Multi-source detection data mileage correction method and system for urban rail comprehensive detection vehicle

Abstract

The invention relates to an urban rail comprehensive detection vehicle multi-source detection data mileage correction method and system. The method comprises the following steps: data feature identification: identifying instantaneous abrupt change feature points at joints of an anchor section by analyzing pantograph-catenary pull-out value change features to obtain a series of feature points corresponding to the positions of the joints of the anchor section in the through mileage; alignment of the pull-out value data features and the through mileage: matching the pantograph-catenary pull-out value data with the through mileage through feature point alignment and resampling so as to correct mileage deviation and enhance data positioning precision; and time-mileage synchronization of multi-source data: reconstructing a time sequence through linear interpolation, ensuring that the resampled pantograph-catenary pull-out value data is synchronized in time and mileage, and endowing new time information into other professional detection data. The system comprises a detection module, a database and a processing module, and the processing module can extract pantograph-catenary pull-out value detection data from the database and complete mileage correction by using a mileage correction method.

Description

Technical Field

[0001] The present invention relates to the technical field of urban rail detection, and in particular to a method and system for correcting mileage of multi-source detection data of an urban rail integrated detection vehicle. Background Art

[0002] The urban rail comprehensive inspection vehicle (hereinafter referred to as the "comprehensive inspection vehicle") is a vehicle specially used for the inspection and maintenance of urban rail transit lines. It is equipped with a variety of advanced inspection equipment and systems, which can realize comprehensive and efficient inspection of the status of facilities such as tracks, catenary, and communications. During operation, the comprehensive inspection vehicle receives inspection data through various professional sensors, and collects time and mileage information at the same time. These data are used to evaluate the geometric parameters of the track (such as gauge, level, height, and direction), the geometric parameters of the contact network (such as pull-out value, guide height, and hard points), and the performance of the signal system (such as signal lights, switch machines, and track circuits). In addition, the comprehensive inspection vehicle is also equipped with high-precision equipment such as laser scanners, inertial navigation systems, high-definition cameras, and data acquisition systems to ensure the accuracy and reliability of the inspection results. Through real-time data transmission and remote monitoring platforms, managers can keep abreast of the inspection situation, achieve precise repairs and efficient maintenance, and ensure the safety and stable operation of urban rail transit.

[0003] During operation, the comprehensive inspection vehicle receives inspection data through various professional sensors, and collects time and mileage information. When a defect is detected, manual re-inspection and repair work are required. This requires that the mileage information of each inspection point in the inspection system is consistent with the on-site mileage. However, the mileage positioning system of the comprehensive inspection vehicle adopts the photoelectric encoder pulse counting method. Its principle is that the urban rail comprehensive inspection wheel pair rotates to drive the photoelectric encoder to rotate synchronously, and the pulse spacing is calculated according to the wheel circumference and the number of pulses, and then the passing distance is calculated by the pulse count and the pulse spacing. Therefore, during the inspection process, the following factors cause deviations in mileage positioning:

[0004] (1) The wheel is not round: the actual circumference of the wheel may deviate from the theoretical circumference;

[0005] (2) Acceleration or braking: Acceleration or braking causes sliding friction between the wheels and the track, resulting in deviation;

[0006] (3) Differences between system mileage calculation and line design drawings: System mileage calculation is based on continuous mileage, while line design drawings contain long and short chain information, resulting in certain deviations.

[0007] Due to the superposition of the above factors, the resulting deviation will gradually accumulate as the detection mileage increases, eventually resulting in a large deviation between the mileage corresponding to the detection data in the system and the on-site mileage, i.e. mileage misalignment, which seriously affects the positioning accuracy of on-site retesting and maintenance.

