A train fusion positioning method based on multi-dimensional data

Abstract

A train fusion positioning method based on multi-dimensional data relates to the technical field of high-speed railway detection, comprising obtaining basic data formed according to manually input line data, and multi-dimensional data for positioning poles; after receiving a trigger signal, generating corresponding expected poles for each dimension in the multi-dimensional data, and each expected pole is assigned a corresponding original weight; according to the results of each expected pole, adding the original weights corresponding to the expected poles with the same result to obtain a comprehensive weight; according to whether the comprehensive weight exceeds a threshold, calibrating each expected pole as a precise positioning pole or a preliminary positioning pole; using basic data and precise positioning poles to obtain actual poles of the line, matching the preliminary positioning poles according to the actual poles of the line; correcting the preliminary positioning poles to the actual poles of the line that have been matched successfully, and generating one-pole-one-file data in combination with the precise positioning poles; the problem of low train positioning accuracy can be solved.

Description

Technical Field

[0001] The present invention relates to the technical field of high-speed railway detection, and in particular to a train fusion positioning method based on multi-dimensional data. Background Art

[0002] With the improvement of railway operation management and technological progress, the overhead contact network suspension status detection and monitoring device has become an important part of the electric railway power supply safety detection and monitoring system. Based on artificial intelligence, machine vision, high-definition imaging and other advanced technologies, the device uses an integrated and platform-based design to perform imaging detection on the components of the overhead contact network and measure the static geometric parameters of the overhead contact network. Through the identification and analysis of the detection data, maintenance suggestions are formed to guide the elimination of hidden dangers of overhead contact network failures.

[0003] When the overhead wire suspension state detection and monitoring device files the data one pole at a time in accordance with the requirements of TJ / GD006-2014 "Interim Technical Conditions for Overhead Wire Suspension State Detection and Monitoring Device (4C)", it is necessary to first locate the location of the detection vehicle, that is, the pole interval where the trigger imaging is generated. The traditional positioning method is to pre-enter the basic data of the line to be detected (such as pole number, station interval, kilometer mark, tunnel, mileage, etc.), select the line and the starting pole number during the detection, and calculate the mileage of the detection vehicle based on the accumulated calculation during the detection process. The mileage is searched from the basic information of the line to obtain the most matching pole interval. This method is easily affected by inaccurate data entry of the detection line and cumulative errors in mileage, resulting in continuous deviations in the located pole interval, and ultimately causing the generated one-pole-one-file data to not match the real line data.

[0004] In addition, during the application of the current contact network suspension status detection and monitoring device, due to the errors in the manually entered line basic data and the cumulative errors of the axle encoder installed on the wheel, the accurate positioning of the detection vehicle is affected, resulting in errors in the archiving of detection data, affecting the analysis and confirmation of subsequent detection data.

[0005] The Chinese invention patent document with the publication number CN114519116A discloses an image acquisition and archiving method and system for detecting the hanging state of the contact network. The invention provides a method for obtaining the corresponding pole positioning information from the line basic data according to the trigger signal, and then storing the contact network hanging image in the corresponding front, back or hanging string image folder according to the corresponding information, and storing the corresponding information in a primary archiving database; then normalizing and analyzing each trigger signal, and creating one-pole-one-file folders one by one, and then classifying the image information in the front, back or hanging string image folder corresponding to each trigger into the corresponding one-pole-one-file folder to form one-pole-one-file data;

[0006] However, in the archive database, only image information is used to determine the pole positioning information, and there is a lack of verification means, and the pole positioning information still has the defect of low accuracy. Summary of the invention

[0007] The purpose of the present invention is to provide a train fusion positioning method based on multi-dimensional data, which can solve the problem of low train positioning accuracy.

[0008] The present invention is achieved through the following technical solutions:

[0009] A train fusion positioning method based on multi-dimensional data, comprising:

[0010] S1. Obtaining basic data formed based on manually entered line data and multi-dimensional data for positioning the pole;

[0011] S2. After receiving the trigger signal, a corresponding expected bar is generated for each dimension in the multi-dimensional data, and each expected bar is assigned a corresponding original weight;

[0012] S3. According to the results of each expected pole, the original weights corresponding to the expected poles with the same results are added together to obtain a comprehensive weight;

[0013] S4. According to whether the comprehensive weight exceeds the threshold, each expected pole is marked as a precise positioning pole or a preliminary positioning pole;

[0014] S5. Use basic data and precise positioning poles to obtain actual poles of the line, and match the initial positioning poles with the actual poles of the line;

[0015] S6. Correct the initial positioning pole to the actual pole of the successfully matched line, and generate one-pole-one-file data in combination with the precise positioning pole.

[0016] Furthermore, the basic data includes: line name, line type, station section name, pole number and kilometer mark field.

