A rail detection method based on a rail surface detection system

Abstract

The application discloses a track detection method based on a track surface detection system, and the detection method comprises the following steps: a work computer establishes three-dimensional coordinates of a track to be detected and establishes a track section area in the width direction of the track; the work computer synthesizes track modeling data and stores normal track surface data according to profile data obtained by each sensor; the work computer collects data of the track section area in real time, compares the data with the normal track surface data, and judges whether each section area is abnormal; the work computer judges whether the gauge changes according to the height of the track section area; and the work computer synthesizes a track surface graph of a track section area where an abnormal point is located and corresponding track section length. The application has the advantages that the detection of abnormal conditions such as track surface foreign matters, track surface flatness, gauge changes, track surface screw fastening degree and tool surface slurry turning can be realized, the efficiency is high, and the safety hidden danger is low.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of subway track rail surface inspection, more particularly to a track detection method based on a track rail surface detection system. BACKGROUND

[0002] In the field of urban rail transit and high-speed rail, in order to ensure the safe and stable operation of trains, routine inspection operation needs to be carried out on the track area, i.e. the main line, through which the train travels. The detection content includes tunnels, bridges, track beds, tracks, overhead contact systems, auxiliary electrical equipment, etc. The detection time is generally during the window period, i.e. from early morning to about 4 o'clock after the train stops running, and the detection cycle is arranged differently in different cities. How to scientifically maintain such a large-scale operation line and ensure the stability and reliability of the infrastructure so that the rail transit can operate safely for a long time is a problem that must be faced and solved in the current stage of rail transit development.

[0003] During the operation of the subway tunnel, it will be affected by the surrounding civil construction, the vibration of the subway itself, the load or disturbance of the overlying soil, and the load of the surrounding buildings, etc., causing the subway tunnel structure to change in stress and easily causing tunnel convergence deformation. Therefore, it is necessary to regularly monitor the subway tunnel. At present, a detection trolley is commonly used in subway tunnels, which carries laser scanners, high-precision cameras, LED fill light modules and other equipment. The main function is to uniformly and continuously scan and shoot the inner wall of the iron tunnel. The focus is on daily inspection of the inner wall of the track. When safety abnormalities are found in the inner wall, they can be promptly investigated. Chinese Patent Publication No. CN110207608A discloses a subway tunnel deformation detection method based on three-dimensional laser scanning, which includes the following steps: S1, running a detection trolley on the track of the subway tunnel, taking the center of the detection trolley as the origin, and establishing a right-angle coordinate system of the subway tunnel section; S2, installing an angle meter, a laser range finder and a laser scanner on the detection trolley, and performing rotational measurement of the starting point tunnel section through the angle meter, the laser range finder and the laser scanner to obtain the coordinate data of all measurement points of the starting point tunnel section. The starting point coordinate data is sent to a data processing module. The method of the patent application can accurately detect the shape and size data of the tunnel section. Through comparative analysis of the tunnel three-dimensional space profile data and historical data, a tunnel section deformation report is formed, which can timely alarm, but lacks inspection of the track below the subway. During the daily operation of the track, abnormal conditions such as track surface foreign matter, track surface flatness, track gauge change, track surface screw fastening degree, and track surface slurry turning still need to be investigated by manual inspection, which is extremely low in efficiency and has high safety risks. SUMMARY

[0004] The technical problem to be solved by the present application is that the existing subway track detection method lacks detection of abnormal conditions such as track surface foreign matter, track surface flatness, track gauge change, track surface screw fastening, and track surface slurry turning, and still needs to be checked by manual inspection, which is extremely low in efficiency and has high safety hazards.

[0005] The present application solves the above technical problems by the following technical means: a track detection method based on a track surface detection system, the detection system is mounted on the bracket on the front side of the track inspection trolley, the detection system includes two sensors for detecting the data of the two tracks, two sensors for detecting the data of the track bed, and an industrial computer for collecting the data of each sensor, the detection method includes:

[0006] Step a): the industrial computer establishes a three-dimensional coordinate of the track to be detected, and establishes a track cross-section area in the track width direction;

[0007] Step b): the industrial computer synthesizes track modeling data and stores normal track surface data according to the profile data obtained by each sensor;

[0008] Step c): the industrial computer collects data of the track cross-section area in real time, compares with the normal track surface data, and judges whether each cross-section area is abnormal;

[0009] Step d): the industrial computer determines whether the track gauge changes according to the height of the track cross-section area;

[0010] Step e): the industrial computer synthesizes the track surface graph of the track section where the abnormal point is located and the corresponding track section length.