[0008] The current methods for railway inspection data mileage correction (alignment) include:

[0009] (1) Modify the detection section based on data characteristics

[0010] For example, Zhao Zhengyang et al. proposed a mileage calibration method for track dynamic geometry and wheel-rail force detection data under unified acquisition and decentralized acquisition modes in the article "Method and Application of Mileage Calibration of Track Dynamic Geometry and Wheel-Rail Force Detection Data" (Zhao Zhengyang, Wang Bing, Li Yang, et al. Mileage Calibration Method and Application of Track Dynamic Geometry and Wheel-Rail Force Detection Data [J]. Railway Construction, 2022, 62(10): 55-59.). In the unified acquisition mode, the mileage of the track dynamic geometry detection data is first calibrated, and then the mileage of the wheel-rail force detection data is synchronized and calibrated with it. In the decentralized acquisition mode, the mileage calibration of the wheel-rail force detection data is realized by using the method of correlation analysis of special sections, line elements and detection waveforms. However, the circular curve involved in the article contains four feature points, and its recognition accuracy is greatly limited; and the mileage error is corrected based on the main point section, which is only accurately corrected near the main point, and there is still a large error in other areas.

[0011] (2) Hardware assistance

[0012] For example, Yang Chao et al. mentioned in the article "Principle and Application of GPS Mileage Automatic Correction System" (Yang Chao, Wang Weidong. Principle and Application of GPS Mileage Automatic Correction System [J]. Railway Computer Application, 2009, 18 (5): 33-35.) that the global positioning system GPS and differential positioning system DGPS are installed on the comprehensive inspection vehicle to automatically correct the mileage deviation in real time online. For another example, Liu Xiaolei et al. mentioned in the article "Application of RFID Technology in High-Precision Positioning of Trains" (Liu Xiaolei, Huang Pu. Application of RFID Technology in High-Precision Positioning of Trains [J]. Urban Rapid Rail Transit, 2017, 30 (3): 107-111.) that radio frequency identification technology and short-range wireless communication technology can be used to install transmitters and receivers with mileage codes on the track line and the car body to control the mileage deviation from the source. However, the above method has the following defects: on the one hand, urban rail transit is mostly located in underground tunnels. If there is no signal transceiver equipment, wireless signal transmission is difficult; on the other hand, whether adding equipment to enhance wireless signal transmission or adding other types of auxiliary equipment, high installation costs will be incurred, and subsequent maintenance costs will also be involved. In addition, some special lines have restrictions that prohibit the installation of external devices.

[0013] Therefore, the existing technology still cannot solve the problem of mileage deviation (i.e. mileage misalignment) caused by various factors during the comprehensive inspection process, which results in errors between the mileage points corresponding to the inspection data and the on-site mileage points, seriously affecting the positioning accuracy of on-site re-testing and maintenance. Therefore, a method is urgently needed to align the mileage information in the system with the actual mileage, so as to achieve accurate maintenance and improve work efficiency.

[0014] In addition, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making the present invention, but due to space limitations, not all details and contents are listed in detail. However, this does not mean that the present invention does not have the characteristics of these prior arts. On the contrary, the present invention already has all the characteristics of the prior art, and the applicant reserves the right to add relevant prior art to the background technology. Summary of the invention

[0015] In view of the deficiencies in the prior art, the present invention provides a method and system for correcting the mileage of multi-source detection data of an urban rail integrated inspection vehicle, and in particular, a method and system for correcting the mileage of multi-source detection data of an urban rail integrated inspection vehicle based on the characteristics of bow-net pull-out value data, so as to solve at least some of the above-mentioned technical problems.

[0016] The present invention discloses a mileage correction method for multi-source detection data of an urban rail comprehensive detection vehicle, which comprises:

[0017] S1. Data feature identification: By analyzing the change characteristics of the bow-net pull-out value, the instantaneous mutation feature points at the anchor section joint are identified to obtain a series of feature points corresponding to the anchor section joint positions in the through mileage;

[0018] S2. Alignment of the pull-out value data features with the through mileage: Through feature point alignment and resampling, the bow-net pull-out value data is matched with the through mileage to correct the mileage deviation and enhance the data positioning accuracy;

[0019] S3. "Time-mileage" synchronization of multi-source data: Reconstruct the time series through linear interpolation to ensure that the resampled bow net pull-out value data is synchronized in time and mileage, and assign new time information to other professional detection data.

