阅读说明：本技术 一种用于抑制钢轨漏磁检测提离干扰的滤波方法 (Filtering method for inhibiting steel rail magnetic flux leakage detection lift-off interference ) 是由 贾银亮 周黄健 王平 张事成 于 2021-08-10 设计创作，主要内容包括：本发明公开了一种用于抑制钢轨漏磁检测提离干扰的滤波方法,包括如下步骤：在磁轭下方排布两列,各N个磁敏传感器,分别检测x、z方向磁场；每个磁敏传感器采样点数为M；采样数据前后各补B个采样点,从第一个实际采样点开始向后推进计算；对某采样点滤波时,以同一行两个磁敏传感器为一组,取该点前后各B个采样点；先找到这些采样点中,两个磁敏传感器差值的绝对值的最大值；再比较N组磁敏传感器的最大值,找到其中最小值所对应的一组；将该组的采样点作为参考信号,得到滤波后的结果。本发明对传感器阵列排布进行合理的布局,并根据每一个采样点的计算和判断,能够有效的抑制提离干扰,增加伤损信号的信噪比,适用于钢轨伤损的漏磁检测。(The invention discloses a filtering method for inhibiting steel rail magnetic flux leakage detection lift-off interference, which comprises the following steps of: two lines are arranged below the magnetic yoke, and N magnetosensors respectively detect magnetic fields in the x direction and the z direction; the number of sampling points of each magnetic sensor is M; b sampling points are respectively supplemented before and after the sampling data, and calculation is carried out backward from the first actual sampling point; when filtering a certain sampling point, taking two magnetic sensors in the same row as a group, and taking B sampling points before and after the point; firstly, finding the maximum value of the absolute value of the difference value of the two magnetic sensors in the sampling points; then, the maximum values of the N groups of magnetic sensors are compared, and a group corresponding to the minimum value is found; and taking the group of sampling points as a reference signal to obtain a filtered result. The invention carries out reasonable layout on the sensor array arrangement, can effectively inhibit lift-off interference according to the calculation and judgment of each sampling point, increases the signal-to-noise ratio of a damage signal, and is suitable for the magnetic leakage detection of the rail damage.)

1. A filtering method for inhibiting steel rail magnetic flux leakage detection lift-off interference is characterized by comprising the following steps:

s1: two lines of magneto-dependent sensors are arranged below the magnetic yoke along the detection direction at intervals, wherein N magneto-dependent sensors are arranged in each two lines and are respectively used for measuring magnetic fields in the x direction and the z direction;

s2: sampling each magnetic sensor, wherein the number of sampling points of each magnetic sensor is M;

s3: b data points are supplemented before and after the sampling data, calculation is carried out by backward advancing from the first point of the actual sampling data, and B sampling points before and after the point are taken;

s4: two magnetic sensors which respectively detect magnetic fields in the x direction and the z direction in the same row are taken as a group, N groups of magnetic sensors are shared, and difference processing is carried out on each group of sampling data to find out the point with the maximum absolute value of the difference;

s5: comparing the N groups of points with the maximum absolute value obtained in step S4 to find a group of channels corresponding to the point with the minimum absolute value;

s6: and taking the sampling point corresponding to the channel obtained in the step S5 as a reference signal, and taking the difference between the original signal and the reference signal to obtain a filtered result.

2. The filtering method for suppressing the lifting-off interference of the magnetic flux leakage detection of the steel rail as claimed in claim 1, wherein the spacing distance between the two arrays of the magnetic sensors in the step S1 is obtained according to the minimum damage size, the inspection speed and the sampling speed, and is specifically obtained according to the equation H'_z(x,z₀) Solution with 0 greater than zero:

wherein the minimum damage width measured is 2a, the depth measured is b, and the lift-off when the sensor is not vibrated is z₀，σ_msThe surface magnetic charge density of the damaged side surface.

