wheel wear is slight, less than 1 mm for a wheel with a diameter of 860 mm. It is impossible to detect such slight uneven wheel wear visually or otherwise. For this reason, wheels have been periodically re-ground at regular intervals or over a set distance to suppress vibration. However, uneven wheel wear can sometimes occur before the re-grounding cycle. For this reason, technologies have been proposed to detect uneven wheel wear based on the vibrations generated when a railway vehicle is traveling.

[0003] For example, Patent Document 1 discloses a technology that measures vibrations of railway structures, extracts vibration components in a frequency range that are susceptible to the effects of uneven wheel wear, and determines that "uneven wear exists" if the vibration components exceed a predetermined threshold.

[0004] Patent Document 2 discloses a technology that measures vibrations on the vehicle floor, filters and envelopes the data, and determines that there is "uneven wear" if the processed data exceeds a predetermined threshold value for a predetermined period of time. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-237348 [Patent Document 2] Japanese Patent Publication No. 2020-8364 Summary of the Invention [Problem to be solved by the invention]

[0006] Uneven wheel wear can occur before the grinding cycle. Therefore, it is desirable to improve the accuracy of determining uneven wheel wear and to properly formulate a wheel grinding plan.

[0007] However, the technology disclosed in Patent Document 1 detects vibrations of track structures to determine uneven wheel wear, which includes the influence of external disturbances. Therefore, there is room for improvement in the accuracy of determining uneven wheel wear in the technology disclosed in Patent Document 1.

[0008] Uneven wheel wear tends to show peaks in the vibration of the wheel rotation period and its higher-order components. The technology disclosed in Patent Document 2 does not use the vibration of the wheel rotation period itself, but acquires vibration data at regular intervals and uses the peak values ​​to determine whether or not there is uneven wheel wear. The vibration data acquired at regular intervals is affected by the vehicle's speed. Therefore, the technology disclosed in Patent Document 2 has variability in the processed data, leaving room for improvement in the accuracy of determining uneven wheel wear.

[0009] Therefore, there is a need for technology that improves the accuracy of determining uneven wheel wear and appropriately formulates wheel grinding plans. [Means for solving the problem]

[0010] A railway vehicle wheel uneven wear detection system developed to solve the above-mentioned problems is configured to have: (1) a sensor unit that measures the vibration of the railway vehicle; a sampling unit that samples measurement values ​​from the sensor unit at regular distances; a calculation unit that extracts a predetermined wheel rotation frequency component caused by uneven wheel wear from the measurement values ​​sampled at regular distances by the sampling unit and calculates a vibration processing value based on the peak value detected from the extracted wheel rotation frequency component; a memory unit that accumulates and stores the vibration processing values ​​calculated by the calculation unit while the railway vehicle travels along a specified running section; a statistical processing unit that calculates a statistical processing value based on the vibration processing values ​​for the specified running section accumulated and stored in the memory unit; and a judgment unit that judges the state of uneven wheel wear by comparing the statistical processing value calculated by the statistical processing unit with a threshold value.

[0011] In the railway vehicle wheel uneven wear detection system configured as described above, since processed vibration values ​​due to uneven wheel wear tend to occur continuously in accordance with the wheel rotation cycle, the sampling unit samples vehicle vibration measurement values ​​from the sensor unit at regular intervals. The calculation unit extracts a predetermined wheel rotation frequency component caused by uneven wheel wear from the sampled measurement values ​​and calculates a processed vibration value based on a peak value detected from the extracted wheel rotation frequency component. The memory unit accumulates and stores the processed vibration values ​​for a predetermined running section. This allows the railway vehicle wheel uneven wear detection system to acquire the processed vibration value based on wheel rotation separately from processed vibration values ​​caused by bogie component failures, rail joints, etc. The statistical processing unit calculates a statistical processing value based on the processed vibration values ​​for the predetermined running section. The determination unit compares the calculated statistical processing value with a threshold value to determine the state of uneven wheel wear. Therefore, the railway vehicle wheel uneven wear detection system configured as described above compares the statistical processing value with a threshold value to determine the state of uneven wheel wear, which is expected to improve determination accuracy and enable appropriate formulation of wheel grinding plans.

[0012] (2) In the railway vehicle wheel uneven wear detection system described in (1), the threshold value is set based on the correlation between the statistically processed value based on the measured vibration data and the uneven wear data based on the measured uneven wear data, which are obtained by actually measuring vibration data and uneven wear data when the railway vehicle repeatedly travels along the specified running section, and the uneven wear data is data obtained by actually measuring the wheel circumference shape, and the uneven wear index value is the amount of change between the maximum uneven wear amount position where the amount of uneven wear is maximum in the uneven wear data and the minimum uneven wear amount position where the amount of uneven wear is minimum.

[0013] The railway vehicle wheel uneven wear detection system configured as described above uses the amount of change between the maximum and minimum positions of uneven wear in the measured uneven wear data as an uneven wear index value, and sets a threshold value based on the correlation between the uneven wear index value and the statistically processed value, thereby estimating the degree of progression of uneven wheel wear.

[0014] (3) In the railway vehicle wheel uneven wear detection system described in (2), it is preferable that, when there is a difference between the wheel center of gravity and the axle center of the wheel for which the uneven wear data is actually measured, the uneven wear index value is corrected so that the wheel center of gravity is set on an extension line connecting the axle center and the position where the uneven wear amount is maximum.

[0015] The railway vehicle wheel uneven wear detection system configured as described above corrects the difference between the wheel center of gravity and the axle center so that the wheel center of gravity is set on the extension line connecting the axle center and the point where the amount of uneven wear is maximum, thereby improving the correlation between the statistical processing value and the uneven wear index value and enabling the degree of progression of uneven wheel wear to be estimated accurately.