[0020] The present invention uses the characteristics of the bow-net pull-out value data to correct the mileage of multi-source detection data. The characteristics of the bow-net pull-out value data are obvious and easy to obtain, because these characteristics are caused by on-site hardware (such as anchor segment joints), and the mileage information of these hardware has been accurately recorded in the through mileage. By drawing the "mileage (x)-pull-out value (y)" curve, the slope mutation points can be intuitively identified, and these mutation points are characteristic points, thereby ensuring the high accuracy of data feature recognition and reducing the possibility of misjudgment. The operation process is mainly divided into three steps: data feature recognition, feature point alignment and multi-source data synchronization. The calculation amount of each step is small, which is suitable for large-scale data processing. The difficulty of data acquisition is low, and the bow-net pull-out value data can be collected in real time by sensors on the comprehensive inspection vehicle without complex equipment or technology. Through the identification and alignment of feature points, the mileage deviation can be effectively eliminated, the positioning accuracy of multi-source detection data can be improved, and the accuracy of the detection results can be improved. High-precision data alignment and synchronization not only optimizes the maintenance strategy, reduces unnecessary maintenance work, and reduces maintenance costs, but also enhances the overall reliability of the system, providing strong technical support for the detection and maintenance of urban rail transit.

[0021] According to a preferred embodiment, the instantaneous mutation feature point at the anchor segment joint is obtained by drawing a "mileage-pull-out value" curve, wherein the slope between each data point in the "mileage-pull-out value" curve and its previous data point is calculated, and when the slope corresponding to any data point exceeds a set slope threshold, the point is taken as a feature point and the mileage of the data is recorded as an identifier of the feature point.

[0022] According to a preferred embodiment, the identified feature points are extracted to form a data feature set, and the difference between the data feature set and the through-mileage anchor segment joint mileage set is calculated to obtain a phase difference set, and the phase difference set is applied as a translation amount to each element of the original data feature set to obtain a translated data feature set, thereby achieving alignment of the bow-net pull-out value data feature information with the anchor segment joint in the through-mileage.

[0023] According to a preferred embodiment, in the process of resampling the bow net pull-out value data, any two adjacent original data feature points are selected from the original data feature set, and the two original data feature points are mapped to the translated data feature set to obtain two corresponding new data feature points, and sampling points are obtained at a preset sampling frequency within the original interval formed between the two original data feature points to form a pull-out value data set.

[0024] According to a preferred embodiment, sampling points are obtained in a resampling interval formed between two new data feature points in a manner maintaining the same sampling frequency, wherein when the original interval length is less than the corresponding resampling interval length, the number of sampling points in the resampling interval increases; and when the original interval length is greater than the corresponding resampling interval length, the number of sampling points in the resampling interval decreases.

[0025] According to a preferred embodiment, for any new sampling point located within the resampling interval, its corresponding bow-net pull-out value can be calculated by linear interpolation, and the calculated bow-net pull-out value can be incorporated into a new pull-out value data set to achieve alignment of the pull-out value data with the through mileage.

[0026] According to a preferred implementation, for any new sampling point, the two closest feature points of the original data are first found, and linear interpolation is used to calculate the new time point so that the mileage value corresponding to the new sampling point is consistent with the new time point, thereby synchronizing the bow net pull-out value data and mileage data in the time dimension.

[0027] According to a preferred embodiment, various professional inspection data collected by the urban rail comprehensive inspection vehicle during operation can be stored in the database as multi-source data in the form of attached inspection mileage and inspection time, wherein the multi-source data includes the bow-net pull-out value inspection data.

[0028] According to a preferred embodiment, when the on-site mileage is inconsistent with the inspection mileage, the on-site mileage and the through mileage are aligned through the database, wherein a correspondence table between the on-site mileage and the through mileage is maintained in the database in advance to establish a mileage comparison table, and then the on-site mileage is extracted through associated retrieval.

