Multi-channel track inspection data adaptive mileage correction and evaluation method

Abstract

The application discloses a multi-channel track detection data adaptive mileage correction and evaluation method, relates to the track detection field, and comprises the following steps: S1, acquiring reference data X and to-be-repaired data Y of a track; S2, extracting special points in the reference data and the to-be-repaired data as base points; S3, adopting digital curve matching of a local section to obtain optimal correction points; S4, correcting mileage error of to-be-repaired data between adjacent base points according to the optimal correction points; and S5, comprehensively evaluating correction results according to the precision of mileage deviation correction of the to-be-repaired data and the change of waveforms before and after correction. According to the track reference data X, the to-be-repaired data Y is accurately corrected in terms of mileage error, higher-precision mileage information is obtained, original data features of track detection data such as track orientation and height, track gauge, level, twist, curvature of left and right tracks are reserved, and the quality of mileage correction is evaluated by using scientific and faithful statistical indexes, so that the real geometric state of the track can be more accurately mastered.

Description

Technical Field

[0001] This invention relates to the field of track inspection, and in particular to a method for adaptive mileage correction and evaluation of multi-channel track inspection data. Background Technology

[0002] With the rapid development of high-speed railway construction and the significant speed increase of conventional railways in my country, ensuring the high-speed, stable, and safe operation of railway trains is the primary task of railway technical management and track maintenance departments. To guarantee the long-term stable and safe operation of trains, railway track maintenance departments regularly conduct track geometry condition inspections and formulate maintenance plans. However, due to factors such as relative creep between wheels and rails, wheel wear, unstable GPS signals, and manual mileage input errors by inspectors, the track mileage information detected by dynamic inspection vehicles, track inspection instruments, or combined navigation track measuring instruments often does not match the actual track mileage. This discrepancy not only affects the accurate location of track defects and delays maintenance work in defective sections, but also causes mileage offsets in dynamic or static track irregularity data detected at different times, making them incomparable for repeated inspections. This directly impacts the effectiveness of track irregularity condition assessment, prediction results, and maintenance plans.

[0003] The mileage error problem in track dynamic and static inspection data has attracted great attention from engineers, technicians and experts, and many effective methods have been proposed to solve the mileage error. The current problem with mileage error correction algorithms is that they stretch or compress the data to be repaired, changing the original waveform characteristics, including amplitude, wavelength and frequency. However, these characteristics are important indicators for assessing track irregularities and diagnosing track defects. Summary of the Invention

[0004] The purpose of this invention is to design a multi-channel track inspection data adaptive mileage correction and evaluation method to solve the above problems.

[0005] The present invention achieves the above objectives through the following technical solutions:

[0006] Adaptive mileage correction and evaluation methods for multi-channel track inspection data, including:

[0007] S1. Obtain the baseline data X and the data to be repaired Y of the track;

[0008] S2. Extract specific points in X and Y as base points, and denot the base point mileage of X as follows: The base point mileage of Y is denoted as ;

[0009] S3. The optimal correction point (or endpoint) of the Y base point is obtained by digital curve matching of local segments before and after the X and Y base point data sequences.

[0010] S4. Based on the optimal calibration point, correct the mileage error of the data to be repaired between adjacent base points, expressed as follows:

[0011]

[0012] In the formula, node (i), node (i+1) represents the mileage of the two adjacent endpoints of Y, and nodeX(i) and nodeX(i+1) represent the distances of node... (i), node The mileage of endpoint X corresponding to (i+1), where N is... The number of all sampling points between adjacent endpoints, node(n) is the node (i) and node The mileage of sampling point n between (i+1) points, where n is the sampling point number;

[0013] S5, according to The accuracy of the mileage deviation correction and the changes in the waveform before and after the correction are used to comprehensively evaluate the correction results.

[0014] The beneficial effects of this invention are as follows: Based on the baseline data or reference data of the track geometry state—baseline data X, the track geometry state data to be repaired—is precisely corrected for mileage error, thereby obtaining higher precision mileage information. At the same time, the original data characteristics of the track detection data, such as the orientation of the left and right rails, elevation, gauge, level, twist, and curvature, are preserved. Furthermore, the quality of mileage correction is evaluated using scientifically accurate statistical indicators, so as to more accurately grasp the true geometric state of the track. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the adaptive mileage correction and evaluation method for multi-channel track inspection data of the present invention.

