Method for identifying wheel tread damage by using shearing force sensor

Abstract

The invention discloses a method for identifying wheel tread damage by using shear sensors, and relates to the technical field of railway safety monitoring, and the method comprises the following steps: S01, a detection area is established on a steel rail, and the detection area comprises a sub-detection area formed by a plurality of mutually independent rail clamping type shear sensors; s02, when the wheel passes through the detection area, each sub-measurement area collects original shearing force data of the wheel in an effective measurement area, and the effective measurement area ranges from + / -100 mm to + / -600 mm; s03, processing the original shearing force data collected by each sub-detection area to obtain continuous vertical wheel-rail force data in the detection area; and S04, based on the obtained continuous vertical wheel-rail force data, calculating the ratio of the amplitude of each point in the effective measurement area to the average amplitude of the area. The method provided by the invention does not depend on an under-rail two-dimensional force sensor, and can realize identification of tread damage only by using a shear force sensor in the TPDS.

Description

Technical Field

[0001] This invention relates to the field of railway safety monitoring technology, specifically a method for identifying wheel tread damage using a shear force sensor. Background Technology

[0002] In the field of railway vehicle inspection, the identification of wheel tread damage mainly relies on the vehicle running quality trackside dynamic monitoring system (hereinafter referred to as TPDS). The two-dimensional force sensor and shear force sensor in this equipment can collect vertical wheel-rail force and rail shear force data. Through data processing and analysis, continuous measurement of vertical wheel-rail force can be achieved, and abnormal wheel impact load on the rail can be identified, thereby realizing the identification of wheel tread damage.

[0003] Referring to Chinese patent publication CN19037497A, published on January 29, 2024, a comprehensive early warning system and method for passenger car wheel rim defects based on TADS and TPDS are disclosed, belonging to the field of railway safety monitoring. The system includes a TADS data interface module, a TPDS data interface module, a wheel status data integration module, a wheel status information history module, a wheel fault identification module, and a fault early warning information push module. The method receives messages generated by corresponding TADS and TPDS devices, recombines the data in the messages according to each wheel; normalizes the combined sound or vibration data of each wheel; iteratively stores the normalized data of each wheel; retrieves the historical iterative data of the corresponding wheel to identify the fault type and severity; after identification, the fault early warning information push module generates a message and pushes it to each monitoring center. This invention improves the wheel rim fault identification capability and accuracy.

[0004] However, the current application of TPDS devices has the following shortcomings:

[0005] 1. The two-dimensional force sensor for measuring vertical wheel-rail force in TPDS is installed directly below the track and fixed to a dedicated steel rail. Generally, there are 10 dedicated steel rails. After replacing the rails, it is necessary to tamp and fine-tune the track in this area. The process is complex and the workload is large.

[0006] 2. This identification method is only applicable to ballasted tracks and cannot be used in ballastless track sections such as subways or high-speed railways, so its applicability is poor.

[0007] In conclusion, TPDS is not applicable to ballastless track. A good solution is needed to address the complex and labor-intensive processes involved in rail replacement, post-tamping, and track fine-tuning. Summary of the Invention

[0008] The purpose of this invention is to provide a method for identifying wheel tread damage using a shear force sensor, in order to overcome the aforementioned shortcomings in the prior art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a method for identifying wheel tread damage using a shear force sensor, comprising the following steps:

[0010] S01. A detection area is created on the rail, the detection area comprising multiple rail-mounted shear force sensors that independently form sub-detection areas;

[0011] S02. When the wheel passes through the detection area, each of the sub-measurement areas collects the original shear force data of the wheel within its effective measurement area, wherein the effective measurement area ranges from ±100mm to ±600mm.

[0012] S03. Process the raw shear force data collected in each of the sub-test areas to obtain consecutive vertical wheel-rail force data within the test area;

[0013] S04. Based on the obtained continuous vertical wheel-rail force data, calculate the ratio of the amplitude of each point in the effective measurement area to the average amplitude of the area. If the ratio exceeds the preset identification threshold, it is determined that there is an abnormal impact load caused by wheel tread damage at that location, and wheel tread damage is identified.