[0029] The present invention also discloses a mileage correction system for multi-source detection data of an urban rail integrated detection vehicle, which includes: a detection module, which includes a plurality of detection devices for acquiring different types of detection data; a database for storing the detection data acquired by each detection device; and a processing module for extracting data from the database and performing data processing to complete mileage correction. Preferably, the bow-net pull-out value detection device configured in the detection module can send the acquired bow-net pull-out value detection data to the database, so that the processing module can extract the bow-net pull-out value detection data from the database and complete the mileage correction using the aforementioned mileage correction method. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a main step diagram of a mileage correction method according to a preferred embodiment of the present invention;

[0031] Figure 2 It is a plan view of the bow-net structure;

[0032] Figure 3 It is a schematic diagram of the mechanism of sudden change of pull-out value at the joint of the anchor segment;

[0033] Figure 4 This is a data feature identification schematic diagram of a preferred implementation mode provided by the present invention;

[0034] Figure 5 It is a schematic diagram of pull-out value resampling according to a preferred embodiment of the present invention;

[0035] Figure 6 It is a schematic diagram of assigning on-site mileage labels according to a preferred embodiment of the present invention;

[0036] Figure 7 It is an operation flow chart of a mileage correction method according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0037] The following is a detailed description with reference to the accompanying drawings.

[0038] Example 1

[0039] like Figure 1 As shown, the present invention discloses a method for correcting the mileage of multi-source detection data of an urban rail comprehensive inspection vehicle (or simply referred to as an integrated inspection vehicle), in particular, a method for correcting the mileage of multi-source detection data of an urban rail comprehensive inspection vehicle based on the characteristics of pantograph-net pull-out value data, which comprises:

[0040] S1, data feature identification;

[0041] S2, the pull-out value data feature is aligned with the through mileage;

[0042] S3. Multi-source data “time-mileage” synchronization.

[0043] Preferably, the comprehensive inspection vehicle can be used to detect and evaluate various infrastructures of the rail transit system in real time to ensure safety and efficiency. The comprehensive inspection vehicle integrates a variety of advanced detection technologies, including laser measurement, photoelectric sensing, high-definition camera, high-speed data acquisition and processing system. It is equipped with multiple sensors and antennas to ensure real-time monitoring and data recording during driving. The comprehensive inspection vehicle runs regularly on the rail line to provide continuous monitoring and fault warning for the rail transit system. This active detection method not only improves the safety of urban rail transit, but also greatly reduces maintenance costs and operational risks.

[0044] Preferably, the mileage correction method of the present invention may further include, before step S1, the following steps: S0, multi-source data collection and storage.

[0045] Preferably, multi-source data is a general term for various professional inspection data collected by the comprehensive inspection vehicle during operation, including track inspection data, catenary inspection data, communication signal inspection data, power system inspection data, etc., to achieve comprehensive inspection of the facility status. Different professional inspection processes can be completed by different inspection devices carried on the comprehensive inspection vehicle. Track inspection may include gauge, level and height (unevenness) inspection, as well as track wear and section inspection to obtain corresponding track inspection data, among which laser ranging sensors, acceleration sensors, gyroscopes and other equipment may be used to perform gauge, level and height (unevenness) inspection; laser scanning technology may be used to capture the true cross-sectional shape of the track for track wear and section inspection. The pantograph detection may include contact wire wear and height detection, pull-out value (contact wire offset) detection, and dynamic pantograph-net relationship detection to obtain pantograph-net detection data, wherein the contact wire wear and height detection may be performed using laser ranging and visual recognition technology; the pull-out value (contact wire offset) detection may be performed using cameras and computer vision technology; the relative movement of the pantograph and the contact wire may be monitored by combining high-speed cameras and sensors to perform dynamic pantograph-net relationship detection. The communication signal detection may include wireless signal strength and quality detection and signal system status detection to obtain communication signal detection data, wherein the wireless signal strength and quality detection may be performed using a radio spectrum analyzer; the connectivity and quality of the communication signal may be analyzed using signal monitoring software to perform signal system status detection. The power system detection may include power supply voltage and current detection to obtain power system detection data, wherein voltage and current sensors may be used for real-time monitoring to perform power supply voltage and current detection.