[0016] (4) In the railway vehicle wheel uneven wear detection system described in (2) or (3), it is preferable to perform curve fitting using an exponential distribution on the correlation between the uneven wear index value and the statistically processed value.

[0017] The railway vehicle wheel uneven wear detection system configured as described above is expected to improve the accuracy of determining the degree of progression of uneven wheel wear by performing curve fitting using an exponential distribution on the correlation between the uneven wear index value and the statistically processed value.

[0018] (5) In the railway vehicle wheel uneven wear detection system described in any one of (2) to (4), it is preferable that the threshold value is set in stages based on the degree of progression of the uneven wheel wear.

[0019] The railway vehicle wheel uneven wear detection system configured as described above can easily determine the degree of progression of uneven wheel wear by setting threshold values ​​in stages based on the degree of progression of uneven wheel wear.

[0020] A method, an apparatus, a program for realizing the functions of the above system, and a computer-readable storage medium storing the program are also novel and useful. [Effects of the Invention]

[0021] Therefore, according to the railway vehicle wheel uneven wear detection system configured as described above, the judgment accuracy is improved in the railway vehicle wheel uneven wear detection system that detects uneven wheel wear based on vibrations while the vehicle is in motion, and wheel grinding plans can be appropriately formulated. [Brief explanation of the drawings]

[0022] [Figure 1] FIG. 1 is a schematic diagram of a wheel uneven wear monitoring system. [Figure 2] FIG. 10 is a diagram illustrating an outline of a procedure for determining uneven wheel wear. [Figure 3] FIG. 10 is a diagram illustrating an example of vibration processing data. [Figure 4] FIG. 10 is a diagram illustrating an example of uneven wear data. [Figure 5] FIG. 10 is a diagram illustrating the variation in statistically processed values ​​depending on the type of uneven wheel wear. [Figure 6] FIG. 4 is a diagram illustrating calculation of an uneven wear index value. [Figure 7]FIG. 4 is a diagram illustrating calculation of an uneven wear index value. [Figure 8] 10A and 10B are diagrams showing examples of the correlation between uneven wear index values ​​and statistically processed values ​​when curve fitting is not performed and when curve fitting is performed. [Figure 9] FIG. 10 is a diagram illustrating an example of a threshold value. [Figure 10] FIG. 10 is a diagram illustrating an example of a notification screen. DETAILED DESCRIPTION OF THE INVENTION

[0023] The present specification will now be described with reference to the accompanying drawings, which illustrate embodiments of a railway vehicle wheel wear detection system and a railway vehicle wheel wear detection method. The present specification will discuss a railway vehicle wheel wear detection system that determines the state of wheel wear based on vibrations measured while the railway vehicle is in motion.

[0024] (Schematic configuration of a railway vehicle wheel uneven wear detection system) The railway vehicle wheel uneven wear detection system 1 (hereinafter abbreviated as "system 1") shown in Figure 1 is a system that detects uneven wheel wear of a railway vehicle traveling on a specified traveling section based on vibrations generated by the railway vehicle.

[0025] The predetermined running section is, for example, a section with a running distance D of 100 km or more. The reason why the system 1 detects uneven wheel wear for each railway vehicle running on the predetermined running section is to suppress variations in the vibration processing value due to the influence of external disturbances by statistically processing the vibration processing value. In this embodiment, the detection target is a train 4 running on the running section from Station A (starting station) to Station B (end station). The running section from Station A to Station B is an example of a "predetermined running section."

[0026] The train 4 is made up of, for example, two railcars 41A and 41B. The railcar 41A has bogies 43A and 43B arranged at the front and rear of a carbody 42A. The bogie 43A has four wheels 45A. The wheels 45A are coupled to both ends of axles 44A arranged at the front and rear of the bogie 43A in the left-right direction of the vehicle (sleeper direction), and are rotatably held on the bogie 43A via the axles 44A. The railcar 41B is provided with bogies 43C and 43D, similar to the railcar 41A. The bogies 43B, 43C, and 43D are provided with wheels 45B, 45C, and 45D and axles 44B, 44C, and 44D, similar to the bogie 43A. The railcars 41A and 41B run when the wheels 45A, 45B, 45C, and 45D roll on the rails 3. In the following description, the wheels 45A, 45B, 45C, and 45D may be collectively referred to as the wheels 45 unless a distinction is particularly required.

[0027] In order to suppress vibration, the wheel 45 has a wheel circumferential shape of the tread that contacts the rail 3, which is formed into a perfect circle. The wheel circumferential shape of the wheel 45 wears due to braking, frictional heat during running, vehicle load, etc., and the degree of roundness changes. In other words, uneven wheel wear occurs on the wheel 45. Uneven wheel wear increases vehicle vibration, and may lead to a deterioration in ride comfort and damage to bogie parts. Therefore, wheels 45 that have experienced uneven wheel wear are reshaped to return their wheel circumferential shape to a perfect circle.

[0028] The system 1 of this embodiment includes a ground-side device 10 and a vehicle-side device 20. The ground-side device 10 and the vehicle-side device 20 can be connected to each other so as to be able to communicate with each other.

[0029] The vehicle-side device 20 is provided on the train 4. The vehicle-side device 20 includes acceleration sensors 21A, 21B, 21C, and 21D, vibration processing devices 22A and 22B, a management device 23, and a vehicle-side communication unit 24.

[0030] Acceleration sensors 21A, 21B, 21C, and 21D are attached to bogies 43A, 43B, 43C, and 43D, respectively, and measure vibrations of railway vehicles 41A and 41B that occur in bogies 43A, 43B, 43C, and 43D, respectively. Acceleration sensors 21A, 21B, 21C, and 21D are examples of "sensor units."