[0016] Figure 2 This is the ultra-high data characteristic curve diagram of the present invention;

[0017] Figure 3 This is a curve diagram illustrating the curvature characteristics of this invention;

[0018] Figure 4 This is a characteristic curve diagram of the ground marker (ALD) of the present invention;

[0019] Figure 5 This is a schematic diagram illustrating the search for the optimal correction point in this invention;

[0020] Figure 6 This is a schematic diagram of the adaptive parameter tuning of the present invention;

[0021] Figure 7This is a schematic diagram of the invention for finding new base points using segment length as the step size;

[0022] Figure 8 This is a schematic diagram of the second correction of mileage error in Embodiment 2 of the present invention;

[0023] Figure 9 It is the adjacent two-endpoint segment of the reference data and the first correction data of this invention;

[0024] Figure 10 This is a schematic diagram illustrating the segment boundary movement of the present invention;

[0025] Figure 11 This is a schematic diagram of the second correction of mileage error in Embodiment 3 of the present invention. Detailed Implementation

[0026] 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 with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0027] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0028] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0029] In the description of this invention, it should be understood that the terms "upper," "lower," "inner," "outer," "left," "right," "front," and "rear," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use, or the orientation or positional relationship commonly understood by those skilled in the art. They are only used to facilitate the description of this invention and to simplify the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0030] Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0031] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, terms such as "set" and "connection" should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0032] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0033] Example 1:

[0034] Adaptive mileage correction and evaluation method for multi-channel track inspection data, such as Figure 1 As shown, it includes:

[0035] S1. Obtain the baseline data X and the data to be repaired Y of the track.

[0036] S0. Determine whether the baseline data X and the data to be repaired Y are from the same source. If they are, proceed directly to S2. Otherwise, preprocess the baseline data to make it from the same source as the data to be repaired before proceeding to S2. When both the baseline data X and the data to be repaired are dynamic inspection data or both are static inspection data, the baseline data and the data to be repaired are from the same source.

[0037] Preprocessing involves mileage correction between dynamic and static, or static and dynamic detection data. This allows for the direct use of dynamic and static detection equipment output data from inertial measurement references, or the use of high-pass and low-pass filters to preprocess the dynamic and static detection data.

[0038] S2. Extract specific points from the baseline data and the data to be repaired as base points. The base point mileage of the baseline data is... The baseline mileage of the data to be repaired is denoted as ,like Figure 2 , Figure 3 and Figure 4 As shown.

[0039] S3. Obtain the optimal correction point (referred to as the endpoint) by using digital curve matching of local segments, such as... Figure 5 As shown; specifically including:

[0040] S31. Taking the base points of the channel data such as the elevation, direction, gauge, superelevation and triangular pit of the left and right rails in Y as the center, extend L / 2 along the data sequence before and after each as the base point local matching segment, and take L before and after the base point of the corresponding channel data in X as the matching search area.

[0041] S32. Calculate the curve fit between the local matching segment of the base point and the matching search area. To determine the correction point for the Y-base point, perform curve matching and locate the position where the sum of the curve fit measurement function values ​​reaches its maximum.

[0042]

[0043]

[0044]

[0045] In the formula, i is the index of the Y base point, nodeY(i) is the mileage of the i-th base point in Y, the corresponding mileage of the X base point is nodeX(i), and L is the matching length. and These are the mileages of the first and last points of the local matching segment of the Y base point. and denoted as the mileages of the first and last points of the X segment for calculating the curve fit with the matching segment, and S as the sliding distance over the matching segment search area. Let be the curve matching degree measure function for the local matching segment of the base points i in X and Y. This represents the sum of curve matching function values ​​for the left elevation, right elevation, left track orientation, right track orientation, track gauge, superelevation, and triangular pit channel data of the X and Y track inspection data in the local matching section of the base point i. To determine the sliding distance S required to achieve the highest curve fit within the matching search area;

[0046] S33. Determine the curve matching degree Is it less than the preset curve fit threshold? If the condition is met, skip the local matching section of base point i, do not use it as the correction point for mileage error, and return to S31 to calculate the multi-channel curve matching degree of the next base point i+1; otherwise, proceed to S34.

[0047] S34. Take the midpoint of the sliding distance S as the correction point of the Y base point, and denot its mileage as... The mileage of the endpoint is .

[0048] S4. Based on the optimal calibration point, perform mileage error correction on the data to be corrected between adjacent base points to obtain the first corrected data. , represented as

[0049]

[0050] In the formula, node (i), node (i+1) represents the mileage of two adjacent correction points of Y, and nodeX(i) and nodeX(i+1) represent the distances of node... (i), node The mileage of endpoint X corresponding to (i+1), where N is... The number of all sampling points between two adjacent correction points, node(n) is the node (i) and node The mileage of sampling point n between (i+1) points, where n is the sampling point number;

[0051] S4', Perform a second mileage correction on the first corrected data to obtain the second corrected data; specifically including:

[0052] S4'1. Determine whether to perform adaptive parameter tuning. If yes, input the range of values ​​for the segment length L and the movable quantity s of the segment boundary, and the reference data and first correction data between the two endpoints. Perform adaptive parameter tuning to obtain the optimal segment length L and segment boundary displacement s; otherwise, manually input the segment length L and segment boundary displacement s, such as... Figure 6 As shown; specifically:

[0053] A. Using the horizontal axis as the matching length variable, the vertical axis as the boundary movement variable, row as the number of rows, and col as the number of columns, arrange the parameters sequentially in the grid. Each intersection point in the grid represents a set of parameter combinations (L). col ,s row );

[0054] B. Calculate the matching length L in seconds. max =d×row max， When d represents the long interval between sampling points, the first correction data... The sum of the curve fit measurement function values ​​of the m channels of data from the sub-segment and the reference data X sub-segment. ,in, , .