[0046] Preferably, the original data recorded by the comprehensive inspection vehicle during operation may include inspection mileage and inspection time, wherein the inspection mileage is used to indicate the relative spatial position of the inspection point, generally starting from "0"; the inspection time is used to record the time information when data is collected at a certain point during the inspection process.

[0047] Preferably, the multi-source data collected by the comprehensive inspection vehicle can be stored in a database, wherein the data stored in the database may include bow-net pull-out value detection data (which may be referred to as bow-net pull-out value or pull-out value for short) and other professional detection data in addition to this. Preferably, the bow-net pull-out value detection data is a type of bow-net detection data, which is the detection value of the lateral offset of the contact line from the center line of the pantograph. The detection of the pull-out value is crucial to ensure the stability of power transmission and reduce the wear between the pantograph and the contact line. Excessive pull-out values ​​may cause the pantograph to detach from the contact line, thereby interrupting the power supply and affecting the normal operation of the train. Preferably, the bow-net pull-out value detection data can be obtained by using a high-definition camera installed on the comprehensive inspection vehicle to shoot the relative position between the pantograph and the contact line in real time, and then calculating the lateral offset between the bow and the catenary through image analysis technology; it can also be obtained by using laser positioning technology to accurately measure the lateral distance between the bow and the catenary, and then processing the laser measurement data. Preferably, the collected image or laser data is analyzed in real time by an on-board computer, and the built-in analysis software of the on-board computer can automatically identify the position of the pantograph and the contact wire, and calculate the pull-out value, wherein the generated pull-out value data can be presented in the form of a time series and used to monitor trends and detect anomalies. In the prior art, the pantograph-net pull-out value detection data is usually used as an important indicator for evaluating the stability of the power supply system during train operation, so that track maintenance personnel can timely discover problems and perform preventive maintenance by monitoring the pull-out value. For example, if the pull-out value of a certain section of track continues to exceed the set safety range, it is necessary to check the installation of the track or contact network, and whether the position of the pantograph needs to be adjusted. The present invention utilizes the unique characteristics of the pantograph-net pull-out value detection data, takes the characteristic points in the pantograph-net pull-out value detection data as the boundary, completes the segmented alignment of the pantograph-net pull-out value detection data and the through mileage, and finally realizes the mileage correction of the multi-source detection data of the urban rail comprehensive inspection vehicle.

[0048] In urban rail transit, rigid contact network is a commonly used power supply method, especially in underground tunnel environments. The rigid contact network is composed of metal rods and crossbars, with high strength and sturdiness in structure. It is installed in a fixed suspension manner, so that the contact line does not have the tension problem in the traditional suspension method. Therefore, during the dynamic detection process, the "mileage-pull-out value" curve between the pantograph and the contact line shows a uniform regularity and has significant periodicity. However, at the anchor section joint, the situation is different. As Figure 2 As shown in the figure, the anchor segment joint is the key structure for achieving a smooth transition between the contact network anchor segments. It connects the front and rear contact lines through the parallel overlap of short segments. Figure 3As shown in the figure, when the pantograph moves to the anchor section joint, the power supply line transitions to the next section of the contact wire, and the pull-out value at this time will show an instantaneous mutation. This mutation is caused by the sudden change in the physical connection mode of the contact wire, which is an important feature that cannot be ignored in the detection process. This mutation is the unique characteristic of the bow-net pull-out value detection data. By reversely analyzing the changes in the bow-net pull-out value, the position of the anchor section joint can be identified. This identification can be based on the comparison of the detected sudden change in the bow-net pull-out value with the through mileage information, because the mileage information of the anchor section joint is queryable and accurate in the through mileage. At the same time, since the comprehensive inspection vehicle is equipped with a variety of professional equipment during the inspection process, the mileage information consistency of each professional inspection data collected at the same time is guaranteed. Therefore, the through mileage and the on-site mileage can be aligned by a step-by-step association method. This alignment of multi-source detection data with on-site mileage not only improves the accuracy of the detection data, but also enhances the reliability and operability of the analysis results. The through mileage refers to the length of the subway project that is physically connected after the construction is completed. The mileage records of hardware facilities such as "anchor joints" are also stored in it. The on-site mileage refers to the actual mileage mark on site, which includes information such as the length of the chain and the kilometer mark, such as "K20+300". Such signs have physical signs on site to facilitate identification by staff.