[0031] The vibration processing devices 22A and 22B have a function of detecting vibrations caused by uneven wheel wear. The vibration processing device 22A is mounted on the railway vehicle 41A and includes a sampling unit 221A and a calculation unit 222A. The sampling unit 221A has a function of sampling vibrations measured by the acceleration sensors 21A and 21B under predetermined conditions. The calculation unit 222A has a function of calculating a vibration processing value corresponding to the wheel rotation period based on the vibrations sampled by the sampling unit 221A. The vibration processing device 22B is mounted on the railway vehicle 41B and, like the vibration processing device 22A, includes a sampling unit 221B and a calculation unit 222B. The vibration processing devices 22A and 22B are connected to the management device 23.

[0032] The management device 23 is a device that manages the equipment mounted on the railway vehicle 41. The management device 23 has a function of accumulating and storing vibration processing values ​​calculated by the calculation units 222A, 222B of the vibration processing devices 22A, 22B while the railway vehicle 41 travels on a predetermined traveling section. The management device 23 is an example of a "storage unit."

[0033] The vehicle-side communication unit 24 is a device that controls communication with external devices such as the ground-side device 10. The management device 23 can transmit the vibration processing value to the ground-side device 10 using the vehicle-side communication unit 24.

[0034] The number of railcars constituting train 4 is not limited to that in this embodiment, and may be one, or three or more. In the following description, railcars 41A, 41B, bogies 43A, 43B, 43C, 43D, acceleration sensors 21A, 21B, 21C, 21D, vibration processing devices 22A, 22B, sampling units 221A, 221B, and calculation units 222A, 222B may be collectively referred to as railcar 41, bogie 43, acceleration sensor 21, vibration processing device 22, sampling unit 221, and calculation unit 222 unless there is a particular need to distinguish between them.

[0035] The ground-side device 10 includes a ground-side communication unit 11 that controls communication with external devices such as the vehicle-side device 20, and an analysis device 12 that analyzes data. The analysis device 12 includes a vibration processing data acquisition unit 121 that acquires vibration processing values ​​from the vehicle-side device 20, a statistical processing unit 122 that statistically processes the vibration processing values ​​to calculate statistical processing values, a determination unit 123 that determines the state of uneven wheel wear based on the statistical processing values ​​and a threshold value, and a notification unit 124 that notifies the determination result. The ground-side device 10 can be installed in any location.

[0036] (Explanation of the operation of the railway vehicle wheel uneven wear detection system) The operation of the system 1 will be specifically described with reference to Fig. 2. The system 1 can determine the state of uneven wheel wear for each bogie 43 of the train 4. Here, an example will be described in which the uneven wheel wear is determined for the wheel 45A of the bogie 43A.

[0037] The vehicle-side device 20 uses the acceleration sensor 21A to measure vibrations of the railway vehicle 41 that occur in the bogie 43A at regular time intervals (for example, every second) (S11).

[0038] The sampling unit 221A of the vibration processing device 22A samples the measurement value (vibration data) from the acceleration sensor 21A every time the railway vehicle 41 travels a certain distance (for example, 4 m) (S12). S12 is an example of a "sampling step".

[0039] The process shown in S12 is executed in the same manner as in, for example, Japanese Patent No. 4388594. Briefly, the sampling unit 221A initializes the travel distance [m] and travel time [s], for example, at the start point of a predetermined travel section (Station A in this embodiment). Based on the speed data received from the management device 23, the sampling unit 221A adds the product of the travel time and the travel speed [m / s] to the travel distance, and then initializes the travel time [s]. If the added travel distance is less than a predetermined equidistant sampling period [m] (e.g., 4 m), the sampling unit 221A further calculates and adds the travel distance based on the travel time and the travel speed, as described above, and then initializes the travel time. Thereafter, the sampling unit 221A repeats the procedure of comparing the added travel distance with the equidistant sampling period. If the added travel distance is equal to or greater than the equidistant sampling period, the sampling unit 221A collects vibration data from the acceleration sensor 21A, initializes the travel distance, and repeats the sampling operation.

[0040] The calculation unit 222A executes bandpass filtering to extract wheel rotation frequency components caused by uneven wheel wear from the vibration data sampled by the sampling unit 221A (S13). Next, the calculation unit 222A calculates a vibration processing value from the extracted wheel rotation frequency components (S14).

[0041] The processing shown in S13 to S14 is performed in the same manner as, for example, Japanese Patent No. 4388594. Simply put, the spatial frequency [1 / m] of vibrations generated by different causes of vibrations varies, and uneven wheel wear tends to generate vibrations according to the wheel rotation period. Therefore, the calculation unit 222A processes the equidistant sampling data collected from the sampling unit 221A using a band-pass filter (BPF) to divide the spatial frequency components into regions related to the wheel rotation frequency.

[0042] The calculation unit 222A calculates the absolute value of the vibration peak for each extracted wheel rotation frequency component, and calculates the average of the absolute values ​​(average of vibration peak absolute values) for each wheel rotation frequency component as a vibration processing value. This vibration processing value is calculated for each traveling speed range in 5 km / h increments, such as 100-105 km / h, 105-110 km / h, and 110-115 km / h.

[0043] The management device 23 accumulates and stores the vibration processing values ​​calculated by the calculation unit 222A (S15). S15 is an example of a "storage step". For example, the management device 23 stores vibration processing data for each piece of bogie identification information. The vibration processing data can be stored by associating the vibration processing value with the distance from the start point of a specified running section, as shown in FIG. 3, for example. Note that FIG. 3 conceptually illustrates the vibration processing values, and the number of vibration processing values ​​differs from the actual number.