[0014] Preferably, multiple rail-mounted shear force sensors are installed at the web of the rail at the center of the rail gap.

[0015] Preferably, the original shear force data in step S02 is the output of the target wheel when it approaches the rail-mounted shear force sensor, which increases non-linearly to a positive peak value A. max When the target wheel is directly above the rail-mounted shear force sensor, the output rapidly reverses to reach the reverse peak value A. min .

[0016] Preferably, the processing of the multiple raw shear force data collected in step S03 includes amplifying the amplitude of the data in the effective regions to the left and right of the zero point and taking the absolute value of the data to the right of the zero point, wherein:

[0017] The range to the left of the zero point is -600mm≤n≤-100mm;

[0018] The range to the right of the zero point is 100mm≤n≤600mm.

[0019] Preferably, the processing of the multiple raw shear force data in step S03 employs the function amplification method, which specifically includes the following steps:

[0020] S31a. Use the positive peak point A to the left of the zero point in the original data curve or the specified calibration point as the reference point;

[0021] S32a. Based on the data set collected within the effective measurement area to the left of the zero point, construct at least one fitted straight line to approximate the original data curve;

[0022] S33a. Based on the parameters of the fitted straight line and the amplitude of the reference point, calculate Z0(x) for any original data point within the region. Z y Z The corresponding amplitude amplification factor m is used to obtain the amplified data point Z1(x). Z y Z *m)

[0023] S34a. Perform the same amplification process on the data within the effective measurement area to the right of the zero point as in steps S32a to S33a, and then take the absolute value of the processed data.

[0024] Preferably, in step S32a, the fitted straight line approximates at least two original data curves, namely, the fitted straight line AB and the fitted straight line BC, wherein the slope of segment AB is tanα, the slope of segment BC is tanβ, and any point Z on the fitted straight line... n The magnification factor of the ordinate replaces the magnification factor of any point Z0 in the original data, where the coordinates of point A are (-90, y...). A The coordinates of point B are (x, y). B y B Point C has coordinates (-600, 0), Z... n Let Z0 be the projection of the fitted line, where:

[0025] When -600≤n≤x B At that time, based on the coordinates of point C and the slope, we know that the point on BC conforms to a linear function: y = tanβ*x + 600tanβ;

[0026] Any point Z on BC n The ordinate is: y = tanβ*x Z +600tanβ;

[0027] Then any point Z n The magnification factor m of the ordinate is: m = y A / (tanβ*x Z +600tanβ);

[0028] Any point Z after processing n The coordinates are: (x Z y Z *m);

[0029] Then, when x B When n ≤ -100, the points on AB conform to a linear function:

[0030] y = tanα*x + (y A / tanα+100)tanα;

[0031] Z, any point on AB n The ordinate is: y = tanα*x Z +(y A / tanα+100)tanα;

[0032] any point Z n The magnification factor m of the ordinate is: (x Z y Z *m);

[0033] The x-coordinate of the dividing point B is... B It can be calculated using the function corresponding to BC:

[0034] x B =600tanβ-(yA / tanα+100)tanα / tanα-tanβ.

[0035] Preferably, the processing of the multiple raw shear force data in step S03 employs a partitioned amplification method, which specifically includes the following steps:

[0036] S31b: Divide the effective measurement areas to the left and right of the zero point into several sub-regions;

[0037] S32b. Use a vertical force loading device to perform calibration loading at the center point of each sub-region and record the sensor output amplitude at each center point.

[0038] S33b. Using the calibration amplitude of the selected reference sub-region center point as the reference value, calculate the proportionality coefficient of the calibration amplitude of each other sub-region center point relative to the reference value.

[0039] S34b, Based on the original data point Z0(x) Z y Z The sub-region where ) is located is magnified using the scaling factor m of that sub-region: Z1(x) Z y Z *m)

[0040] S35b: Take the absolute value of the data processed to the right of zero.

[0041] Preferably, in step S03, the continuous vertical wheel-rail force data within the detection area are obtained as curves.