[0049] Preferably, if Figure 4 As shown, the bow-net pull-out value detection data with unique characteristics can be manifested as an instantaneous mutation between the two detection points. This instantaneous mutation can be intuitively seen by drawing the "mileage (x)-pull-out value (y)" curve. The absolute value of the slope of the curve at the position where the instantaneous mutation occurs is significantly larger than that at other positions. Furthermore, the data for drawing the "mileage (x)-pull-out value (y)" curve is high-precision data obtained by extracting the bow-net pull-out value related data from the original data collected by the comprehensive inspection vehicle after pre-processing operations such as smoothing to reduce the impact of noise. The smoothing process can, for example, use a moving average method or a low-pass filter.

[0050] Preferably, in step S1, based on the drawn “mileage (x)-pullout value (y)” curve, the slope between each data point and its previous data point can be calculated, wherein the slope can be calculated by the following formula:

[0051]

[0052] Among them, y i and i-1 are the extracted values ​​of the i-th and i-1-th data points, respectively, x i and x i-1 are the mileages of the i-th and i-1-th data points respectively.

[0053] Furthermore, a slope threshold θ is set to distinguish normal slopes from sudden slopes. By traversing all calculated slopes, check whether each slope exceeds the threshold θ. If a slope exceeds the threshold θ, the position is considered to be a feature point and the mileage x of the data is recorded. i As the identification of feature points.

[0054] Preferably, in order to further improve the accuracy of recognition, the initially identified feature points can be verified. The verification method includes but is not limited to checking the continuity and stability of the data before and after the feature point, and comparing with the known anchor segment joint position. If the pull-out value data around the feature point conforms to the expected mutation pattern and is consistent with the known anchor segment joint position, the validity of the feature point is confirmed.

[0055] Preferably, the confirmed valid feature points are recorded to form a data feature identification list. Each feature point may include its mileage position and the corresponding change in the pantograph pull-out value. These feature points can be used in the subsequent mileage alignment step.

[0056] Preferably, by identifying the characteristics of the bow-net pullout value data, a series of characteristic points can be obtained, and these characteristic points correspond to the joint positions of the anchor segments in the through mileage. Preferably, after obtaining a series of characteristic points, step S2 can be executed, wherein step S2 can include the following sub-steps: characteristic data alignment; pullout value data resampling.

[0057] Preferably, in the coordinate system specified by the “mileage (x)-pullout value (y)” curve, the identified feature points are extracted from the pantograph pullout value data to form a data feature set B = {b1, b2, …, b n These feature points are represented by slope mutation points on the "mileage (x)-pullout value (y)" curve. These feature points are identified by calculating the slope between adjacent points and setting a threshold. Each feature point contains its mileage position b i And the corresponding changes in the pantograph pull-out value. At the same time, the known anchor segment joint position information is extracted from the through-mileage database to form a through-mileage anchor segment joint mileage set R = {r1, r2, …, r n}. This information includes the mileage position r of each anchor segment joint i Among them, the through mileage data usually comes from the full-length information recorded after the completion of the subway project construction, which has higher accuracy.

[0058] Preferably, for each pair of data, the difference between the x-coordinate of the data feature and the x-coordinate of the anchor segment joint mileage in the through mileage is calculated to obtain the phase difference set Δ. The specific calculation formula is as follows:

[0059] δ i =r i -b i

[0060] Δ={δ1,δ2,…,δ n}

[0061] Among them, δ i is the difference between the i-th data pair, i is an integer and 1≤i≤n; Δ is the phase difference set.