[0044] The vehicle-side device 20 repeats the above procedure until the vehicle reaches the end point of the predetermined travel section. As a result, the management device 23 accumulates and stores vibration processing values ​​for the predetermined travel section (in this embodiment, the travel section from station A to station B).

[0045] 2, the vehicle-side device 20 transmits the vibration processing value stored in the management device 23 to the ground-side device 10 (S16). The ground-side device 10 receives the vibration processing value from the vehicle-side device 20 using the ground-side communication unit 11 (S31).

[0046] For example, the vibration processing data acquisition unit 121 of the analysis device 12 requests the vehicle-side device 20 to transmit vibration processing data using the ground-side communication unit 11. In response to the request, the management device 23 of the vehicle-side device 20 transmits the vibration processing data stored in the management device 23 to the ground-side device 10 using the vehicle-side communication unit 24. Using the ground-side communication unit 11, the vibration processing data acquisition unit 121 receives vibration processing values ​​calculated by the vibration processing devices 22A and 22B based on the vibration data of the acceleration sensors 21A, 21B, 21C, and 21D while the train 4 is traveling in a predetermined traveling section (in this embodiment, the traveling section from station A to station B).

[0047] The statistical processing unit 122 executes statistical processing to calculate a statistical processing value based on the vibration processing data received from the vehicle-side device 20 (S32). S32 is an example of a "statistical step." For example, when determining the state of uneven wheel wear for the wheels 45A of the bogie 43A, the statistical processing unit 122 reads out the vibration processing data associated with the bogie identification number of the bogie 43A and calculates a statistical processing value of the vibration processing value included in the vibration processing data. This allows the vibration processing value of the vibration generated due to the rotation of the wheels 45A that constitute the bogie 43A to be evaluated. An example of a "statistical processing value" is the average value of the vibration processing values.

[0048] To explain in detail how to evaluate the vibration tolerance value, as shown in Figure 3, vehicle vibrations are caused by a variety of factors, resulting in variations in the magnitude of the vibration tolerance value. The vibration tolerance value due to uneven wheel wear may be smaller than the vibration tolerance value due to the influence of external disturbances, etc. When the maximum vibration tolerance value shown in P1 in Figure 3 is extracted from the vibration tolerance data, there is a possibility that the vibration tolerance value is due to factors other than uneven wheel wear, such as the vibration tolerance value due to the influence of external disturbances.

[0049] On the other hand, vibrations due to uneven wheel wear tend to occur continuously according to the wheel rotation cycle. Therefore, as shown in P2 of Fig. 3, the statistical processing unit 122 evaluates the processed vibration value by statistically processing, for example, averaging, the processed vibration values ​​for a predetermined driving section. By having the statistical processing unit 122 evaluate the processed vibration value using the average value, the influence of variations in the processed vibration value due to the causes of vibration is suppressed, and it becomes more likely that the processed vibration value due to uneven wheel wear can be evaluated separately from the processed vibration values ​​due to other causes.

[0050] In this embodiment, the vibration processing data includes a large amount of vibration processing values ​​calculated based on vibration data sampled at regular intervals while the train 4 travels from station A to station B. The statistical processing unit 122 calculates statistically processed values ​​of this large amount of vibration processing values. The statistically processed values ​​of the large amount of vibration processing values ​​are close to the vibration processing values ​​based on uneven wheel wear that occurs continuously in accordance with the wheel rotation cycle while traveling, and variations in the vibration processing values ​​due to the influence of disturbances are suppressed.

[0051] 2, the determination unit 123 executes a determination process to determine the state of uneven wheel wear by comparing the statistical processing value calculated by the statistical processing unit 122 with a threshold value TH (S33). S33 is an example of a "determination step".

[0052] The procedure for setting the threshold value TH will now be described in detail. The inventors use a computer to calculate an uneven wear index value for evaluating uneven wheel wear based on uneven wear data actually measured on the wheels of a railway vehicle.

[0053] The uneven wear data is data obtained by actually measuring the circumferential shape of the wheel, and is measured, for example, by jacking up the axle and rotating the wheel, and using a measuring instrument such as a dial gauge from the ground.

[0054] Figure 4(A) is an example of uneven wear data for a wheel that has experienced uneven wear. SF01 is a perfect circle circumscribing the maximum radius of the worn wheel circumference. SF10 indicates the worn wheel circumference. F indicates the amount of wear of the tread in the wheel radial direction (hereinafter referred to as "runout amount"). Note that the actual wheel uneven wear is less than 1 mm, and Figure 4(A) exaggerates the wheel uneven wear. Figure 4(B) is a diagram showing the uneven wear data shown in Figure 4(A) expanded in the circumferential direction. In the figure, IP11 to IP17 indicate inflection points of the wheel circumference. In the figure, dotted lines C11 to C16 indicate the amount of change between two adjacent inflection points. In the figure, E1 to E6 indicate the angle difference between two adjacent inflection points.

[0055] The inventor analyzed the shape of the uneven wheel wear based on each uneven wear data. As a result, it was found that there are two types of uneven wheel wear. The first type, as shown in Figure 5(A), is a type in which the uneven wheel wear progresses steeply in one part of the wheel. The second type, as shown in Figure 5(B), is a type in which the uneven wheel wear progresses gradually over a wide area of ​​the wheel.