[0011] The present application mounts a track surface detection system on the inspection trolley, establishes a track cross-section area in the track width direction, realizes modeling of each area, and judges whether each cross-section area is abnormal through the method of track profile image data modeling recognition comparison, which is convenient for daily safety inspection of track surface foreign matter, track surface flatness, track surface screw fastening, and track surface slurry turning, determines whether the track gauge changes according to the height of the track cross-section area, and the whole scheme has high inspection efficiency and low safety hazards.

[0012] Further, the two sensors for detecting the data of the two tracks are installed at the two ends of the bracket and above the two tracks respectively, the two sensors for detecting the data of the track bed are installed at the middle part of the bracket and above the middle part of the track, and the industrial computer is installed on the bracket between the sensors for detecting the data of the track bed and the sensors for detecting the data of the track.

[0013] Further, the step a) includes:

[0014] The industrial computer establishes the X-axis, Y-axis and Z-axis coordinates of the track to be detected, the X-axis value represents the lateral length of the track, the Z-axis represents the longitudinal length of the track, and the Y-axis value represents the height of the track. The track section areas z1, z2, z3, z4, z5, z6 and z7 are established in the Z-axis direction, z1 is the right rail outer fastener area, z2 is the right rail surface area, z3 is the right rail inner fastener area, z4 is the track bed area, z5 is the left rail inner fastener area, z6 is the left rail surface area, and z7 is the left rail outer fastener area. The track lateral direction represents the distance direction between the two rails, and the track longitudinal direction represents the advancing direction of the track inspection trolley on the track.

[0015] Further, the step b) comprises:

[0016] The track inspection trolley moves on the track at a speed of V 车 The industrial computer establishes the X-axis section modeling data X 基 (i) of the track according to the profile data obtained by each sensor, the modeling data is the Y-axis height data, i is the section granularity of the track to be detected after being detected by the track inspection trolley, and the granularity has a linear proportional relationship with the track length L, K=L / i.

[0017] Further, the step b) further comprises:

[0018] After the full-length detection of the track to be detected is completed, the industrial computer stores the normal track surface data X 基 (i) of the track.

[0019]

[0020] Each section contains Y-axis height values of 7 zones divided along the z-axis, and the sampling numbers of the 7 zone height values are J, K, L, M, N, O and P, respectively. Y i1J represents the Jth height value of the z1 zone at the i point position of the track length. Y i2K represents the Kth height value of the z2 zone at the i point position of the track length. Y i3L represents the Lth height value of the z3 zone at the i point position of the track length. Y i4M represents the Mth height value of the z4 zone at the i point position of the track length. Y i5N represents the Nth height value of the z5 zone at the i point position of the track length. Y i6O represents the Oth height value of the z6 zone at the i point position of the track length. Y i7P represents the Pth height value of the z7 zone at the i point position of the track length.

[0021] Further, the step c) comprises:

[0022] During daily inspection, the track surface detection system collects track surface data along the track to be inspected, and synthesizes cross-section data X 采 (j) in real time. 基 (j) to see if the difference between the height value in the 7 regions in the j cross-section and the corresponding normal track surface data is greater than the threshold value of the corresponding region, and if it is greater than the threshold value, it is considered that the region has a safety anomaly, and the industrial computer obtains that the track has a certain region anomaly at the K*j length position of the track through the length coefficient, otherwise, if the difference is less than the threshold value, it is considered that the track surface condition is normal.

[0023] Further, the step d) comprises:

[0024] The sum of the sampling values in the z2 and z6 regions of the j cross-section is obtained as G 采 (j). 基 (j), if the difference is greater than the threshold value THG, it is judged that the track gauge has changed.

[0025] Further, G 采 (j) is obtained by the formula: K α＝1 Y j2α +∑ O β＝1 Y j6β The sum of the sampling values in the z2 and z6 regions of the j cross-section is obtained.

[0026] Further, G 基 (j) is obtained by the formula: K α＝1 Y' j2α +∑ O β＝1 Y' j6β The corresponding normal track surface data of G 采 (j) is obtained.