[0017] Furthermore, the multi-dimensional data includes: current pole information, pole number image recognition results, mileage increment, Beidou positioning data and GYK kilometer mark.

[0018] Furthermore, after receiving the trigger signal in S2, false trigger filtering is performed on the trigger signal.

[0019] Furthermore, the specific steps of the false trigger filtering include: each time a trigger signal is received, the mileage of the current train is synchronously recorded; if the result of subtracting the mileage of the current train from the mileage of the previous train is less than the minimum pole spacing, the current trigger signal is falsely triggered and filtered.

[0020] Furthermore, the specific method of S4 includes:

[0021] S41. Determine whether the expected pole with the highest comprehensive weight is unique;

[0022] S42. If not, then the expected rods are all initial positioning rods;

[0023] S43. If so, determine whether the comprehensive weight is greater than the threshold;

[0024] S44. If the comprehensive weight is greater than the threshold, the corresponding desired pole is calibrated as the precise positioning pole;

[0025] S45. If the comprehensive weight is less than the threshold, the corresponding expected pole is the initial positioning pole.

[0026] Furthermore, the specific process of matching the initial positioning pole according to the actual pole of the line in S5 includes:

[0027] S51. Determine the actual pole of the line to be matched;

[0028] S52. Calculate the actual line pole mileage x according to the precise positioning pole mileage;

[0029] S53. Calculate the mileage of each initial positioning pole;

[0030] S54. According to the formula Calculate and obtain the error range Initial positioning rod inside;

[0031] S55. Update the actual pole data of the line to be matched to the initial positioning pole obtained in S54.

[0032] The technical solution of the present invention has at least the following advantages and beneficial effects:

[0033] The present invention discloses a train fusion positioning method based on multi-dimensional data, which obtains the desired pole through multi-dimensional data, and performs fusion positioning on the desired pole based on weight assignment, which can effectively solve the problem of inaccurate positioning;

[0034] Through positioning correction, the poles with poor positioning conditions are corrected to improve the accuracy of positioning in the entire detection range;

[0035] And according to the process of calibrating the precise positioning rod, the initial positioning rod and the final positioning rod, and data matching, it is possible to generate relatively accurate positioning information when the positioning conditions are unfavorable, generate accurate positioning information when the positioning conditions are more favorable, and correct the positioning information generated when the positioning conditions are unfavorable. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A schematic diagram of a method flow of the present invention;

[0037] Figure 2It is a distribution diagram of the precise positioning rod, the preliminary positioning rod and the actual line rod of the present invention;

[0038] Figure 3 It is a schematic diagram of the matching relationship between the initial positioning pole and the actual pole of the line of the present invention. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0040] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0042] Example 1

[0043] As attached Figure 1 A train fusion positioning method based on multi-dimensional data is shown, comprising:

[0044] S1. Obtain basic data formed according to manually entered line data, and multi-dimensional data for positioning the pole; in particular, the basic data includes: line name, line type, station interval name, pole number and kilometer mark fields, and the basic data is the basis for other data to generate poles; the multi-dimensional data includes: current pole information, pole number image recognition results, mileage increment, Beidou positioning data and GYK kilometer mark; in addition, the weights corresponding to each dimension in the multi-dimensional data, one of which is allocated in the following ways: 15%, 40%, 15%, 25% and 5%; and the weight allocation can be dynamically adjusted according to the line conditions;

[0045] S2. After receiving the trigger signal, a corresponding expected bar is generated for each dimension in the multi-dimensional data, and each expected bar is assigned a corresponding original weight;

[0046] The following is a detailed description of the method of generating each data bar:

[0047] 1) Generate an expected pole based on the current pole: Based on the line, line type, station interval, pole number and other information of the current pole, find the next pole of the current pole from the basic data, and the next pole found is the expected pole.

[0048] 2) Generate the expected pole according to the pole number image recognition result:

[0049] a) The inspection vehicle passes through a support device and receives multiple pole number image recognition results;

[0050] b) removing Chinese characters from each recognition result, and comparing them, and recording the number of repetitions of each recognition result;

[0051] i. When the result with the most repetitions is unique, the expected pole is generated by matching it with the basic data according to the pole number it identifies;

[0052] ii. When the result with the most repetitions is not unique, the expected poles are generated by matching the pole numbers identified from the basic data. In this case, multiple expected poles are generated;

[0053] iii. If the matching from the basic data based on the pole number image recognition result fails, it means that the expected pole generated based on the pole number image recognition result fails.