3. The filtering method for suppressing the steel rail magnetic flux leakage detection lift-off interference according to claim 1, wherein the calculation formula of B in the step S3 is as follows:

wherein, it is v meter/second to patrol and examine speed, and sampling speed is s point/second, and 2L are the range of distribution of damage leakage magnetic field, and L is calculated by following formula:

wherein the minimum damage width measured is 2a, the depth measured is b, and the lift-off when the sensor is not vibrated is z₀。

4. The filtering method for suppressing the steel rail magnetic flux leakage detection lift-off interference according to claim 1 or 4, wherein the point at which the absolute value of the difference value is maximum in step S4 is obtained by:

two magnetic sensors in the same group are respectively set as S_x[i]、S_z[i](i＝1、2、......、N)，S_x[i]、S_z[i]The sampling result of (a) is an array: s_x[i，j]、S_z[i，j](j ═ 1, 2,... and M), the array R is calculated_x[i，k]、R_z[i，k](k＝1、2、......、M+2B)：

Setting a cyclic variable q, wherein the initial value is B + 1;

obtaining | R_x[i，q-B]-R_z[i，q-B]|、|R_x[i，q-B+1]-R_z[i，q-B+1]|、......、|R_x[i，q+B]-R_z[i，q+B]The maximum value in | is denoted MAXi.

Technical Field

The invention belongs to the technical field of nondestructive testing, relates to a digital filtering technology, and particularly relates to a filtering method for inhibiting steel rail magnetic flux leakage testing lift-off interference.

Background

Nondestructive testing is a new discipline for evaluating structural abnormalities and damage, that is, detecting the presence or absence of damage such as cracks or inclusions in the internal structure, physical properties, or state of a workpiece or material to be tested, without damaging the workpiece or material, by using changes in the response to heat, sound, electricity, light, magnetism, or the like, caused by the presence of the abnormalities and damage in the internal structure of the material. The magnetic flux leakage nondestructive detection method can detect the surface and near-surface damage of ferromagnetic material workpieces, has the advantages of high detection sensitivity, high speed, low requirement on the surface cleanliness of the workpieces, low cost, simple operation and the like, and is widely applied to nondestructive detection of surface damage of ferromagnetic materials, such as steel rails, steel pipes and other workpieces.

The vertical distance between the magnetic sensor and the measured workpiece is called lift-off, and the distribution of leakage magnetic fields is different under different lift-off conditions. When the probe is used for circular detection on the surface of a workpiece, the lift-off is changed under the influence of factors such as vibration and the like, so that the output change of the magnetic sensor is called lift-off interference. The lift-off interference is superimposed on the detection signal, which makes the damage signal difficult to distinguish and is not beneficial to the measurement of the damage. The lift-off interference can be suppressed by a hardware circuit such as a differential circuit, but since the lift-off interference is affected by factors such as inspection speed and the surface state of the steel rail, the same circuit is difficult to adapt to various situations, and the effect of suppressing the lift-off interference is affected, digital filtering is generally required in addition to hardware filtering. Because the amplitude of the lift-off interference is influenced by factors such as inspection speed and the like, and the frequency spectrum of the lift-off interference is often superposed with a damaged signal, the existing digital filtering technology is difficult to effectively filter. Since the leakage magnetic field of the damage is distributed in a certain range, the farther the damage is, the weaker the leakage magnetic field is, the more easily the leakage magnetic field is interfered to cause misjudgment, and in order to effectively detect, identify and reconstruct the damage, a new digital filtering technology is needed to effectively inhibit the influence of lift-off interference.

Disclosure of Invention

The purpose of the invention is as follows: in order to overcome the defects in the prior art, the filtering method for inhibiting the lifting-off interference of the magnetic flux leakage detection of the steel rail is provided, the lifting-off interference in the routing inspection process can be effectively inhibited, the signal-to-noise ratio of a damage signal is increased, and the filtering method is suitable for the magnetic flux leakage detection of the damage of the steel rail.