[0055] C. Search results for s max The sum of the similarity measurement function values ​​in this row Which column (col) corresponds to the maximum matching length L?

[0056] D. In this column, the fixed matching length L remains unchanged, i.e., L=L col This causes the row value to change from s. max Starting from the largest to the smallest, calculate the parameter combinations (L) one by one. col ,s row The sum of the values ​​of the fit measurement function under ) If the latter combination (L) col ,s row-1The sum of the similarity measure function values ​​of (L) is compared with the previous combination (L) col ,s row If the sum of the similarity measurement function values ​​is small, the calculation stops. col ,s row This is the optimal parameter combination for the second odometer correction;

[0057] S4'2, Read the baseline data X between the two endpoints nodeX(i) and nodeX(i+1). and Data between two endpoints ,in ;

[0058] S4'3, Adjust the first correction data according to the segment length L. A second mileage correction was performed to obtain the second corrected data; specifically:

[0059] (1) Judgment and If the length between the two endpoints is greater than the segment length L, proceed to (2); otherwise, use it as the second correction data. Enter S5;

[0060] (2) From Start by following a step size kL (k∈the set of positive integers) at mileage and Add a new base point between ,like Figure 7 As shown, it is represented as

[0061] ;

[0062] (3) Order As a new , and return (4);

[0063] (4) Perform steps S3 and S4 to perform a second mileage correction and obtain the second correction data. Then return to (1).

[0064] S5. A comprehensive evaluation of the correction results is conducted based on the accuracy of the mileage deviation correction for the data to be repaired and the changes in the waveform before and after the correction. This includes...

[0065] Precision is expressed as:

[0066] ;

[0067] In the formula, X is the baseline data, x i Let X be the amplitude of the irregularity at the i-th sampling point of the X sequence. This is the second corrected data. for The magnitude of the irregularity at the i-th sampling point in the sequence, [X; X and Arrange them in rows to form a 2-row matrix, with the first row being X and the second row being X. , For x i 2 and The summation is defined by SVD, where SVD represents the singular value decomposition process, and SUM represents the summation of the eigenvalues ​​obtained from the singular value decomposition. The precision factor ranges from [0, 1].

[0068] The waveform changes are represented as follows;

[0069] ;

[0070] In the formula, Y represents the data to be repaired. For the second corrected data, ||Y|| and || ||Y and respectively The square root of the sum of the absolute values ​​of the squares, i.e., the Euclidean norm, The minimum value between u and v;

[0071] The overall evaluation is expressed as follows:

[0072]

[0073] The comprehensive factor ranges from [0, 2], and the larger the value, the more ideal the mileage alignment effect of the two sets of data.

[0074] Example 2:

[0075] Adaptive mileage correction and evaluation method for multi-channel track inspection data, such as Figure 1 As shown, it includes:

[0076] S1. Obtain the baseline data X and the data to be repaired Y of the track.

[0077] S0. Determine whether the baseline data X and the data to be repaired Y are from the same source. If they are, proceed directly to S2. Otherwise, preprocess the baseline data to make it from the same source as the data to be repaired before proceeding to S2. When both the baseline data X and the data to be repaired are dynamic inspection data or both are static inspection data, the baseline data and the data to be repaired are from the same source.

[0078] Preprocessing involves mileage correction between dynamic and static, or static and dynamic detection data. This allows for the direct use of dynamic and static detection equipment output data from inertial measurement references, or the use of high-pass and low-pass filters to preprocess the dynamic and static detection data.

[0079] S2. Extract specific points from the baseline data and the data to be repaired as base points. The base point mileage of the baseline data is... The baseline mileage of the data to be repaired is denoted as ,like Figure 2 , Figure 3 and Figure 4 As shown.

[0080] S3. Obtain the optimal correction point (referred to as the endpoint) by using digital curve matching of local segments, such as... Figure 5 As shown; specifically including:

[0081] S31. Taking the base points of the channel data such as the elevation, direction, gauge, superelevation and triangular pit of the left and right rails of Y as the center, extend L / 2 along the data sequence before and after each as the base point local matching segment, and take L before and after the base point of the corresponding channel data in X as the matching search area.