[0042] Preferably, the preset identification threshold in step S04 is the measured value obtained from the actual measurement of vehicle type, running speed and track status.

[0043] In the above technical solution, the method for identifying wheel tread damage using a shear force sensor provided by the present invention has the following beneficial effects:

[0044] 1. This system completely abandons the two-dimensional force sensor required by traditional TPDS systems to be installed directly beneath the track on a dedicated rail, instead employing a universal rail-mounted shear force sensor. The rail-mounted shear force sensor is directly clamped onto the rail web at the gap in the rail, eliminating the need to replace the dedicated rail or perform subsequent tamping and track fine-tuning. This greatly simplifies the process, significantly shortens construction time, and, more importantly, substantially reduces the amount of engineering work, material costs, and post-construction maintenance difficulty, offering outstanding economic and engineering convenience advantages.

[0045] 2. It can effectively eliminate the nonlinear attenuation characteristics of sensor measurements caused by changes in wheel position, and successfully reconstruct the continuous and accurate vertical wheel-rail force variation curve within the test area. By analyzing the ratio of wheel-rail force to average wheel-rail force at each location point, it can accurately capture the peak value of abnormal impact load caused by wheel tread damage. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0047] Figure 1 A schematic diagram of the installation layout structure of the rail and the rail-mounted shear force sensor provided in an embodiment of the present invention;

[0048] Figure 2 This is a line graph showing the amplitude of each acquisition point provided in Embodiment 1 of the present invention;

[0049] Figure 3 This is a graph showing the relationship between the sensor output and the wheel position provided in Embodiment 2 of the present invention.

[0050] Figure 4 This is a line graph showing the data processing to the right of zero point provided in Embodiment 2 of the present invention;

[0051] Figure 5 This is a line graph for absolute value extraction provided in Embodiment 2 of the present invention;

[0052] Figure 6 This is an amplitude diagram provided in Embodiment 3 of the present invention.

[0053] Explanation of reference numerals in the attached figures:

[0054] 1. Rail 1; 2. Rail-mounted shear force sensor. Detailed Implementation

[0055] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0056] This invention provides a technical solution: a method for identifying wheel tread damage using a shear force sensor, such as... Figure 1 As shown, it includes the following steps:

[0057] S01. A detection area is created on rail 1. The detection area includes sub-detection areas composed of multiple rail-mounted shear force sensors 2 that are independent of each other.

[0058] Specifically, the rail-mounted shear force sensor 2 typically consists of three parts: a clamping structure, a strain gauge, and a clamp, see... Figure 1 The rail clamping structure forms the basis for installing the strain gauge, which is bolted to the rail 1. The clamping structure applies a load to the strain gauge, ensuring it adheres tightly to the web of the rail 1, thereby capturing the tangential deformation of the rail 1. Specifically, the rail clamping shear sensor 2 is installed at the center of the sleeper gap.

[0059] S02. When the wheel passes through the detection area, each sub-measurement area collects the original shear force data of the wheel within its effective measurement area. The effective measurement area ranges from ±100mm to ±600mm.

[0060] Specifically, the characteristics of the raw data collected by the shear force sensor are as follows: when the wheel passes the shear force sensor, the output of the shear force sensor gradually increases to a peak value A. max The change process is non-linear, and the curvature gradually increases. When the wheel is directly above the shear force sensor (horizontal axis is 0), the shear force rapidly changes direction, from the positive peak value A. max Point becomes the reverse peak A min Point, then the output is the reverse maximum peak value A min The radius gradually decreases and the curvature gradually decreases. The effective measurement area is the region from 100 to 600 mm and from -100 to -600 mm. The region from -100 to 100 mm is called the blind zone. Generally, the relationship curve between the sensor output and the wheel position can be found in [reference needed]. Figure 2 .

[0061] S03. Process the raw shear force data collected in each sub-test area to obtain consecutive vertical wheel-rail force data within the test area (the raw curve is obtained based on the consecutive vertical wheel-rail force data). The processing of the collected raw shear force data includes amplifying the amplitude of the data in the effective areas to the left and right of the zero point and taking the absolute value of the data to the right of the zero point, wherein:

[0062] The range to the left of zero is -600mm ≤ n ≤ -100mm;

[0063] The range to the right of zero is 100mm≤n≤600mm.