[0054] 3) Generate expected pole according to mileage increment:

[0055] a) Get the mileage d0 of the last precision positioning rod, and the mileage when the latest trigger signal is generated is d1;

[0056] b) If the difference between d1 and d0 is within a certain range, it is considered that the expected pole can be generated through mileage increment; if it is out of the range, it is considered that the expected pole cannot be generated through mileage increment;

[0057] c) If the difference between d1 and d0 is within a certain range, the positioning pole is matched from the basic data according to the information of the precise positioning pole, such as mileage d0, line, station interval, line type, pole number, etc., combined with the difference between d1 and d0. If pole X is matched, the expected pole is X; if all poles fail to match, the expected pole failure is generated.

[0058] 4) Generate expected pole based on Beidou positioning data:

[0059] a) Obtain the latitude and longitude data of Beidou positioning when the trigger signal is generated;

[0060] b) According to the acquired longitude and latitude data, combined with the route and line type in the mission information, the longitude and latitude of the pole in the basic data are matched. If a match is found, the expected pole is generated; otherwise, the expected pole generation fails.

[0061] 5) Generate expected poles based on GYK kilometer mark:

[0062] a) Obtain the GYK kilometer mark when the trigger signal is generated;

[0063] b) Based on the obtained kilometer mark, combined with the route and type of line in the task information, match it with the kilometer mark of the pole in the basic data. If a match is found, the expected pole is generated; otherwise, the generation of the expected pole fails.

[0064] S3. According to the results of each expected pole, the original weights corresponding to the expected poles with the same results are added together to obtain a comprehensive weight;

[0065] After generating the expected stroke through each data respectively, the expected stroke generated by each data is counted. If the generated expected stroke is the same result, the weights are added together. The following example is shown in Table 1.

[0066] Table 1 Example table of weight generation positioning rod

[0067]

[0068] S4. According to whether the comprehensive weight exceeds the threshold, each expected pole is marked as a precise positioning pole or a preliminary positioning pole;

[0069] In combination with Table 1, the processing flow of generating the positioning rod according to the weight is as follows:

[0070] S41. Determine whether the expected pole with the highest comprehensive weight is unique;

[0071] S42. If not, then the expected rods are all initial positioning rods;

[0072] S43. If so, determine whether the comprehensive weight is greater than the threshold;

[0073] S44. If the comprehensive weight is greater than the threshold, the corresponding desired pole is calibrated as the precise positioning pole;

[0074] S45. If the comprehensive weight is less than the threshold, the corresponding expected pole is the initial positioning pole;

[0075] That is, when the expected pole with the highest weight is unique: determine whether its weight value exceeds the set threshold (the default value is 70%). If it exceeds, it is a precision positioning pole, otherwise it is a preliminary positioning pole. For example, in the above example 1, the pole finally generated is pole a, and the precision positioning pole is pole a; in the above example 2, the pole finally generated is pole b, and the precision positioning pole is pole b;

[0076] When the expected pole with the highest weight is not unique: compare according to the size of the data source weight value, and take the pole with the larger data source weight value as the final pole. At the same time, regardless of the weight value of the final pole, the expected pole can only be the initial positioning pole for subsequent correction. As in Example 3 above, the weights of poles a and b are both 70%, but the data source weight values ​​of pole a are 40%, 15%, and 15% from large to small, and the data source weight values ​​of pole b are 40%, 25%, and 5% from large to small. Since 25%>15%, the pole finally generated is pole b. At the same time, although the weight of pole b is ≥70%, it is still the initial positioning pole; the final state of the pole is shown in Table 2;

[0077] Table 2 Final state of positioning rod

[0078]

[0079] S5. Use basic data and precise positioning poles to obtain actual poles of the line, and match the initial positioning poles with the actual poles of the line;

[0080] The initial positioning rod matching correction is carried out in the following way:

[0081] 1) Pole information correction: Correct the pole number or mileage of the generated initial positioning pole according to the unique identification of the pole;

[0082] 2) Rod fusion: Due to false triggering, multiple rods that do not physically exist may be generated. For these rods, they need to be marked as invalid rods during calibration.

[0083] The overall process of rod calibration based on precise positioning rods is as follows:

[0084] 1) Assume that there are precision positioning rods J0 (mileage m0) and J1 (mileage m1):

[0085] Five initial positioning poles are generated between J0 and J1, namely C0 (mileage t0), C1 (mileage t1), C2 (mileage t2), C3 (mileage t3), and C4 (mileage t4). From the basic data, it is found that there are three pole data between J0 and J1, namely A, B, and C, and the pole spacing is d0, d1, d2, and d3 respectively. The data distribution is as follows: Figure 2 As shown;

[0086] 2) Match the initial positioning rods according to the rod spacing in the basic data;

[0087] a. Determine the actual pole of the line to be matched, for example, the actual pole of the line to be matched is pole C;

[0088] b. According to the mileage of the precise positioning pole, calculate the mileage x of the actual pole of the line, x=m1-d3

[0089] c. Calculate the mileage of each initial positioning pole , ;