[0082] S32. Calculate the curve fit between the local matching segment at the base point and the matching search area. Perform curve matching, find the location of the maximum curve fit, and determine the correction point of the Y-base point:

[0083]

[0084]

[0085]

[0086] In the formula, i is the index of the Y base point, nodeY(i) is the mileage of the i-th base point in Y, the corresponding mileage of the X base point is nodeX(i), and L is the matching length. and These are the mileages of the first and last points of the local matching segment of the Y base point. and denoted as the mileages of the first and last points of the X segment for calculating the curve fit with the matching segment, and S as the sliding distance over the matching segment search area. Let be the curve matching degree measure function for the local matching segment of the base points i in X and Y. This represents the sum of curve matching function values ​​for the left elevation, right elevation, left track orientation, right track orientation, track gauge, superelevation, and triangular pit channel data of the X and Y track inspection data in the local matching section of the base point i. To determine the sliding distance S required to achieve the highest curve fit within the matching search area;

[0087] S33. Determine the curve matching degree Is it less than the preset curve fit threshold? If the condition is met, skip the local matching section of base point i, do not use it as the correction point for mileage error, and return to S31 to calculate the multi-channel curve matching degree of the next base point i+1; otherwise, proceed to S34.

[0088] S34. Take the midpoint of the sliding distance S as the correction point of the data nodeY(i) to be repaired, and denot its mileage as... The mileage of the endpoint is .

[0089] S4. Based on the optimal calibration point, perform mileage error correction on the data to be corrected between adjacent base points to obtain the first corrected data Y', denoted as...

[0090] ,

[0091] In the formula, node (i), node (i+1) represents the mileage of two adjacent correction points of Y, and nodeX(i) and nodeX(i+1) represent the distances of node... (i), node The mileage of endpoint X corresponding to (i+1), where N is... The number of all sampling points between two adjacent correction points, node(n) is the node (i) and node The mileage of sampling point n between (i+1) points, where n is the sampling point number;

[0092] S4', Perform a second mileage correction on the first corrected data to obtain the second corrected data; specifically including:

[0093] S4'1. Determine whether to perform adaptive parameter tuning. If yes, input the range of values ​​for the segment length L and the movable quantity s of the segment boundary, and the reference data and first correction data between the two endpoints. Perform adaptive parameter tuning to obtain the optimal segment length L and segment boundary displacement s; otherwise, manually input the segment length L and segment boundary displacement s, such as... Figure 6 As shown; specifically:

[0094] A. Using the horizontal axis as the matching length variable, the vertical axis as the boundary movement variable, row as the number of rows, and col as the number of columns, arrange the parameters sequentially in the grid. Each intersection point in the grid represents a set of parameter combinations (L). col ,s row );

[0095] B. Calculate the matching length L in seconds. max =d×row max， When d represents the long interval between sampling points, the first correction data... Sub-segments and baseline data The sum of the curve fit measurement function values ​​of m channels of data in a sub-segment in, , ;

[0096] C. Search results for s max The sum of the similarity measurement function values ​​in this row Which column (col) corresponds to the maximum matching length L?

[0097] In this column, the fixed matching length L remains unchanged, i.e., L=L col This causes the row value to change from s. max Starting from the largest to the smallest, calculate the parameter combinations (L) one by one. col ,s row The sum of the values ​​of the fit measurement function under ) If the latter combination (L) col ,s row-1 The sum of the similarity measure function values ​​of (L) is compared with the previous combination (L) col ,s row If the sum of the similarity measurement function values ​​is small, the calculation stops. col ,s row This is the optimal parameter combination for the second mileage correction.

[0098] S4'2, Read the baseline data X between the two endpoints nodeX(i) and nodeX(i+1). and First correction data between the two endpoints ,in ;

[0099] S4'3, Adjust the first correction data according to the segment length L. A second mileage correction was performed to obtain the second corrected data; specifically:

[0100] ①.Judgment and Is the length between the two endpoints greater than the segment length L? If yes, proceed to step ②; otherwise, proceed to step ③. The data in the corresponding segment is used as the second correction data. And enter S5;

[0101] ② Based on the segment length L, read the baseline data X and the first correction data. Divided into several sub-segments;

[0102] ③ Find the benchmark data sub-segment with the largest sum of curve fit measurement function values ​​among numerous sub-segments, and use the center of this sub-segment as the new odometer correction point nodeX(i). k ,

[0103] ;

[0104] In the formula, nodeX(i) k To add new mileage correction points, nodeX(i) and nodeX(i+1) are the baseline data endpoints, respectively, x k and The starting and ending mileages a, b, c, and e are the mileages of the kth sub-segment boundary, n is the total number of segments divided by the baseline data, and L is the segment length. The sub-segment that maximizes the sum of the curve fit measurement function values ​​across multiple sub-segments;

[0105] ④ The center of the benchmark data sub-segment with the largest sum of curve fit function values ​​is taken as the newly added mileage correction point nodeX(i). k And store them in the original correction point sequence nodeX according to the mileage size and renumber them;

[0106] ⑤ Based on the newly added mileage correction points, adjust the first correction data. Perform a second mileage correction, such as Figure 8 As shown;

[0107] ⑥ Replace the newly added mileage correction point with the new endpoint, and let As a new , and return ①.