[0064] S04. Based on the obtained continuous vertical wheel-rail force data, calculate the ratio of the amplitude of each point in the effective measurement area to the average amplitude of the area. If the ratio exceeds the preset identification threshold (the preset identification threshold is the actual measured value obtained based on the actual measurement of vehicle type, running speed and track condition), it is determined that there is an abnormal impact load caused by wheel tread damage at this location, and wheel tread damage is identified.

[0065] Specifically, when the wheel tread is damaged, and the damaged area comes into contact with rail 1, it will generate an impact load on rail 1. In the sensor data, the amplitude of the impact load at that location will abnormally increase. (See...) Figure 2 By amplifying and restoring the amplitude of each acquisition point and dividing it by the average amplitude within the effective area, the ratio of wheel-rail force to average wheel-rail force at different locations can be obtained. By setting the recognition range of the ratio, it can be determined whether there is tread damage.

[0066] In summary, this embodiment completely eliminates the two-dimensional force sensor required by traditional TPDS systems to be installed directly beneath the track on a dedicated rail 1, instead employing a universal rail-mounted shear sensor 2. The rail-mounted shear sensor 2 is directly mounted on the web of the rail 1 at the gap in the rail 1, eliminating the need to replace the dedicated rail 1 or perform subsequent tamping and track fine-tuning. This significantly simplifies the process, shortens construction time, and, more importantly, drastically reduces the workload, material costs, and post-installation maintenance difficulty, offering outstanding economic and engineering convenience advantages. It effectively eliminates the nonlinear attenuation characteristics of sensor measurements caused by changes in wheel position, successfully reconstructing a continuous and accurate vertical wheel-rail force variation curve within the measurement area. By analyzing the ratio of wheel-rail force to the average wheel-rail force at each location point, it can accurately capture the peak value of abnormal impact loads caused by wheel tread damage.

[0067] Example 2

[0068] Based on the above embodiment one, this embodiment aims to provide a function amplification method for processing the multiple raw shear force data collected in step S03. When the wheel passes through the measurement area, without tread damage, the vertical wheel-rail force is basically the same at any position in the measurement area. However, due to the characteristics of the shear force sensor, the voltage output decreases as the wheel is farther from the sensor. In order to obtain the true vertical wheel-rail force variation law in the measurement area, it is necessary to amplify the ordinate of each collection point sequentially with the peak point A as the reference. Let the coordinate of any collection point Z before amplification to the left of the zero point in the effective measurement area be (x Z y Z Let m be the magnification factor that needs to be multiplied by the ordinate of any point Z, and n be the relative position of the wheel and the sensor. The function of m with respect to n is m = f(n) (-600 ≤ n ≤ -90). Then, after proportional magnification, the coordinates of point Z1 are Z1(x...). Z y Z *m), the detailed steps include:

[0069] S31a. Use the positive peak point A to the left of the zero point in the original data curve or the specified calibration point as the reference point;

[0070] S32a. Based on the data set collected within the effective measurement area to the left of the zero point, construct at least one fitted straight line to approximate the original data curve;

[0071] S33a. Based on the parameters of the fitted straight line and the amplitude of the reference point, calculate Z0(x) for any original data point within the region. Z y Z The corresponding amplitude amplification factor m is used to obtain the amplified data point Z1(x). Z y Z *m)

[0072] S34a. Perform the same amplification process on the data within the effective measurement area to the right of the zero point as in steps S32a to S33a, and then take the absolute value of the processed data.