The technical scheme is as follows: in order to achieve the above object, the present invention provides a filtering method for suppressing the lift-off interference in the magnetic flux leakage detection of a steel rail, comprising the steps of:

s1: two lines of magneto-dependent sensors are arranged below the magnetic yoke along the detection direction at intervals, wherein N magneto-dependent sensors are arranged in each two lines and are respectively used for measuring magnetic fields in the x direction and the z direction;

s2: sampling each magnetic sensor, wherein the number of sampling points of each magnetic sensor is M;

s3: b data points are supplemented before and after the sampling data, calculation is carried out by backward advancing from the first point of the actual sampling data, and B sampling points before and after the point are taken;

s4: two magnetic sensors which respectively detect magnetic fields in the x direction and the z direction in the same row are taken as a group, N groups of magnetic sensors are shared, and difference processing is carried out on each group of sampling data to find out the point with the maximum absolute value of the difference;

s5: comparing the N groups of points with the maximum absolute value obtained in step S4 to find a group of channels corresponding to the point with the minimum absolute value;

s6: and taking the sampling point corresponding to the channel obtained in the step S5 as a reference signal, and taking the difference between the original signal and the reference signal to obtain a filtered result.

Further, the separation distance between the two columns of magnetic sensors in step S1 is obtained according to the minimum damage size, the inspection speed and the sampling speed, specifically, the separation distance is obtained through an equation H'_z(x，z₀) Solution with 0 greater than zero:

wherein the minimum damage width measured is 2a, the depth measured is b, and the lift-off when the sensor is not vibrated is z₀，σ_msThe surface magnetic charge density of the damaged side surface.

Further, σ in the step S1_msIs calculated from the following formula:

where μ is the permeability of the material and H is the applied magnetic field strength.

Further, the calculation formula of B in step S3 is:

wherein, it is v meter/second to patrol and examine speed, and sampling speed is s point/second, and 2L are the range of distribution of damage leakage magnetic field, and L is calculated by following formula:

wherein the minimum damage width measured is 2a, the depth measured is b, and the lift-off when the sensor is not vibrated is z₀。

Further, the obtaining manner of the point at which the absolute value of the difference is maximum in step S4 is as follows:

two magnetic sensors in the same group are respectively set as S_x[i]、S_z[i](i＝1、2、......、N)，S_x[i]、S_z[i]The sampling result of (a) is an array: s_x[i，j]、S_z[i，j](j ═ 1, 2,... and M), the array R is calculated_x[i，k]、R_z[i，k](k＝1、2、......、M+2B)：

Setting a cyclic variable q, wherein the initial value is B + 1;

obtaining | R_x[i，q-B]-R_z[f，q-B]|、|R_x[i，q-B+1]-R_z[i，q-B+1]|、......、|R_x[i，q+B]-R_z[i，q+B]Maximum value in |, noted as MAX_i。

Has the advantages that: compared with the prior art, the sensor array is reasonably arranged, and lifting interference can be effectively inhibited according to calculation and judgment of each sampling point; the scheme of the invention can be used independently, can also be combined with the existing hardware circuits such as a differential circuit and the like, effectively inhibits lift-off interference, can improve the detection rate of the damage when small damage needs to be detected or the damage needs to be accurately reconstructed, reduces the false alarm rate, improves the reconstruction accuracy, and is suitable for the surface of ferromagnetic materials such as steel rails, steel pipes and the like or the itinerant detection only indicating the damage.

Drawings

FIG. 1 is a perspective view of a detection device;

FIG. 2 is a schematic plan view of the detecting device;

FIG. 3 is a flow chart illustrating a filtering method according to the present invention;

FIG. 4 is a magnetic field distribution plot of the damage in the x and z directions;

FIG. 5 is a schematic view of the minimum lesion width and depth;

fig. 6 is a diagram showing distribution of leakage magnetic fields of different sizes.

Detailed Description

The present invention is further illustrated by the following figures and specific examples, which are to be understood as illustrative only and not as limiting the scope of the invention, which is to be given the full breadth of the appended claims and any and all equivalent modifications thereof which may occur to those skilled in the art upon reading the present specification.