[0108] S5. A comprehensive evaluation of the correction results is conducted based on the accuracy of the mileage deviation correction for the data to be repaired and the changes in the waveform before and after the correction. This includes...

[0109] Precision is expressed as:

[0110]

[0111] In the formula, X is the baseline data, x i Let X be the amplitude of the irregularity at the i-th sampling point of the X sequence. This is data from the second revision. for The magnitude of the irregularity at the i-th sampling point in the sequence, [X; X and Arrange the rows to form a 2-row matrix, with the first row being X and the second row being X. , For x i 2 and The summation is defined by SVD, where SVD represents the singular value decomposition process, and SUM represents the summation of the eigenvalues ​​obtained from the singular value decomposition. The precision factor ranges from [0, 1].

[0112] The waveform changes are represented as follows;

[0113]

[0114] In the formula, Y represents the data to be repaired. For the second corrected data, ||Y|| and || ||Y and respectively The square root of the sum of the absolute values ​​of the squares, i.e., the Euclidean norm, The minimum value between u and v;

[0115] The overall evaluation is expressed as follows:

[0116]

[0117] The comprehensive factor ranges from [0, 2], and the larger the value, the more ideal the mileage alignment effect of the two sets of data.

[0118] Example 3

[0119] Adaptive mileage correction and evaluation methods for multi-channel track inspection data, including:

[0120] S1. Obtain the baseline data X and the data to be repaired Y of the track.

[0121] S0. Determine whether the baseline data X and the data to be repaired Y are from the same source. If they are, proceed directly to S2. Otherwise, preprocess the baseline data to make it from the same source as the data to be repaired before proceeding to S2. When both the baseline data X and the data to be repaired are dynamic inspection data or both are static inspection data, the baseline data and the data to be repaired are from the same source.

[0122] Preprocessing involves mileage correction between dynamic and static, or static and dynamic detection data. This allows for the direct use of dynamic and static detection equipment output data from inertial measurement references, or the use of high-pass and low-pass filters to preprocess the dynamic and static detection data.

[0123] S2. Extract specific points from the baseline data and the data to be repaired as base points. The base point mileage of the baseline data is... The baseline mileage of the data to be repaired is denoted as .

[0124] S3. Obtain the optimal correction point (referred to as the endpoint) by using digital curve matching of local segments; specifically including:

[0125] S31. Taking the base points of the channel data such as the elevation, direction, gauge, superelevation and triangular pit of the left and right rails in Y as the center, extend L / 2 along the data sequence before and after each as the base point local matching segment, and take L before and after the base point of the corresponding channel data in X as the matching search area.

[0126] S32. Calculate the curve fit between the local matching segment at the base point and the matching search area. Perform curve matching, find the location of the maximum curve fit, and determine the correction point of the Y-base point:

[0127]

[0128]

[0129]

[0130] In the formula, i is the index of the Y base point, nodeY(i) is the mileage of the i-th base point in Y, the corresponding mileage of the X base point is nodeX(i), and L is the matching length. and These are the mileages of the first and last points of the local matching segment of the Y base point. and denoted as the mileages of the first and last points of the X segment for calculating the curve fit with the matching segment, and S as the sliding distance over the matching segment search area. Let be the curve matching degree measure function for the local matching segment of the base points i in X and Y. This represents the sum of curve matching function values ​​for the left elevation, right elevation, left track orientation, right track orientation, track gauge, superelevation, and triangular pit channel data of the X and Y track inspection data in the local matching section of the base point i. To determine the sliding distance S required to achieve the highest curve fit within the matching search area;

[0131] S33. Determine the curve matching degree Is it less than the preset curve fit threshold? If the condition is met, skip the local matching section of base point i, do not use it as the correction point for mileage error, and return to S31 to calculate the multi-channel curve matching degree of the next base point i+1; otherwise, proceed to S34.

[0132] S34. Take the midpoint of the sliding distance S as the correction point of the Y base point, and denot its mileage as... The mileage of the endpoint is .