[0056] The inventors used a computer to measure the vibration of the vehicle at regular intervals while repeatedly traveling along a predetermined travel section (in this embodiment, the travel section from Station A to Station B), calculate vibration processing values, and evaluate statistically processed values ​​calculated based on the calculated vibration processing values. The inventors used a computer to calculate the maximum amount of runout (hereinafter referred to as "maximum runout amount") for each piece of uneven wear data, and performed a regression analysis between the maximum amount of runout and the statistically processed value to derive a correlation. Figure 5(C) shows an example of the correlation between the maximum amount of runout and the statistically processed value. P11 in Figure 5(C) is the statistically processed value for a wheel that experienced the type of uneven wheel wear shown in Figure 5(A). P12 is the statistically processed value for a wheel that experienced the type of uneven wheel wear shown in Figure 5(B).

[0057] The wheels corresponding to statistically processed values ​​P11 and P12 had approximately the same maximum runout amounts F1 and F2, but there was a difference in the statistically processed values. The statistically processed value P12 of the wheel that experienced the type of uneven wheel wear shown in Figure 5(B) was smaller than the statistically processed value P11 of the wheel that experienced the type of uneven wheel wear shown in Figure 5(A). Therefore, if a threshold value for determining uneven wheel wear is set based on the correlation between the maximum runout amount and the statistically processed value, the determination accuracy may deteriorate.

[0058] On the other hand, the type shown in Figure 5(B) has a wheel circumferential shape that is gentler than the type shown in Figure 5(A), so it is thought that the statistically processed value P12 is smaller than the statistically processed value P11. Therefore, uneven wheel wear can be evaluated not by the maximum amount of runout but by the amount of change in the wheel circumferential shape. Therefore, in this embodiment, a threshold value for determining uneven wheel wear is set based on the correlation between the amount of change in the wheel circumferential shape and the statistically processed value.

[0059] For example, Figure 6(A) shows an example of uneven wear data for wheel WH10, which has experienced uneven wheel wear. In the figure, O1 is the axle center. In the figure, O2 is the wheel center of gravity. In the figure, D1 indicates the distance between the axle center O1 and the wheel center of gravity O2. In the figure, φ indicates the angular difference in the wheel circumferential direction between the axle center O1 and the wheel center of gravity O2. In the figure, SF01 is a perfect circle. In the figure, SF10 is the wheel circumferential shape. In the figure, P21 is the maximum runout amount position where the amount of runout is maximum. The maximum runout amount position P21 is an example of a "maximum uneven wear amount position."

[0060] When the maximum deflection amount position P21 of the wheel WH10 comes into contact with the rail, the wheel mass acts on the maximum deflection amount position P21, causing a peak value of vibration due to uneven wheel wear. Therefore, an index of how much the wheel mass contributes to the vibration processing value when the maximum deflection amount position P21 comes into contact with the rail is calculated as an uneven wear index value.

[0061] When the wheel WH10's maximum deflection point P21 contacts the rail 3 and the wheel center of gravity O2 is on the extension line L1 connecting the axle center O1 and the maximum deflection point P21, the axle position is most significantly shifted toward the rail 3 (downward), and the wheel mass contributes significantly to the vibration processing value. However, uneven wear data is measured by lifting and rotating the wheel via the axle and measuring it from the ground with a measuring instrument. Therefore, the wheel center of gravity O2 in the uneven wear data may be slightly shifted by less than 1 mm from the axle center O1. This difference differs from the actual wheel condition and may undermine the reliability of the uneven wheel wear assessment.

[0062] Therefore, in order to allow the wheel mass to contribute to the maximum deflection amount position P21, the wheel center of gravity O2 is set on the extension line L1 using, for example, the phase coefficient relational expression shown in FIG. 6(B). The wheel center of gravity set on the extension line L1 is referred to as the "wheel center of gravity O2x." In the phase coefficient relational expression, if the wheel center of gravity O2 is on the extension line L1 extending from the maximum deflection amount position P21 toward the axle center O1 and the angular difference φ between the wheel center of gravity O2 and the axle center O1 is 0°, i.e., 0 (rad), the phase coefficient is set to "1." In the phase coefficient relational expression, if the wheel center of gravity O2 is deviated from the extension line L1 and the angular difference φ between the wheel center of gravity O2 and the axle center O1 is 90°, i.e., π / 2 (rad), the phase coefficient is set to "0." In the relational expression of the phase coefficient, when the wheel center of gravity O2 is on an extension line L1 extending from the maximum deflection amount position P21 to the opposite side of the axle center O1, and the angular difference φ between the wheel center of gravity O2 and the axle center O1 is 180°, that is, π (rad), the phase coefficient is set to "-1".

[0063] In Figures 7(A) and 7(B), the position adjacent to the maximum runout amount position P21 where the runout amount is smallest is defined as the minimum runout amount position P22. The minimum runout amount position P22 is an example of a "minimum uneven wear amount position." When calculating the amount of change [mm / rad] between the maximum runout amount position P21 and the minimum runout amount position P22 without considering the difference between the axle center O1 and the wheel center of gravity O2, the amount of change can be calculated using Equation 1. As shown in Figure 7(A), R represents the distance between the axle center O1 and the minimum runout amount position P22. β represents the distance between the axle center O1 and the maximum runout amount position P21. θ represents the angle formed by a first chord L21 connecting the axle center O1 and the maximum runout amount position P21 and a second chord L22 connecting the axle center O1 and the minimum runout amount position P22.

[0064]

number

[0065] Using Equation 1, the amount of change per unit angle [mm / rad] between the maximum deflection amount position P21 and the minimum deflection amount position P22 is calculated based on the distance β between the axle center O1 and the maximum deflection amount position P21 and the distance R between the axle center O1 and the minimum deflection amount position P22.