The invention provides a filtering method for inhibiting steel rail magnetic flux leakage detection lift-off interference, which comprises the following steps of:

s1: as shown in fig. 1 and 2, two lines of magnetosensitive sensors are arranged on the surface of the rail head of the detected steel rail at intervals along the detection direction right below the magnetic yoke and are respectively used for measuring magnetic fields in the x and z directions, wherein the distance between the two lines of magnetosensitive sensors is l;

s2: sampling each magnetic sensor, wherein the number of sampling points of each magnetic sensor is M;

s3: b data points are supplemented before and after the sampling data, calculation is carried out by backward advancing from the first point of the actual sampling data, and B sampling points before and after the point are taken;

s4: two magnetic sensors which respectively detect magnetic fields in the x direction and the z direction in the same row are taken as a group, N groups of magnetic sensors are shared, and difference processing is carried out on each group of sampling data to find out the point with the maximum absolute value of the difference;

s5: comparing the N groups of points with the maximum absolute value obtained in step S4 to find a group of channels corresponding to the point with the minimum absolute value;

s6: and taking the sampling point corresponding to the channel obtained in the step S5 as a reference signal, and taking the difference between the original signal and the reference signal to obtain a filtered result.

Based on the above scheme, the filtering method of the present invention is analyzed by combining the existing method as follows:

because the leakage magnetic field of the damage is weak, the distribution range is small, and the detection range of a single magnetic sensor is limited. In order to cover the entire rail head surface with the detection range, the conventional method arranges a row of magnetic sensors along the y direction, so that the output of the magnetic sensors is not only affected by the damage signal, but also affected by the lift-off interference. The sampled data of the magnetic sensor can be changed due to the change of the damage and the lift-off, so that misjudgment is easily caused, and the interference caused by the lift-off is difficult to eliminate by a general filtering mode.

Because the magneto-sensitive sensors are close in distance, the data changes caused by lift-off changes are similar. The rail head surface damage is generally distributed at the gauge corners or in the middle of the top surface, and generally no damage throughout the entire surface occurs. Even if such damage occurs, it is generally not perpendicular to the x-direction. When a row of magnetic sensors pass through a damaged part, some magnetic sensors can detect the leakage magnetic field of the damage, and the output data of the magnetic sensors can change due to the damage; some magnetic sensors cannot measure the leakage magnetic field of the damage, and the data of the magnetic sensors are not changed due to the damage. The data of the magnetic sensor which detects the leakage magnetic field is influenced by the damage leakage magnetic field and the lift-off change, and the data of the magnetic sensor which does not detect the leakage magnetic field is only influenced by the lift-off change.

Since the common influence of the lift-off variation and the damage leakage magnetic field on the magnetic sensors may be positive superposition or negative cancellation, it is not preferable to simply take the minimum value as the reference signal for the detection data of a row of magnetic sensors. Considering the characteristics of the x-direction and z-direction magnetosensitive sensor damage signals, the x-direction magnetosensitive sensor signal is approximately unimodal and has a maximum value directly above the damage, while the z-direction magnetosensitive sensor signal is bimodal. The peak of the x-direction impairment signal and the negative peak of the z-direction signal impairment are inverted. At the damage, if one sensor measures the positive peak value of the leakage magnetic signal in the x direction and the other sensor measures the negative peak value of the leakage magnetic signal in the z direction, the absolute value of the difference of the output signals of the two sensors is larger than that of the single-path signal.

At the non-damaged position, the lifting-off interference signals in the x direction and the z direction are in the same direction. When the lift-off becomes smaller, the leakage magnetic signals in the x direction and the z direction become larger; when the lift-off becomes large, the leakage magnetic signals in both the x-direction and the z-direction become small. At the non-damage position, if one sensor measures a leakage flux signal in the x direction and the other sensor measures a leakage flux signal in the z direction, the absolute value of the output signals of the two sensors after the difference is obtained is smaller than that of the one-way signal.

Obviously, if two sensors respectively measuring magnetic flux leakage signals in the x direction and the z direction are taken as a pair, and a plurality of pairs of sensors are sequentially arranged in the y direction, as shown in fig. 1 and fig. 2, as long as the distance between the pair of sensors is reasonably set, the larger the absolute value after the output signals are differentiated, the smaller the possibility that damage is not measured is; the smaller the absolute value, the greater the likelihood that no damage is measured. As described above, during each sampling, several of the sensors in a row of the magnetosensitive sensors always detect no damage signal, and a pair of sensors with the smallest absolute value after the output signals are subtracted can be considered as a pair of sensors that do not detect a damage, and the output of the pair of sensors is taken as a reference signal that does not include a damage signal and only includes a lift-off interference signal, and the output signals of other pairs of sensors subtract the reference signal as a filtered result.