[0027] Further, the step e) comprises:

[0028] The industrial computer synthesizes the track surface graph of the track section in the cross-section region of the abnormal point and the corresponding track section length ΔL, and generates an abnormal data set {X 采 (j-ΔL / 2K), X 采 (j-ΔL / 2K+1),..., X 采 (j+ΔL / 2K)}, wherein X 采 (j-ΔL / 2K) represents a set of height values collected at the j-ΔL / 2K point position of the track length.

[0029] The advantages of the present application are:

[0030] (1) The application carries a track surface detection system on the inspection trolley, establishes a track section area in the width direction of the track, realizes modeling of each area, and judges whether each section area is abnormal through the modeling recognition and comparison method of the rail profile image data, so as to facilitate the daily safety inspection of the track surface foreign matter, track surface flatness, track surface screw fastening degree and track surface slurry turning of the track surface, and determine whether the track gauge changes according to the height of the track section area. The whole scheme has high inspection efficiency and low safety hidden danger.

[0031] (2) Since the width of the left and right rails of the track is fixed, the track surfaces of the left and right rails should be fixed as the z2 area and the z6 area of the z axis under normal circumstances, and the Y axis height of the two areas should be a fixed value. Once the track gauge changes, whether it is single rail displacement, double rail displacement or double rail parallel displacement, one or both of the two track surfaces will deviate from the z2 area and the z6 area of the z axis, resulting in a change in the Y value in the track surface area. Therefore, the application obtains G 采 (j) by summing the sampling values in the z2 and z6 areas of the two track surfaces of the j section, compares the corresponding normal track surface data G 基 (j), and if the difference between the two is greater than the threshold value THG, it is judged that the track gauge has changed. The detection method is simple and reliable. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 The installation position diagram of the detection system in the track detection method based on the track surface detection system disclosed by the embodiment of the application is shown;

[0033] Figure 2 The modeling reference diagram of the track surface to be detected in the track detection method based on the track surface detection system disclosed by the embodiment of the application is shown;

[0034] Figure 3 The comparison diagram of the detection reference data and the abnormal data of the track surface to be detected in the track detection method based on the track surface detection system disclosed by the embodiment of the application is shown;

[0035] Figure 4 The track gauge detection diagram of the track to be detected in the track detection method based on the track surface detection system disclosed by the embodiment of the application is shown;

[0036] Figure 5 The work flow chart of the track detection method based on the track surface detection system disclosed by the embodiment of the application is shown. DETAILED DESCRIPTION

[0037] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] like Figure 1 As shown, a track inspection method based on a track surface detection system is disclosed. The detection system is mounted on a bracket at the front of a track inspection trolley. It should be noted that the track inspection trolley and the bracket are not improvements of this invention. The track inspection trolley can be any existing trolley capable of performing track inspection operations, and the bracket can be any support structure that can be mounted at the front of the track inspection trolley; it merely supports the sensors and industrial control computer, and its structure is not particularly limited. The detection system includes two sensors for detecting data from the rails on both sides. Figure 1 The two sensors, numbered 1 and 4, are used to detect track bed data. Figure 1 The two sensors numbered 2 and 3, and the industrial control computer that collects data from each sensor ( Figure 1 (Ref. 5). The two sensors for detecting rail data on both sides are installed at both ends of the bracket and above the two rails respectively. The two sensors for detecting track bed data are installed in the middle of the bracket and above the middle of the rails. The industrial control computer is installed on the bracket between the sensors for detecting track bed data and the sensors for detecting rail data. All sensors are laser sensors, and the probes of the laser sensors are all pointed downwards at the rails.

[0039] The detection method includes the following steps:

[0040] Step a): The industrial computer establishes the three-dimensional coordinates of the track to be inspected and creates a track cross-sectional area in the track width direction; the specific process is as follows:

[0041] The industrial control computer establishes the X, Y, and Z axis coordinates of the track to be inspected. The X-axis value represents the lateral length of the track, the Z-axis value represents the longitudinal length of the track, and the Y-axis value represents the height of the track. A track cross-sectional region z1, z2, z3, z4, z5, z6, and z7 is established along the Z-axis direction. Figure 2 As shown, z1 is the outer fastener area of ​​the right rail, z2 is the rail surface area of ​​the right rail, z3 is the inner fastener area of ​​the right rail, z4 is the track bed area, z5 is the inner fastener area of ​​the left rail, z6 is the rail surface area of ​​the left rail, and z7 is the outer fastener area of ​​the left rail. The horizontal direction of the track indicates the distance between the two rails, and the vertical direction of the track indicates the direction of movement of the track inspection trolley on the track.