[0066] On the other hand, when calculating the amount of change [mm / rad] between the maximum deflection amount position P21 and the minimum deflection amount position P22, taking into consideration the difference between the axle center O1 and the wheel center of gravity O2, the amount of change can be calculated using Equation 2. α indicates the distance between the wheel center of gravity O2x and the minimum deflection amount position P22, as shown in FIG. 7(B). D G ·cohe indicates the distance between the wheel center of gravity O2x and the axle center O1. β+D G ·cohe indicates the distance between the wheel center of gravity O2x and the maximum deflection position P21. G indicates the angle formed by the third chord L31 connecting the wheel center of gravity O2x and the maximum deflection amount position P21 and the fourth chord L32 connecting the wheel center of gravity O2x and the minimum deflection amount position P22.

[0067]

number

[0068] Using Equation 2, the amount of change per unit angle [mm / rad] between the maximum deflection amount position P21 and the minimum deflection amount position P22 is calculated, reflecting the distance D1 between the wheel center of gravity O2 and the axle center O1 and the angle difference φ in the wheel circumferential direction.

[0069] In order to set a threshold indicating the degree of progression of uneven wheel wear based on the measured statistically processed value during actual operation, it is easier to manage if the statistically processed value and the uneven wear index value have a linear positive correlation. Therefore, the amount of change calculated using Equation 2 is evaluated as an uneven wear index value [mm / rad], and a regression analysis is performed on the uneven wear index value and the statistically processed value calculated in the driving test to derive a linear correlation equation. An example of the derived correlation is shown in Figure 8(A). Compared to the correlation between the statistically processed value and the maximum runout amount shown in Figure 5(C), Figure 8(A) shows a smaller variation in the statistically processed value. Therefore, the threshold set based on the correlation in Figure 8(A) can provide higher accuracy in determining uneven wheel wear than the threshold set based on the correlation in Figure 5(C). The coefficient of determination R of the correlation shown in Figure 8(A) is 2 is 0.88, which means there is room for improvement in the correlation.

[0070] Therefore, the uneven wear index value [mm / rad] calculated using Equation 2 was curve-fitted using an exponential distribution. That is, the curve fitting was performed using Equation 3. e is the base of the natural logarithm. λ is a parameter of the exponential distribution. x is the uneven wear index value.

[0071]

number

[0072] In the curve fitting, λ is adjusted. An example of the correlation between the curve-fitted uneven wear index value and the statistically processed value is shown in Figure 8(B). The coefficient of determination R of the correlation shown in Figure 8(B) 2 was 0.95, which is a higher correlation between the uneven wear index value and the statistically processed value than the correlation shown in Figure 8(A). Therefore, the threshold value set based on the correlation in Figure 8(B) can improve the accuracy of determining uneven wheel wear compared to the threshold value set based on the correlation in Figure 8(A).

[0073] Figure 9 shows an example of thresholds set in stages based on the correlation shown in Figure 8(B). For example, in this embodiment, the range of statistical processing values ​​from 0 to Q1 is Level 1, indicating no signs of uneven wear progression. The range of statistical processing values ​​from Q1 to Q2 is Level 2, indicating early signs of uneven wear progression. The range of statistical processing values ​​from Q2 to Q3 is Level 3, indicating signs of uneven wear progression and requiring grinding at the next timing. The range of statistical processing values ​​greater than Q3 is Level 4, indicating the grinding limit where grinding is necessary and immediate grinding is required.

[0074] The analysis device 12 stores the statistically processed values ​​Q1, Q2, and Q3 in the determination unit 123 as threshold values ​​TH for determining levels 1 to 4. In this embodiment, the threshold values ​​TH are set for each of the carriages 43.

[0075] The determination process will be described in detail. For example, the determination unit 123 of the analysis device 12 reads out the threshold value TH corresponding to the bogie identification information of the bogie 43A. The determination unit 123 compares the read-out threshold value TH with the statistical processing value of the bogie 43A calculated by the statistical processing unit 122, and determines the state of uneven wheel wear.

[0076] For example, if the statistical processing value is equal to or less than Q1, the determination unit 123 determines that the uneven wheel wear state of the four wheels 45A that constitute the bogie 43A is level 1. For example, if the statistical processing value is greater than Q1 and equal to or less than Q2, the determination unit 123 determines that the uneven wheel wear state of the four wheels 45A that constitute the bogie 43A is level 2. For example, if the statistical processing value is greater than Q2 and equal to or less than Q3, the determination unit 123 determines that the uneven wheel wear state of the four wheels 45A that constitute the bogie 43A is level 3. For example, if the statistical processing value is greater than Q3, the determination unit 123 determines that the uneven wheel wear state of the four wheels 45A that constitute the bogie 43A is level 4.

[0077] The determination unit 123 stores the determination result in association with the train number of the train 4 and the bogie identification information of the bogie 43A.

[0078] The analysis device 12 deletes the vibration processed data of the carriage 43A from the vibration processed data acquisition unit 121. This completes the determination of the state of uneven wheel wear for the four wheels 45A that constitute the carriage 43A.

[0079] The analysis device 12 also determines the state of uneven wheel wear for the wheels of the other bogies 43B, 43C, and 43D in the same manner as above, and stores the determination results.

[0080] 2, the analysis device 12 of the ground device 10 uses the notification unit 124 to notify the determination result of the determination unit 123 (S34). For example, if the notification unit 124 has a function to display information, it displays the determination result of the determination unit 123 on a notification screen 110 shown in FIG.

[0081] The notification screen 110 includes, for example, a determination result display area 111, a guidance area 112, and a close button 113. The determination result display area 111 includes, for example, a train number display section 111a and a determination result list 111b. The train number display section 111a displays the train number of the train 4. The determination result list 111b displays the determined level for each bogie identification information of the bogies 43A, 43B, 43C, and 43D that make up the train 4. The guidance area 112 includes, for example, a level list 112b. The level list 112b displays, for each level, the degree of progression of uneven wheel wear and the time for grinding.