[0133] S4. Based on the optimal calibration point, perform mileage error correction on the data to be corrected between adjacent base points to obtain the first corrected data. , represented as

[0134]

[0135] In the formula, node (i), node (i+1) represents the mileage of two adjacent correction points of Y, and nodeX(i) and nodeX(i+1) represent the distances of node... (i), node The mileage of endpoint X corresponding to (i+1), where N is... The number of all sampling points between two adjacent correction points, node(n) is the node (i) and node The mileage of sampling point n between (i+1) points, where n is the sampling point number;

[0136] S4', Perform a second mileage correction on the first corrected data to obtain the second corrected data; specifically including:

[0137] S4'1. Determine whether to perform adaptive parameter tuning. If yes, input the range of values ​​for the segment length L and the movable quantity s of the segment boundary, and the reference data and first correction data between the two endpoints. Perform adaptive parameter tuning to obtain the optimal segment length L and segment boundary displacement s; otherwise, manually input the segment length L and segment boundary displacement s. Specifically:

[0138] A. Using the horizontal axis as the matching length variable, the vertical axis as the boundary movement variable, row as the number of rows, and col as the number of columns, arrange the parameters sequentially in the grid. Each intersection point in the grid represents a set of parameter combinations (L). col ,s row );

[0139] B. Calculate the matching length L in seconds. max =d×row max， When d is the sampling interval, the first correction data... The sum of the curve fit measurement function values ​​of the m channels of data from the sub-segment and the reference data X sub-segment. ,in, , .

[0140] C. Search results for s max The sum of the similarity measurement function values ​​in this row Which column (col) corresponds to the maximum matching length L?

[0141] D. In this column, the fixed matching length L remains unchanged, i.e., L=L col This causes the row value to change from s. max Starting from the largest to the smallest, calculate the parameter combinations (L) one by one. col ,s row The sum of the values ​​of the fit measurement function under ) If the latter combination (L) col ,s row-1The sum of the similarity measure function values ​​of (L) is compared with the previous combination (L) col ,s row If the sum of the similarity measurement function values ​​is small, the calculation stops. col ,s row This is the optimal parameter combination for the second odometer correction;

[0142] S4'2, Read the baseline data X between the two endpoints nodeX(i) and nodeX(i+1). and Data between two endpoints ,in ;

[0143] S4'3, Adjust the first correction data according to the segment length L and the segment boundary mobility s. A second mileage correction was performed to obtain the second corrected data; specifically:

[0144] 1) Based on the segment length L, combine the baseline data X and the first correction data between the two endpoints. The area is divided into segments, and each segment is called a sub-segment, such as... Figure 9 As shown;

[0145] 2) Move the first correction data forward and backward along the data column sequence. The segmentation boundaries make the first corrected data The number of sampling points in the sub-segment changes, such as Figure 10 As shown;

[0146] 3) Determine the first correction data after the segment boundary shift. Check if the data volume of the sub-segment and the baseline data X sub-segment are consistent. If they are consistent, proceed directly to step 5; otherwise, proceed to step 4.

[0147] 4) For the first correction data The sub-segment undergoes interpolation resampling to match the number of sampling points in the corresponding sub-segment of the baseline data X, and then proceeds to step 5); interpolation resampling is represented as... Among them, New First correction data The sequence obtained after interpolation and resampling of the sub-segments, where interp() is the interpolation function. First correction data The mileage of the sub-segment First correction data The amplitude of track irregularity in the sub-segment, where num is the number of sampling points;

[0148] 5) After each sampling interval moves the segment boundary, the starting and ending mileage of the sub-segment is adjusted by d, and the first corrected data after the segment boundary movement is calculated. The sum of the curve fit measurement function values ​​of the m channels of data from the sub-segment and the reference data X sub-segment. ,like Figure 11 As shown. Figure 11 In the case of k = -1, 0, 1, X and Endpoint i matches the start and end points of the segment. a = nodeX(i) - L / 2, b = nodeX(i) + L / 2 + kd, c = node (i)-L / 2, e=node The sum of curve matching degrees calculated by (i)+L / 2 is used to select the maximum value, record the k value at the maximum value, and update the segment boundary position of the sub-segment.

[0149] 6) Find out how many sampling points k the segment boundary moved when the sum of the curve fit measurement function values ​​reaches its maximum value, and determine the segment boundary position of that sub-segment. Update the value and record the k value;

[0150] 7) Based on the new segment boundary positions, adjust the first correction data. Perform steps S3 and S4 to perform a second mileage correction and obtain the second corrected data. .

[0151] S5. A comprehensive evaluation of the correction results is conducted based on the accuracy of the mileage deviation correction for the data to be repaired and the changes in the waveform before and after the correction. This includes...

[0152] Precision is expressed as:

[0153]

[0154] In the formula, X is the baseline data, x i Let X be the amplitude of the irregularity at the i-th sampling point of the X sequence. This is data from the second revision. for The magnitude of the irregularity at the i-th sampling point in the sequence, [X; X and Arrange the rows to form a 2-row matrix, with the first row being X and the second row being X. , For x i 2 and The summation is defined by SVD, where SVD represents the singular value decomposition process, and SUM represents the summation of the eigenvalues ​​obtained from the singular value decomposition. The precision factor ranges from [0, 1].