To ensure that a suitable reference signal is found, the absolute value of a pair of sensors after differencing the output signals at the lesion should be as large as possible. As shown in FIG. 5, when the minimum required width of the damage is 2a and the depth is b, the intensity of the leakage magnetic field at a point P (x, z) above the damage is H (x, z), and the component in the x direction is H_x(x, z) with a z-direction component of H_z(x, z) can be obtained from the expressions (1) and (2), respectively, and the change tendency of the damage is shown in FIG. 4.

σ_msThe surface magnetic charge density of the damaged side surface can be calculated by the following formula:

where μ is the permeability of the material and H is the applied magnetic field strength.

When the x-direction sensor detects a positive peak value, and the z-direction sensor detects a negative peak value, the absolute value of the difference between the output signals of the pair of sensors at the damage part is maximum, and the absolute value is set as Q.

According to formula (1), when extracted from z ═ z₀Invariably, H when x is 0_x(x，z₀) There is a maximum value. According to FIG. 4, H_z(x，z₀) Is given in the positive half-axis of x, deriving equation (2) and let H'_z(x，z₀) Solving for 0 yields a solution l that is greater than zero. Obviously, when the distance between two sensorsAt l, when the x-direction sensor detects a positive peak, the z-direction sensor detects a negative peak.

In practical engineering application, the distance l between the two sensors is determined according to the minimum damage width and the minimum damage depth measured according to requirements. The distribution of the leakage magnetic field of different sizes of the damage is shown in fig. 6. The distance l is calculated according to the minimum damage, the positive peak value in the x direction is larger for the damage with larger size, and when the positive peak value is detected by the x-direction sensor, the signal of the z-direction sensor with the distance l is smaller than the negative peak value of the minimum damage although the negative peak value is not detected by the z-direction sensor, namely the absolute value of the difference of the output signals of the sensors is larger than Q, and the value is larger, so that the selection of the reference signal is not interfered.

As can be seen from fig. 4, the leakage field of the damage has a certain range, and it is desirable to measure the leakage field as completely as possible for accurate identification and reconstruction of the damage. However, the leakage magnetic field is weak at a position far from the damage, and it is difficult to distinguish the leakage magnetic field from various kinds of interference. And the boundary condition of the leakage magnetic field is that the absolute value of the leakage magnetic field signal in the x direction reaches 5% of the absolute value of the peak value in the x direction. According to formula (1), when extracted from z ═ z₀Invariably, H when x is 0_x(x，z₀) Having a maximum value, i.e. the leakage field in the x-direction is greatest just above the centre of the damage, the maximum value being H_x(0，z₀). The farther from the center of the lesion, H_x(x，z₀) The smaller. When H is present_x(x，z₀) Is reduced toIn the meantime, it is considered that there is no distribution of the leakage magnetic field in the x direction, that is, the distribution of the leakage magnetic field in the x direction is within a range of 2L centered on the damage. According to equation (1), the boundary L of the minimum damage x is calculated by:

if the inspection speed is v m/s and the sampling speed is s point/s, the jth sensor of the x-direction sensor₀If the sampling point is at the leakage field boundary of the damageWithin, then areThe positive peak of the x-direction leakage field must be included in each sample point. Since the x-direction sensor is at the positive peak, the absolute value is the largest after the difference of the output signals of the pair of sensors, which is most beneficial to find out the reference signal, so that the reference signal is found at the j pair₀When filtering at the sampling point, [ j ]₀-B，j₀+B]，Within each sampling point, the maximum value MAXi (i is 1, 2, … … and N) of the absolute value after the output signals of the pair of sensors are differentiated is found out, and then the pair of sensors with the minimum MAXi is found out, and the output of the pair of sensors is used as a reference signal.