[0090] d. According to the formula Calculate and obtain the error range The initial positioning rod within the range; the error range can be set according to the actual situation, such as (0,10%), then the calculation formula is ,Will Substitute in sequence , judge which initial positioning rod's mileage can satisfy the inequality, and then determine the initial positioning rod in e;

[0091] e. Update the actual pole data of the line to be matched to the initial positioning pole obtained in d;

[0092] 3) The process of rod calibration based on the matching results. Figure 2 For example, after the rod data is matched, the distribution of the matched data is as follows Figure 3 Shown

[0093] Rod A in the basic data matches the initial positioning rod C1, rod B matches the initial positioning rod C3, and rod C matches the initial positioning rod C4:

[0094] a) Update the A, B, and C pole data in the basic data to the initial positioning poles C1, C2, and C4;

[0095] Rod fusion was performed for the unmatched initially positioned rods C0 and C3: J0 was fused with C0 based on the J0 rod; C2 was fused with C3 based on the C2 rod.

[0096] S6. Correct the initial positioning pole to the actual pole of the line that matches successfully, and generate one-pole-one-file data in combination with the precise positioning pole. In addition, when replacing and generating one-pole-one-file data, the host computer needs to be notified;

[0097] The one-pole-one-file data obtained by this method has higher accuracy, thus providing accurate positioning data for subsequent train positioning.

[0098] Example 2

[0099] After receiving the trigger signal in S2, the trigger signal is subjected to false trigger filtering; in addition, the specific steps of the false trigger filtering include: each time the trigger signal is received, the mileage of the current train is synchronously recorded, and if the result of subtracting the mileage of the current train from the mileage of the previous train is less than the minimum pole spacing, the current trigger signal is subjected to false trigger filtering;

[0100] It should be noted that in the railway system, although the spacing between adjacent contact networks varies, there is a minimum pole spacing, which is generally 40-70 meters. This is because if the difference in mileage generated when a train receives two trigger signals is less than the minimum pole spacing, it means that the second trigger signal is a false trigger signal, and the signal needs to be filtered out to ensure the accuracy of the trigger signal.

[0101] Example 3

[0102] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable, wherein the processor implements the above-mentioned train fusion positioning method when executing the program.

[0103] A computer-readable storage medium stores a computer program, characterized in that when the program is executed by a processor, the above-mentioned train fusion positioning method is implemented.

[0104] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A train fusion positioning method based on multi-dimensional data, characterized in that: include: S1. Obtaining basic data formed based on manually entered line data and multi-dimensional data for positioning the pole; S2. After receiving the trigger signal, a corresponding expected bar is generated for each dimension in the multi-dimensional data, and each expected bar is assigned a corresponding original weight; S3. According to the results of each expected pole, the original weights corresponding to the expected poles with the same results are added together to obtain a comprehensive weight; S4. According to whether the comprehensive weight exceeds the threshold, each expected pole is marked as a precise positioning pole or a preliminary positioning pole; S41. Determine whether the expected pole with the highest comprehensive weight is unique; S42. If not, then the expected rods are all initial positioning rods; S43. If so, determine whether the comprehensive weight is greater than the threshold; S44. If the comprehensive weight is greater than the threshold, the corresponding desired pole is calibrated as the precise positioning pole; S45. If the comprehensive weight is less than the threshold, the corresponding expected pole is the initial positioning pole; S5. Use basic data and precise positioning poles to obtain actual poles of the line, and match the initial positioning poles with the actual poles of the line; S51. Determine the actual pole of the line to be matched; S52. Calculate the actual line pole mileage x according to the precise positioning pole mileage; S53. Calculate the mileage of each initial positioning pole ; S54. According to the formula Calculate and obtain the error range The initial positioning rod inside; S55. Update the actual pole data of the line to be matched to the initial positioning pole obtained in S54; S6. Correct the initial positioning pole to the actual pole of the successfully matched line, and generate one-pole-one-file data in combination with the precise positioning pole.

2. The train fusion positioning method based on multidimensional data according to claim 1 is characterized in that: The basic data include: line name, line type, station section name, pole number and kilometer mark field.

3. The train fusion positioning method based on multidimensional data according to claim 1 is characterized in that: The multi-dimensional data includes: current pole information, pole number image recognition results, mileage increment, Beidou positioning data and GYK kilometer mark.

4. The train fusion positioning method based on multidimensional data according to claim 1 is characterized in that: After receiving the trigger signal in S2, false trigger filtering is performed on the trigger signal.

5. The train fusion positioning method based on multi-dimensional data according to claim 4 is characterized in that: The specific steps of the false trigger filtering include: each time a trigger signal is received, the mileage of the current train is synchronously recorded; if the result of subtracting the mileage of the current train from the mileage of the previous train is less than the minimum pole spacing, the current trigger signal is falsely triggered and filtered.

Images