[0155] The waveform changes are represented as follows;

[0156]

[0157] In the formula, Y represents the data to be repaired. For the second corrected data, ||Y|| and || ||Y and respectively The square root of the sum of the absolute values ​​of the squares, i.e., the Euclidean norm, The minimum value between u and v;

[0158] The overall evaluation is expressed as follows:

[0159]

[0160] The comprehensive factor ranges from [0, 2], and the larger the value, the more ideal the mileage alignment effect of the two sets of data.

[0161] The technical solutions of the present invention are not limited to the specific embodiments described above. Any technical modifications made in accordance with the technical solutions of the present invention fall within the protection scope of the present invention.

Claims

1. A method for multi-lane track data adaptive milepost correction and evaluation, characterized in that, Comprise: S1, obtain the datum X and the data Y to be repaired of track reference; S2, extract special point in X and Y as base point, X base point mileage is recorded as , Y base point mileage is recorded as ; S3, adopt the digital curve matching of local section to obtain the best correction point position of Y base point along X, Y base point data sequence before and after; Specifically includes: S31, with the high, track, track, super high and triangular pit channel data of Y left and right track base point as the center, extend L / 2 along the data sequence before and after, and mark as the base point local matching section, and take the base point before and after L of the corresponding channel data in X as the matching search area; S32, calculate the curve goodness of fit of both the base point local matching section and the matching search region , perform curve matching to find the position of the maximum curve goodness of fit to obtain the correction point of the Y base point: ； ； ； where i is the index of Y base point, nodeY(i) is the mileage of the i th Y base point, and the corresponding X base point mileage is nodeX(i), L is the matching length, and are the mileages of the start and end points of the local matching section of the Y base point, respectively, and are the mileages of the start and end points of the X section for curve fitting degree calculation, respectively, and S is the sliding distance on the matching section search area, is the curve fitting degree measurement function of the local matching section of the X and Y base points i, represents the sum of the curve fitting degree measurement function values of the left high-low, right high-low, left track direction, right track direction, track gauge, super-elevation, and triangular pit channel data of the X and Y track inspection data in the local matching section of the base point i, is the sliding distance S when the curve fitting degree in the matching search area is the highest. S33, judging the curve fitting degree whether less than a preset curve fitting degree threshold If yes, skip the local matching section of the base point i, do not take it as a correction point of the mileage error, and return to S31 to calculate the multi-channel curve fitting degree of the next base point i+1. Otherwise, enter S34. S34, the midpoint of the sliding distance S is taken as the correction point of Y base point, and its mileage is recorded as node (i), the mileage value is ; S4, according to the best correction point, the Y between adjacent base points is corrected, and is expressed as ； wherein node (i), node (i+1) are the mileages of the two adjacent X endpoints of Y, and nodeX(i), nodeX(i+1) are the X endpoints of Y, respectively (i), node (i+1) are the mileages of the two adjacent X endpoints of Y, and nodeX(i), nodeX(i+1) are the X endpoints of Y, respectively N is the number of all sampling points between the two adjacent endpoints, and node(n) is the endpoint node (i), and node (i+1) are the mileages of the two adjacent X endpoints of Y, and nodeX(i), nodeX(i+1) are the X endpoints of Y, respectively S5、According to The precision of the mileage deviation correction and the change of the waveform before and after the correction are comprehensively evaluated.

2. The method of claim 1, wherein, Between S1 and S2, S0 is also included, that is, whether the datum X and the data Y to be repaired are homologous data, if so, directly enter S2, otherwise, pre-process X, so that X and Y after processing become homologous data and then enter S2.

3. The method of claim 1, wherein, S4, the first correction data is obtained by correcting the mileage error Between S4 and S5, S4' is further included, the second correction data is obtained by correcting the first correction data .

4. The method of claim 3, wherein, In S4', it includes: S4'1, judging whether to perform adaptive parameter adjustment, if yes, inputting the value range of the segment length L and the segment boundary movable amount s, and obtaining the first correction data corresponding to the reference data between the two endpoints performing adaptive parameter adjustment to obtain the optimal segment length L and the segment boundary movable amount s; otherwise, manually inputting the segment length L and the segment boundary movable amount s; S4'2, read reference data X between the two endpoints nodeX(i) and nodeX(i+1), node (i) and node (i+1) two endpoints first correction data wherein ; S4'3, the first correction data according to the segment length L and the segment boundary shift s performing a second distance correction to obtain second correction data .

5. The method of claim 4, wherein, In S4'1, adaptive parameter adjustment is carried out, specifically including: A, with the horizontal axis as the matching length variable, the vertical axis as the segmented boundary movable variable, row as the number of rows, col as the number of columns, arranging the parameters in the grid in turn, and each intersection point in the grid being a set of parameter combinations (L col ,s row ); B, the matching length L is calculated for each max = d x row max， d is the sampling point interval length, the first correction data The sum of the curve fitting degree function values of the m channel data of the sub-section and the reference data X sub-section wherein: , ; C, search out s max The sum of the values of the similarity measure function in this row The maximum matching length L is which column col; D. In this column, the fixed matching length L is unchanged, i.e. L = L col , the row value s max is sequentially decreased from large to small, and the sum of the fitness measure function values under the parameter combination (L col , s row ) is calculated respectively ; if the sum of the fitness measure function values under the next combination (L col , s row-1 ) is smaller than that under the previous combination (L col , s row ), the calculation is stopped, and (L col , s row ) is the optimal parameter combination for the second mileage correction.