[0082] For example, when an operator sees that the bogie identification information "XXXA01" for bogie 43A is displayed as "Level 2" in the judgment result list, the operator can confirm that there are early signs of uneven wheel wear on wheel 45A of bogie 43A, but that there is no need to perform grinding.

[0083] Furthermore, for example, when an operator sees that "Level 1" is displayed for the bogie identification information "XXXA02" of bogie 43B in the judgment result list, the operator can confirm that there is no uneven wheel wear on wheel 45B of bogie 43B and that there is no need to perform grinding.

[0084] Furthermore, for example, when an operator sees that "Level 3" is displayed for the bogie identification information "XXXB01" of bogie 43C in the list of judgment results, the operator can confirm that uneven wheel wear is progressing on wheel 45C of bogie 43C and that grinding is required at the next timing.

[0085] Furthermore, for example, when an operator sees that "Level 4" is displayed for the bogie identification information "XXXB02" of bogie 43D in the list of judgment results, the operator can confirm that the uneven wheel wear on wheel 45D of bogie 43D has progressed to the grinding limit and that immediate grinding is required.

[0086] In this way, the system 1 can determine the degree of progression of uneven wheel wear for each wheel 45 of the bogies 43A, 43B, 43C, and 43D, and appropriately formulate a wheel reconditioning plan.

[0087] For example, in the past, the wheel grinding cycle was set to the travel distance at which uneven wheel wear was likely to begin to occur. In contrast, the system 1 of this embodiment can detect wheel 45D of bogie 43D, which is experiencing uneven wheel wear that requires grinding, even before the travel distance reaches the grinding cycle, and notify the operator that wheel 45D needs to be grinded. Furthermore, the system 1 can extend the grinding cycle for wheel 45A of bogie 43A, where uneven wheel wear is progressing slowly, and for wheel 45B of bogie 43B, where there are no signs of uneven wheel wear. In other words, the system 1 can appropriately formulate a wheel grinding plan depending on the degree of progression of uneven wheel wear of wheel 45.

[0088] As described above, in the system 1 of this embodiment, since the vibration processing value due to uneven wheel wear tends to occur continuously according to the wheel rotation period, the sampling unit 221A samples the vehicle vibration measurement value from the acceleration sensor 21A at regular intervals. The calculation unit 222A extracts a predetermined wheel rotation frequency component due to uneven wheel wear from the sampled measurement value and calculates the vibration processing value based on the peak value detected from the extracted wheel rotation frequency component. This allows the system 1 to acquire the vibration processing value due to wheel rotation separately from vibration processing values ​​generated by bogie component failures, rail joints, etc. The management device 23 accumulates and stores the vibration processing values ​​for a predetermined traveling section. The statistical processing unit 122 calculates a statistical processing value based on the vibration processing value for the predetermined traveling section. The determination unit 123 compares the calculated statistical processing value with a threshold to determine the state of uneven wheel wear. Therefore, the system 1 of this embodiment compares the statistical processing value with a threshold to determine the state of uneven wheel wear, which is expected to improve the determination accuracy and appropriately formulate a wheel 45 grinding plan.

[0089] The present invention is not limited to the above embodiment and can be applied in various ways. For example, the predetermined running section is not limited to the line section from the starting station to the terminal station, but may be a section between any points. Also, for example, the acceleration sensor 21 may be attached to the car body 42 or the axle box instead of the bogie 43.

[0090] For example, the configuration of the system 1 may differ from that of the above embodiment. For example, the vehicle-side device 20 may perform the statistical processing and determination processing executed by the ground-side device 10 based on the vibration processing value calculated by the vehicle-side device 20 itself. Furthermore, for example, the system 1 may install one or more relay devices connectable to the vehicle-side device 20 and a server at any location within a predetermined travel section, and aggregate or integrate vibration processing data in the server via the relay devices. In this case, the ground-side device 10 may access the server in S31 of FIG. 2 to obtain the vibration processing data and execute the processes of S32 to S34. In this case, the vehicle-side device 20 does not need to store a large number of vibration processing values ​​for the predetermined travel section, thereby reducing the memory load.

[0091] For example, the statistical processing performed by the statistical processing unit 122 may be processing different from the processing for calculating the average value of the vibration processing values.

[0092] For example, the determination process in S33 and the notification process in S34 in FIG. 2 may be performed collectively on data for a plurality of predetermined travel sections.

[0093] For example, the vibration processing device 22 is not limited to being installed in one unit per railway vehicle 41. For example, depending on the processing capacity, one vibration processing device 22 may be installed in one bogie 43 or one in one train.

[0094] For example, the management device 23 and the vehicle-side communication unit 24 do not have to be one per train, but may be one per railway car 41, for example.

[0095] For example, the threshold value TH may be uniform regardless of the type of bogie 43, or may be varied depending on the type of vehicle or type of bogie.

[0096] The predetermined travel distance may be less than 100 km. However, in the case of a travel distance of 100 km or more, the vehicle-side device 20 calculates a large number of vibration processing values, and the ground-side device 10 statistically processes the large number of vibration processing values, so that the statistically processed values ​​approach the vibration processing values ​​corresponding to the state of uneven wheel wear. By determining the state of uneven wheel wear based on the statistically processed values, the system 1 is expected to improve its determination accuracy.

[0097] The threshold value TH does not have to be set based on the correlation between the uneven wear index value calculated from the actual measurement data and the statistically processed value. However, the system 1 can estimate the degree of progression of uneven wheel wear by using the amount of change between the maximum runout amount position P21 and the minimum runout amount position P22 of the actually measured uneven wear data as the uneven wear index value and setting the threshold value TH based on the correlation between the uneven wear index value and the statistically processed value.