[0073] Furthermore, in step S32a, the fitted lines (not limited to two; the more lines, the higher the fit to the curve) approximate the original data curve with at least two lines: a fitted line AB and a fitted line BC. The slope of segment AB is tanα, and the slope of segment BC is tanβ. Any point Z on the fitted line... n The magnification factor of the ordinate replaces the magnification factor of any point Z0 in the original data, where the coordinates of point A are (-90, y...). A The coordinates of point B are (x, y). B y B Point C has coordinates (-600, 0), Z... n Let Z0 be the projection of the fitted line, where:

[0074] When -600≤n≤x B At that time, based on the coordinates of point C and the slope, we know that the point on BC conforms to a linear function: y = tanβ*x + 600tanβ;

[0075] Any point Z on BC n The ordinate is: y = tanβ*x Z +600tanβ;

[0076] Then any point Z n The magnification factor m of the ordinate is: m = y A / (tanβ*x Z +600tanβ);

[0077] Any point Z after processing n The coordinates are: (x Z y Z *m), such as Figure 4 As shown;

[0078] Then, when x B When n ≤ -100, the points on AB conform to a linear function:

[0079] y = tanα*x + (y A / tanα+100)tanα;

[0080] Z, any point on AB n The ordinate is: y = tanα*x Z +(y A / tanα+100)tanα;

[0081] Any point Z n The magnification factor m of the ordinate is: (x Z y Z *m), such as Figure 5 As shown;

[0082] The x-coordinate of the dividing point B is x B It can be calculated using the function corresponding to BC:

[0083] x B =600tanβ-(yA / tanα+100)tanα / tanα-tanβ.

[0084] Example 3

[0085] Based on the above embodiment one, this embodiment aims to provide a method for processing multiple raw shear force data in step S03 using a partitioned amplification method, which specifically includes the following steps:

[0086] S31b: Divide the effective measurement areas to the left and right of the zero point into several sub-regions;

[0087] S32b. Use a vertical force loading device to perform calibration loading at the center point of each sub-region and record the sensor output amplitude at each center point.

[0088] S33b. Using the calibration amplitude of the selected reference sub-region center point as the reference value, calculate the proportionality coefficient of the calibration amplitude of each other sub-region center point relative to the reference value.

[0089] S34b, Based on the original data point Z0(x) Z y Z The sub-region where ) is located is magnified using the scaling factor m of that sub-region: Z1(x) Z y Z *m)

[0090] S35b: Take the absolute value of the data processed to the right of zero.

[0091] Specifically, the effective measurement area of ​​the sensor is divided into 10 sub-regions, each 100mm in diameter, designated as ab and a1-b1. However, the number of sub-regions is not limited to 10; the denser the sub-regions, the more accurate the processed data. A vertical force loading device is used on the rail to load the center points AE and A1-E1 of each region, and the amplitude of the sensor output is recorded. (See attached diagram.) Figure 6 Taking the data to the left of point 0 as an example, calculate the ratio of the amplitude of the center calibration point of each region to that of the calibration point in region a, and obtain the amplification ratio of the sensor measurement data for each region. Based on the ratio of each region, restore the original data to obtain the true changes in wheel-rail force within the test area. Repeat steps S03 and S04 in Scheme 1 to complete the identification of impact load and tread damage.

[0092] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for identifying wheel tread damage using a shear force sensor, characterized in that, Includes the following steps: S01. A detection area is created on the rail, the detection area comprising multiple rail-mounted shear force sensors that independently form sub-detection areas; S02. When the wheel passes through the detection area, each of the sub-measurement areas collects the original shear force data of the wheel within its effective measurement area, wherein the effective measurement area ranges from ±100mm to ±600mm. S03. Process the raw shear force data collected in each of the sub-test areas to obtain consecutive vertical wheel-rail force data within the test area; S04. Based on the obtained continuous vertical wheel-rail force data, calculate the ratio of the amplitude of each point in the effective measurement area to the average amplitude of the area. If the ratio exceeds the preset identification threshold, it is determined that there is an abnormal impact load caused by wheel tread damage at that location, and wheel tread damage is identified.

2. The method for identifying wheel tread damage using a shear force sensor according to claim 1, characterized in that, Multiple rail-mounted shear force sensors are installed at the web of the rail at the center of the rail gap.