[0062] Optionally, the phase difference value Δ may be found by performing statistical analysis on the phase difference set Δ avg , which is used to reflect the overall offset between the bow-net pull-out value data and the penetration mileage, where the overall offset can be more accurately estimated by calculating the average or median.

[0063] Preferably, the phase difference set Δ is applied as a translation amount to each element of the original data feature set B to obtain a translated data feature set B'. The specific calculation formula is as follows:

[0064] b′ i =b i +δ i

[0065] B′={b′1,b′2,…,b′ n}

[0066] Among them, b' i is the value of the i-th data in the original data feature set B after translation, i is an integer and 1≤i≤n; B' is the data feature set after translation.

[0067] In order to ensure the accuracy of the alignment results, the resampled data can be verified. The specific methods include: feature point comparison, that is, comparing the translated feature point B' with the anchor segment joint position R in the through mileage to check whether the two are consistent; data visualization, that is, drawing the translated "mileage (x)-pullout value (y)" curve, observing the smoothness of the curve and the mutation of the feature points, to ensure that the aligned data meets expectations.

[0068] Through the above steps, the characteristic information (i.e., characteristic points) of the pantograph-net pullout value data can be aligned with the joint of the anchor section in the through mileage. This method is not only simple to operate and has a small amount of calculation, but also can effectively eliminate mileage deviation, improve the positioning accuracy of multi-source detection data, and provide reliable support for urban rail inspection and maintenance.

[0069] Preferably, in the process of aligning the characteristic data, the characteristic points of the pantograph-catwalk pullout value data are aligned with the joints of the anchor sections in the through mileage, and a new characteristic point set B'={b'1, b'2, ..., b' n In order to further ensure the accuracy of the pulled value data at the new mileage position, the pulled value data needs to be resampled.

[0070] Preferably, when resampling the extracted value data, the extracted value data set L can be defined first. k,k+1 ={l1,l2,…,l m}, where m represents any two adjacent data feature points (i.e. data points) in set B. k and b k+1 The number of sampling points between, and 1≤k<n, k, m, n are positive integers. When the set B is aligned with the anchor segment joint mileage set R in the through mileage, the data feature point b k and b k+1 are mapped to new positions b' k and b' k+1 In this process, the original pull-out value data set L k,k+1 Corresponding resampling is also required to adapt to the new data point distribution.

[0071] Preferably, during the resampling process of the pantograph-catwalk pullout value data, the sampling frequency is kept unchanged so that the pullout value data set L k,k+1 The elements are evenly distributed, and the sampling frequency (unit length interval) is f s , that is, there is f per meter s The distance between sampling points or each point is Δk = 1 / f s . When the original interval length is u=b k+1 -b k When the corresponding resampling interval length is u'=b' k+1 -b' k , the stretching and compression processing methods of data are as follows:

[0072] When u'>u, it means that the interval of the aligned data is larger than that of the original data, that is, the original data is stretched. In order to maintain the same frequency, the data is mapped to a longer interval length based on the upsampling method. k ,b' k+1 ] to generate more sampling points, that is, to increase the interval [b' k ,b' k+1 ]The number of sampling points within;

[0073] When u'<u, it means that the interval of the aligned data is smaller than that of the original data, that is, the original data is compressed. In order to maintain the same frequency, the data is mapped to a shorter interval length based on the downsampling method. k ,b' k+1 ] generates fewer sampling points, that is, reduces the interval [b' k ,b' k+1 ] is the number of sampling points within.

[0074] Preferably, for any new sampling point d' j , and its corresponding pull-out value l' j The linear interpolation method can be calculated using the following formula:

[0075]

[0076] Among them, b' k ≤d' j <b' k+1 , that is, the new sampling point is located between the new data feature points.

[0077] Through the above steps, a new pull value data set L' is generated k,k+1 ={l'1,l'2,…,l' j},like Figure 5 So far, the pull-out value data and the through mileage are aligned to obtain the associated data of the two.

[0078] Preferably, in order to ensure the accuracy of the resampled data, the resampled data may be verified by feature point comparison and / or data visualization.