[0098] The difference between the axle center O1 and the wheel center of gravity O2 does not need to be corrected. However, by correcting the difference between the wheel center of gravity O2 and the axle center O1 so that the wheel center of gravity O2 is set on the extension line L1 connecting the axle center O1 and the maximum runout amount position P21, the system 1 improves the correlation between the statistical processing value and the uneven wear index value, enabling the degree of progression of uneven wheel wear to be estimated with high accuracy.

[0099] The correlation between the uneven wear index value and the statistically processed value shown in Figure 8(A) does not have to be curve-fitted using an exponential distribution. However, by curve-fitting the correlation between the uneven wear index value and the statistically processed value using an exponential distribution, the system 1 is expected to improve the accuracy of determining the degree of progression of uneven wheel wear.

[0100] The threshold value TH does not have to be set in stages based on the degree of progression of uneven wheel wear, but by setting the threshold value TH in stages based on the degree of progression of uneven wheel wear, the system 1 can easily determine the degree of progression of uneven wheel wear.

[0101] 2 may be performed by a method other than displaying the notification screen 110, such as printing the determination result or outputting it as audio. The notification process shown in S34 may be omitted, and the determination result may not be automatically notified. For example, a notification instruction acceptance button may be provided on a screen provided by the analysis device 12, and the determination result may be notified at any timing in response to button operation. The configuration of the notification screen 110 shown in FIG. 10 is not limited to this embodiment.

[0102] Furthermore, in any of the flowcharts disclosed in the embodiments, any of the multiple processes can be executed in any order or in parallel, as long as no contradiction occurs in the process content.

[0103] The processes disclosed in the embodiments may be executed by hardware such as a single CPU, multiple CPUs, or ASIC, or a combination thereof. The processes disclosed in the embodiments may be realized in various ways, such as a recording medium on which a program for executing the processes is recorded, or a method. [Explanation of symbols]

[0104] 1. Uneven Wheel Wear Detection System for Railway Vehicles 21, 21A, 21B, 21C, 21D Acceleration sensor (example of sensor part) 23 Management device (an example of a storage unit) 41, 41A, 41B Railway vehicles 45,45A,45B,45C,45D wheels 122 Statistical Processing Unit 123 Judgment section 221 Sampling section 222 Arithmetic section

Claims

1. A railway vehicle wheel uneven wear detection system that detects uneven wheel wear based on vibrations generated when the railway vehicle is running, a sensor unit that measures vibrations of the railway vehicle; a sampling unit that samples measurement values ​​at regular intervals from the sensor unit; a calculation unit that extracts a predetermined wheel rotation frequency component caused by uneven wheel wear from the measurement values ​​sampled at regular distances by the sampling unit, and calculates a vibration processing value based on a peak value detected from the extracted wheel rotation frequency component; a storage unit that accumulates and stores the vibration processing value calculated by the calculation unit while the railway vehicle is traveling along a predetermined traveling section; a statistical processing unit that calculates a statistical processing value based on the vibration processing value for the predetermined traveling section that is accumulated and stored in the storage unit; a determination unit that determines a state of uneven wheel wear by comparing the statistically processed value calculated by the statistical processing unit with a threshold value; having The uneven wear detection system for railway vehicles is configured as follows.

2. 2. The railway vehicle wheel uneven wear detection system according to claim 1, the threshold value is set based on a correlation between a statistically processed value based on the measured vibration data and an uneven wear index value based on the measured uneven wear data, the statistically processed value being obtained by actually measuring vibration data and uneven wear data when the railway vehicle repeatedly travels along the predetermined travel section; The uneven wear data is data obtained by actually measuring the wheel circumferential shape, The uneven wear index value is a change amount between a maximum uneven wear amount position where the uneven wear amount is maximum and a minimum uneven wear amount position where the uneven wear amount is minimum in the uneven wear data. The uneven wear detection system for railway vehicles is configured as follows.

3. 3. The railway vehicle wheel uneven wear detection system according to claim 2, When there is a difference between the center of gravity of the wheel and the center of the axle of the wheel for which the uneven wear data is actually measured, the uneven wear index value is corrected so that the difference is set to the center of gravity of the wheel on an extension line connecting the center of the axle and the position where the uneven wear amount is maximum. The uneven wear detection system for railway vehicles is configured as follows.

4. 3. The railway vehicle wheel uneven wear detection system according to claim 2, Curve fitting is performed using an exponential distribution on the correlation between the uneven wear index value and the statistically processed value. The uneven wear detection system for railway vehicles is configured as follows.

5. The railway vehicle wheel uneven wear detection system according to any one of claims 2 to 4, The threshold value is set in stages based on the degree of progression of the uneven wheel wear. The uneven wear detection system for railway vehicles is configured as follows.

6. A method for detecting uneven wheel wear for a railway vehicle, which detects uneven wheel wear based on vibrations generated when the railway vehicle is running, comprising: a sampling step of sampling measurement values ​​at regular distances from a sensor unit that measures vibrations of the railway vehicle; a calculation step of extracting a predetermined wheel rotation frequency component caused by uneven wheel wear from the measurement values ​​sampled at regular distances in the sampling step, and calculating a vibration processing value based on a peak value detected from the extracted wheel rotation frequency component; a storage step of accumulating and storing the vibration processing value calculated in the calculation step while the railway vehicle is traveling in a predetermined traveling section; a statistical step of calculating a statistically processed value based on the vibration processing value for the predetermined traveling section accumulated and stored in the storage step; a determining step of determining a state of uneven wheel wear by comparing the statistically processed value calculated in the statistical step with a threshold value; To do The method for detecting uneven wear on a railway vehicle wheel is configured as follows.

Images