[0042] Step b): the industrial computer synthesizes track modeling data according to the profile data obtained by each sensor and stores normal track surface data; the specific process is as follows:

[0043] The track inspection trolley moves on the track at a speed of V 车 The industrial computer establishes the section modeling data X 基 (i)' of the track X-axis trend according to the profile data obtained by each sensor, the modeling data is the Y-axis height data, i is the section granularity of the track to be detected after being detected by the track inspection trolley according to its length, the granularity has a linear proportional relationship with the track length L, K=L / i. When i is determined, the length coefficient K is also determined, which is used for positioning the fault position when the track surface anomaly occurs. The track inspection trolley is located at the starting end of the track to be detected, and the moving trolley speed V 车 (Km / h) meets the requirements of collection frequency F(Hz) and resolution R(mm / ms), and the corresponding calculation formula is as follows

[0044] V 车 *(10000 / 36)=(1 / F)*1000*R

[0045] After the full-length detection of the track to be detected is completed, the industrial computer stores the normal track surface data X 基 (i),

[0046]

[0047] Each section contains 7 Y-axis height values divided along the z-axis, and the sampling numbers of the 7 height values are J, K, L, M, N, O, and P, respectively. i1J Y' i2K represents the Jth height value of the z2 zone at the i-length point position of the track, i3L represents the Lth height value of the z3 zone at the i-length point position of the track, i4M represents the Mth height value of the z4 zone at the i-length point position of the track, i5N represents the Nth height value of the z5 zone at the i-length point position of the track, i6O represents the Oth height value of the z6 zone at the i-length point position of the track, i7P represents the Pth height value of the z7 zone at the i-length point position of the track.

[0048] Step c): the industrial computer collects the data of the track section area in real time, compares it with the normal track surface data, and judges whether each section area has an anomaly; the specific process is as follows:

[0049] During daily inspection, the track surface detection system collects track surface data along the track to be detected and synthesizes section data X采 (j), and then compare this data with the corresponding normal track surface data X. 基 (j) Compare, such as Figure 3 As shown (dashed lines in the collected data represent abnormal data), check whether the difference between the height value in the 7 regions of the j-section and the corresponding normal track surface data is greater than the threshold of the corresponding region. If it is greater than the threshold, it is considered that there is a safety anomaly in the region. The industrial control computer uses the length coefficient to determine that an anomaly occurs in a certain region of the track at the K*j length position. Conversely, if the difference is less than the threshold, the track surface condition is considered normal.

[0050] The threshold values ​​for the seven zones correspond to TH1 to TH7, and the magnitude of these thresholds determines the sensitivity of the track surface detection system. A smaller threshold results in higher sensitivity, while a larger threshold results in lower sensitivity. This step can detect foreign objects on the track surface, track surface flatness, track bolt tightness, and safety risks such as slippage on the track edge that could lead to abnormal track surface height. Specifically... Figure 5 As shown, when the difference between the height value in region z1 of section j and the corresponding normal rail surface data is greater than the threshold TH1 of the corresponding region, the right rail external bolt is determined to be loose. When the difference between the height value in region z2 of section j and the corresponding normal rail surface data is greater than the threshold TH2 of the corresponding region, the right rail surface flatness is determined to be abnormal or there are foreign objects. When the difference between the height value in region z3 of section j and the corresponding normal rail surface data is greater than the threshold TH3 of the corresponding region, the right rail internal bolt is determined to be loose. When the difference between the height value in region z4 of section j and the corresponding normal rail surface data is greater than the threshold TH4 of the corresponding region, the track bed surface is determined to be abnormal or the cutting surface is frost-prone. When the difference between the height value in region z5 of section j and the corresponding normal rail surface data is greater than the threshold TH5 of the corresponding region, the left rail internal bolt is determined to be loose. When the difference between the height value in region z6 of section j and the corresponding normal rail surface data is greater than the threshold TH6 of the corresponding region, the left rail surface flatness is determined to be abnormal or there are foreign objects. If the difference between the height value in region z7 of the j-section and the corresponding normal rail surface data is greater than the threshold TH7 of the corresponding region, then the left rail outer fastener is determined to be loose.