3. The method for identifying wheel tread damage using a shear force sensor according to claim 1, characterized in that, In step S02, the original shear force data is the output of the target wheel, which increases non-linearly to a positive peak value A when it approaches the rail-mounted shear force sensor. max When the target wheel is directly above the rail-mounted shear force sensor, the output rapidly reverses to reach the reverse peak value A. min .

4. The method for identifying wheel tread damage using a shear force sensor according to claim 1, characterized in that, The processing of the multiple raw shear force data collected in step S03 includes amplifying the amplitude of the data in the effective regions to the left and right of the zero point and taking the absolute value of the data to the right of the zero point, wherein: The range to the left of the zero point is -600mm≤n≤-100mm; The range to the right of the zero point is 100mm≤n≤600mm.

5. A method for identifying wheel tread damage using a shear force sensor according to claim 4, characterized in that, The processing of the multiple raw shear force data in step S03 using the function amplification method specifically includes the following steps: S31a. Use the positive peak point A to the left of the zero point in the original data curve or the specified calibration point as the reference point; S32a. Based on the data set collected within the effective measurement area to the left of the zero point, construct at least one fitted straight line to approximate the original data curve; S33a. Based on the parameters of the fitted straight line and the amplitude of the reference point, calculate Z0(x) for any original data point within the region. Z y Z The corresponding amplitude amplification factor m is used to obtain the amplified data point Z1(x). Z y Z *m) S34a. Perform the same amplification process on the data within the effective measurement area to the right of the zero point as in steps S32a to S33a, and then take the absolute value of the processed data.

6. A method for identifying wheel tread damage using a shear force sensor according to claim 5, characterized in that, In step S32a, the fitted straight line approximates at least two original data curves, namely the fitted straight line AB and the fitted straight line BC, wherein the slope of segment AB is tanα, the slope of segment BC is tanβ, and any point Z on the fitted straight line... n The magnification factor of the ordinate replaces the magnification factor of any point Z0 in the original data, where the coordinates of point A are (-90, y...). A The coordinates of point B are (x, y). B y B Point C has coordinates (-600, 0), Z... n Let Z0 be the projection of the fitted line, where: When -600≤n≤x B At that time, based on the coordinates of point C and the slope, we know that the point on BC conforms to a linear function: y = tanβ*x + 600tanβ; Any point Z on BC n The ordinate is: y = tanβ*x Z +600tanβ; Then any point Z n The magnification factor m of the ordinate is: m = y A / (tanβ*x Z +600tanβ); Any point Z after processing n The coordinates are: (x Z y Z *m); Then, when x B When n ≤ -100, the points on AB conform to a linear function: y=tanα*x+(y A / tanα+100)tanα; Z, any point on AB n The ordinate is: y = tanα*x Z +(y A / tanα+100)tanα; any point Z n The magnification factor m of the ordinate is: (x Z y Z *m); The x-coordinate of the dividing point B is x B It can be calculated using the function corresponding to BC: x B =600tanβ-(yA / tanα+100)tanα / tanα-tanβ。 7. A method for identifying wheel tread damage using a shear force sensor according to claim 4, characterized in that, The step S03, which involves processing the collected raw shear force data using a partitioned amplification method, specifically includes the following steps: S31b: Divide the effective measurement areas to the left and right of the zero point into several sub-regions; S32b. Use a vertical force loading device to perform calibration loading at the center point of each sub-region and record the sensor output amplitude at each center point. S33b. Using the calibration amplitude of the selected reference sub-region center point as the reference value, calculate the proportionality coefficient of the calibration amplitude of each other sub-region center point relative to the reference value. S34b, Based on the original data point Z0(x) Z y Z The sub-region where ) is located is magnified using the scaling factor m of that sub-region: Z1(x) Z y Z *m) S35b: Take the absolute value of the data processed to the right of zero.

8. A method for identifying wheel tread damage using a shear force sensor according to claim 1, characterized in that, In step S03, obtaining consecutive vertical wheel-rail force data within the detection area is a curve.

9. A method for identifying wheel tread damage using a shear force sensor according to claim 1, characterized in that, The preset identification threshold in step S04 is the measured value obtained based on the actual measurement of vehicle type, running speed and track status.

Images