6. The method of claim 5, wherein, In S4'3, the second mileage correction is carried out, specifically including: (1), judge node (i) and node (i+1) two endpoints between the length is greater than the length of the segment L, if, then enter (2), otherwise, as the second correction data , enter S5; (2) from node (i) start increasing new base points node (i) and node (i+1) by step kL, k e set of positive integers, in the distance node (i) k , is expressed as ; (3), let node (i) k as the new node (i), go to (4) ; (4) performing steps S3 and S4 to obtain second correction data Return to (1).

7. The method of claim 5, wherein, In S4'3, the second mileage correction is carried out, specifically including: ①, judge whether the length between the two end points of nodeX(i) and nodeX(i+1) is greater than the segment length L, if yes, go to ②, otherwise, go to The corresponding segment data is as the second correction data and go to S5; ②、according to the segment length L, the read reference data X and the first correction data are divided into several sub-segments; ③、 search the center of the reference data sub-section with the maximum sum of the curve goodness-of-fit function value in the numerous sub-sections as the new milepost correction point nodeX(i) k , ； wherein nodeX(i) k For the newly added mileage correction point, nodeX(i) and nodeX(i+1) are the endpoints of the reference data, respectively, x k For the kth sub-section of the reference data, For the kth sub-section of the first modified data, x k and The start and end points a, b, c, and e of the kth sub-section are the demarcation point mileages of the kth sub-section, n is the total number of sections divided by the reference data, and L is the section length. The kth sub-section is the sub-section with the maximum sum of the curve fitting degree measurement function values of the plurality of sub-sections. IV. The center of the reference data sub-section with the maximum sum of the curve goodness of fit function values is taken as the newly added mileage correction point nodeX(i) k and stored in the original correction point sequence nodeX in order of the mileage size and renumbered; V. According to the newly added mileage correction point, the first correction data is subjected to a second mileage correction. ​ ⑥ Replace the newly added odometer correction point with the new endpoint, and set node... (i) k As a new node (i), and return ①.

8. The method of claim 5, wherein, In S4'3, the second mileage correction is carried out, specifically including: 1) According to the segment length L, the reference data X and the first correction data Segment division is performed, and each of the divided segments is denoted as a subsegment. 2) move the first correction data along the data column order forward and backward across the segment boundary, so that the number of sampling points of the first correction data sub-segment changes; 2) move the first correction data along the data column order forward and backward across the segment boundary, so that the number of sampling points of the first correction data sub-segment changes; 2) move the first correction data along the data column order forward and backward across 3) judging the first modified data after moving the segment boundary whether the data amount of the subsegment and the reference data X subsegment is consistent, if consistent, directly entering 5), otherwise entering 4); 4) For the first correction data The sub-segment undergoes interpolation resampling to match the number of sampling points in the corresponding sub-segment of the baseline data X, and then proceeds to step 5); interpolation resampling is represented as... Among them, New First correction data The sequence obtained after interpolation and resampling of the sub-segment, where interp() is the interpolation function, l i First correction data The mileage of the sub-segment First correction data The amplitude of track irregularity in the sub-segment, where num is the number of sampling points; 5) After the segment boundary moves one sampling interval, the start and end points of the subsegment are adjusted by d, and the first correction data after the segment boundary moves is calculated The sum of the values of the curve fitting degree measurement function of the m channel data of the subsegment and the reference data X 6) search the maximum value of the sum of the curve goodness-of-fit function, and how many sampling points k the segment boundary moves, record the segment boundary position of the subsegment update and record the k value; 7), based on the new segment boundary position, to the first correction data Steps S3 and S4 are performed to obtain second correction data .

9. The method of claim 1, wherein, In S5, The precision is expressed as: ； wherein X is reference data, x i is the uneven amplitude of the i-th sampling point of the X sequence, is the second correction data, is the first correction data, is the uneven amplitude of the i-th sampling point of the X sequence, X and x are arranged in rows to form a 2-row matrix, the first row being X and the second row being x , is x i 2 and is the cumulative sum of x and x, svd is a singular value decomposition process, and sum indicates summing the characteristic values obtained by singular value decomposition; the precision factor has a value range of [0, 1]. The waveform change condition is expressed as: The waveform change condition is expressed as: ； where Y is the data to be corrected, is the second correction data, ||Y|| and || are the absolute value square sums of Y and respectively, and is the minimum value of u and v. The comprehensive evaluation is expressed as: .

Images