[0051] Step d): The industrial control computer determines whether a change in track gauge has occurred based on the height of the track section area; the specific process is as follows:

[0052] For safety risk points caused by changes in track gauge, the method of determining the height of the track surface window is used, such as... Figure 4As shown, since the width of the left and right rails is fixed, the rail surfaces of the left and right rails should be fixed as z2 and z6 regions of the z axis under normal circumstances, and the Y axis height of the two regions should be a fixed value. Once the track gauge changes, whether it is single rail displacement, double rail displacement, or double rail parallel displacement, one or both of the two rail surfaces will definitely deviate from the z2 and z6 regions of the z axis, resulting in a change in the Y value in the rail surface region. The sum of the sampling values in the z2 and z6 regions of the two rail surfaces of the j section is obtained by G 采 (j), and the corresponding normal rail surface data G 基 (j) is obtained. If the difference between the two is greater than the threshold THG, it is determined that the track gauge has changed. Wherein,

[0053] The sum of the sampling values in the z2 and z6 regions of the two rail surfaces of the j section is obtained by G 采 (j) = ∑ K α＝1 Y j2α + ∑ O β＝1 Y j6β (j) is obtained.

[0054] The sum of the sampling values in the z2 and z6 regions of the two rail surfaces of the j section is obtained by G 基 (j) = ∑ K α＝1 Y' j2α + ∑ O β＝1 Y' j6β (j) is obtained. 采 (j) is obtained.

[0055] The value range of α and β is (1, K) and (1, O) respectively.

[0056] Step e): The industrial computer synthesizes the rail surface graph of the rail section in the section area of the abnormal point and the corresponding rail section length ΔL to generate an abnormal data set {X 采 (j-ΔL / 2K), X 采 (j-ΔL / 2K+1),... X 采 (j+ΔL / 2K)}, wherein X 采 (j-ΔL / 2K) represents a height value set collected at the j-ΔL / 2K point position of the rail length. The staff obtains the fault position and fault region synthesis graph through the background to make specific safety investigation, and realizes efficient and reliable subway rail surface inspection and maintenance.

[0057] Through the technical scheme, the rail surface detection system is carried on the inspection trolley, the rail section area is established in the rail width direction, the modeling of each area is realized, and the method of modeling identification comparison through the rail profile image data is used to judge whether each section area is abnormal, so that the daily safety inspection of the rail surface foreign matter, the rail surface flatness, the rail surface screw fastening degree and the rail surface slurry turning of the rail surface is facilitated, whether the gauge change occurs is determined according to the height of the rail section area, the whole scheme has high inspection efficiency and low safety hidden danger.

[0058] The above examples are only used to illustrate the technical solutions of the present application, but not limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A rail detection method based on a rail head detection system, characterized by, The detection system is mounted on the bracket in front of the track inspection trolley, and comprises two sensors for detecting data of two tracks, two sensors for detecting data of the track bed, and an industrial computer for collecting data of the sensors. Step a): the industrial computer establishes a three-dimensional coordinate of the track to be detected, and establishes a track section area in the track width direction; the industrial computer establishes X-axis, Y-axis and Z-axis coordinates of the track to be detected, the X-axis value represents the track transverse length, the Y-axis value represents the track height, and the Z-axis represents the track longitudinal length; the track section area z1, z2, z3, z4, z5, z6 and z7 are established in the Z-axis direction, z1 is the right rail outer fastener area, z2 is the right rail surface area, z3 is the right rail inner fastener area, z4 is the track bed area, z5 is the left rail inner fastener area, z6 is the left rail surface area, and z7 is the left rail outer fastener area; wherein, the track transverse direction represents the distance direction between the two tracks, and the track longitudinal direction represents the advancing direction of the track inspection trolley on the track; Step b): the industrial computer synthesizes track modeling data according to the profile data obtained by each sensor and stores normal track surface data; the track inspection trolley moves on the track at a V 车 speed, and the industrial computer establishes the section modeling data X 基 (i)' of the track X-axis trend according to the profile data obtained by each sensor, the modeling data is Y-axis height data, and i is the section granularity of the track to be detected after being detected by the track inspection trolley according to the length of the track, the granularity and the full length L of the track have a linear proportional relationship K=L / i. Step c): the industrial computer collects data of the track section area in real time, compares the data with normal track surface data, and judges whether each section area is abnormal; Step d): the industrial computer determines whether the track gauge changes according to the height of the track section area; Step e): the industrial computer synthesizes the track surface graph of the track section area where the abnormal point is located and the corresponding track section length.