Based on the above analysis, the present embodiment describes the above filtering method in detail, and provides a filtering method for suppressing the rail magnetic flux leakage detection lift-off interference, as shown in fig. 3, which includes the following steps:

step 1: two magnetosensitive sensors S for respectively detecting magnetic fields in x-direction and z-direction in the same row_x[i]、S_z[i](i is 1, 2, … …, N) is a group, and N groups of magnetic sensors are shared; sampling each magnetic sensor, wherein the number of sampling points of each magnetic sensor is M and S_x[i]、S_z[i]The sampling result of (a) is an array: s_x[i，j]、S_z[i，j](j＝1、2、……、M)；

The spacing distance l of the two lines of magnetic-sensing sensors is obtained according to the minimum size of the damage, the inspection speed and the sampling speed, and is specifically the spacing distance is obtained through an equation H'_z(x，z₀) Solution with 0 greater than zero:

wherein the minimum damage width measured is 2a, the depth measured is b, and the lift-off when the sensor is not vibrated is z₀，G_msThe surface magnetic charge density of the damaged side surface.

Step 2: calculating an array R_x[i，k]、R_z[i，k](k＝1、2、……、M+2B)：

The inspection speed is v m/s, the sampling speed is s points/s, and L is calculated by the following formula:

as shown in FIG. 5, z₀When the rail flaw detection vehicle is at rest, the lift-off of the magnetic sensor from the surface of the rail is determined as a minimum damage width to be detected of 1/2 and b as a minimum damage depth to be detected, so that the minimum damage width to be detected is 2a and the depth is b, and the lift-off when the sensor does not vibrate is z₀。

And setting a loop variable q, wherein the initial value is B + 1.

If the lift-off of the probe z is detected₀＝1mm，a＝1mm，b＝1mm，The x and z direction magnetic field distribution of the lesion is shown in fig. 4. Obviously, if extracted from z ═ z₀Invariably, H when x is 0_x(x, z) has a maximum value, i.e. the leakage magnetic field in the x direction is maximum just above the center of the damage, and the maximum value is H_x(0，z₀). The farther from the center of the lesion, H_xThe smaller (x, z). When H is present_x(x, z) is reduced toWhen it is, it can be considered thatThe distribution of the leakage magnetic field in the x direction is not damaged, that is, the distribution of the leakage magnetic field in the x direction is considered to be within 2L.

And step 3: to find

|R_x[i，q-B]-R_z[i，q-B]|、|R_x[i，q-B+1]-R_z[i，q-B+1]|……、|R_x[i，q+B]-R_z[i，q+B]Maximum value in |, noted as MAX_i。

And 4, step 4: finding MAX₁、MAX₂、……、MAX_NMinimum of (d), record subscript i of the minimum₀；

And 5: s_x[i，q-B]＝R_x[i，q]-R_x[i₀，q]、s_z[i，q-B]＝R_z[i，q]-R_z[i₀，q]，i＝1、2、……、N；

Step 6: if q is not greater than M + B, turning to the step 3, otherwise, executing the step 7;

and 7: with S_x[i，j]、S_z[i，j]The results of the filtering are (i 1, 2, … …, N, j 1, 2, … …, M).

In this embodiment, the filtering method is applied as an example, and specifically, the following is performed:

the 16-path X-direction and Z-direction magnetic-sensitive sensors are arranged in parallel in the direction perpendicular to the train running direction by a certain steel rail top surface damage and magnetic leakage detection system, and the minimum damage required to be detected is that a is 1mm, b is 1mm, and z is₀With 1mm, L8 mm can be determined from the above equation using the MATLAB tool, and if v 1m/s and the sampling rate f 10khz, the sampling point is determined

Sampling and storing signals output by the 16-path magnetic sensor group, and supplementing 80 sampling points before and after sampling data to ensure that 80 data before and after the sampling data can be calculated from a first point to a last point; respectively obtaining the maximum value Maxi (i is 1, 2, 3 … … 16) of the absolute value of the difference between the sampling points in the x direction and the z direction of 80 actual sampling points in each sensor group, taking the sampling point of the channel corresponding to the minimum value of the maximum value array as a reference signal, and taking the formed reference signal as a noise channel and original data to perform adaptive filtering to obtain filtered data.