2. The track detection method based on the track rail surface detection system according to claim 1, wherein, The two sensors for detecting data of two tracks are installed at two ends of the bracket and above the two tracks, the two sensors for detecting data of the track bed are installed at the middle part of the bracket and above the middle part of the tracks, and the industrial computer is installed on the bracket between the sensors for detecting data of the track bed and the sensors for detecting data of the tracks.

3. The track detection method based on the track rail surface detection system according to claim 1, wherein, The step b) further comprises: After the full track detection is completed, the industrial computer stores the normal track surface data X of the track to be detected 基 (i), X 基 (i)={ { {Y ' 111 , Y ' 112 ,.............. Y ' 11J}, { Y ' 121 , Y ' 122 , Y ' 12K}, { Y ' 131 , Y ' 132 ,.............. Y ' 13L}, { Y ' 141 , Y ' 142 ,.............. Y ' 14M}, { Y ' 151 , Y ' 152 , Y ' 15N}, { Y ' 161 , Y ' 162 , Y ' 16O}, { Y ' 171 , Y ' 172 , Y ' 17P}, }, ...... { { Y ' i11 , Y ' i12 ,.............. Y ' i1J}, { Y ' i21 , Y ' i22 ,.............. Y ' i2K}, { Y ' i31 , Y ' i32 , Y ' i3L}, { Y ' i41 , Y ' i42 , Y ' i4M}, { Y ' i51 , Y ' i52 , Y ' i5N}, { Y ' i61 , Y ' i62 ,.............. Y ' i6O}, { Y ' i71 , Y ' i72 ,.............. Y ' i7P}, }, }, wherein each section contains Y-axis height values of 7 zones divided along the z-axis, and the sampling number of the 7 zone height values are J, K, L, M, N, O, P, respectively, Y ' i1J represents the i-th point position z1 zone the J-th height value, Y ' i2K represents the i-th point position z2 zone the K-th height value, Y ' i3L represents the i-th point position z3 zone the L-th height value, Y ' i4M represents the i-th point position z4 zone the M-th height value, Y ' i5N represents the i-th point position z5 zone the N-th height value, Y ' i6O represents the i-th point position z6 zone the O-th height value, Y ' i7P represents the i-th point position z7 zone the P-th height value.

4. The track detection method based on the track rail surface detection system according to claim 3, characterized in that, The step c) comprises: During daily inspection, the track surface detection system collects track surface data along the track to be inspected, and synthesizes cross-section data X 采 (j) in real time. Then the data is compared with the corresponding normal track surface data X 基 (j) to see if the difference between the height value in the j cross-section and the corresponding normal track surface data in the 7 regions is greater than the threshold value of the corresponding region. If it is greater than the threshold value, it is considered that there is a safety anomaly in the region. The industrial computer obtains that there is a certain regional anomaly in the K*j length position of the track through the length coefficient. Conversely, if the difference is less than the threshold value, it is considered that the track surface condition is normal.

5. The track detection method based on the track rail surface detection system according to claim 4, characterized in that, The step d) comprises: G is obtained by summing the sampling values in the regions of the two rail surfaces z2 and z6 of the j section 采 (j) and the corresponding normal rail surface data G 基 (j), if the difference between the two is greater than a threshold value THG, it is determined that the gauge has changed.

6. The track detection method based on the track rail surface detection system according to claim 5, wherein, By formula G 采 (j) =∑ K α=1 Y j2α +∑ O β=1 Y j6β Obtain the sum of the sampling values in the regions of the two tracks z2 and z6 in the j-th section.

7. The track detection method based on the track rail surface detection system according to claim 5, wherein, By formula G 基 (j) =∑ K α=1 Y ' j2α +∑ O β=1 Y ' j6β Obtain G 采 (j) corresponding normal track surface data.

8. The track detection method based on the track rail surface detection system according to claim 5, wherein, The step e) comprises: The industrial computer synthesizes the track section surface graph of the section area where the abnormal point is located and the corresponding track section length AL, and generates an abnormal data set {X 采 (j-ΔL / 2K), X 采 (j-ΔL / 2K+1),... X 采 (j+ΔL / 2K)}, wherein, X 采 (j-ΔL / 2K) represents the track length j-ΔL / 2K point position collection height value set.

Images