[0079] In the process of each professional inspection, in addition to the corresponding inspection indicators, the original data also includes time information and mileage information. Each inspection device of the comprehensive inspection vehicle uses the same set of time encoders, that is, all the inspection data collected by each professional inspection process at the same time correspond to the same mileage information, so as to ensure the smooth implementation of the "time-mileage" synchronization of multi-source data. Preferably, step S2 has achieved dimensional alignment of the pulled value data and the through mileage. Due to the resampling process, the number of sampling points has changed, resulting in the original time series no longer being applicable. Therefore, in step S3, the time series can be reconstructed by linear interpolation.

[0080] Preferably, in the initial state, the time series t of each detection device i and the mileage value d at the time of sampling i Correspondingly, a "time-mileage" correspondence is formed:

[0081]

[0082] Among them, t i represents the i-th time point, d i represents the mileage value corresponding to the i-th time point. Since the data collection of each inspection system is synchronous during the operation of the comprehensive inspection vehicle, the data of each inspection system is collected synchronously at each time point t i The corresponding mileage value d i It should be consistent.

[0083] After mileage calibration and resampling of the pantograph and catenary pullout values, the new mileage value d' j Replaces the original mileage value d i In addition to u'=u, the resampling process will change the number and distribution of sampling points, so the time series also needs to be recalculated based on the new mileage distribution. Preferably, time synchronization can use linear interpolation to calculate the new time point t' j , the calculation formula is as follows:

[0084]

[0085] Among them, t k and t k+1 is the corresponding feature point b k and b k+1 The original time, and b k ≤d' j <b k+1 .

[0086] Therefore, for each new mileage value d' j , we can first find the two original mileage values ​​b closest to it k and b k+1 , so that b k ≤d' j <b k+1 , and then use linear interpolation to calculate the new time point t' j , through this method, the new mileage d' j can be compared with the new time point t' j Keep consistent so that the bow network pull-out value data and mileage data are synchronized in the time dimension.

[0087] The present invention ensures that each detection system has consistent mileage data at the same time point by assigning new time information to other professional detection data, achieving the effect of "time-mileage" synchronization, and finally realizing the alignment of multi-source data with through mileage. Through the time synchronization method, the mileage data of each detection system at the same time point is ensured to be consistent, the error caused by the asynchronous data acquisition time is eliminated, and the consistency of the data is improved. The linear interpolation method is simple and efficient, with low computational complexity. It is suitable for large-scale data processing, can quickly complete the reconstruction of time series, and improves computational efficiency. Through the interpolation method, the missing sampling points caused by resampling are filled, ensuring the integrity and continuity of the data. The time synchronization method is based on the unified time encoder of the comprehensive inspection vehicle, which ensures the reliability of data synchronization and improves the positioning accuracy of multi-source detection data.

[0088] Preferably, the mileage correction method of the present invention may further include, after step S3: S4, assigning on-site mileage labels, wherein step S4 is mainly used for situations where the on-site mileage is inconsistent with the detection mileage. Furthermore, since most on-site mileages are "broken chains", that is, in railway or road projects, due to local line changes, segmented measurements, etc., the pile numbers (i.e., mileage pile numbers) are discontinuous, while the mileage recorded by the comprehensive inspection vehicle equipment during the inspection process is continuous, resulting in inconsistency between the on-site mileage and the detection mileage, which seriously affects on-site re-measurement and maintenance positioning. The present invention calculates the length of the long and short chains, performs mileage difference correction on the original inspection data, and then uses the label mapping method to finally achieve "multi-source inspection data-on-site mileage" association alignment.

[0089] Preferably, in step S4, the on-site mileage information is assumed to be K = {K1, K2, ..., K n}, no matter whether there is a "break" in the on-site mileage, the length of the on-site mileage is consistent with the length of the through mileage, so there must be a mileage point corresponding to the on-site mileage on the through mileage. The on-site mileage can be used as a label to map the entire mileage to the through mileage to achieve the purpose of aligning the on-site mileage with the through mileage, thereby achieving the effect of aligning multi-source data with the on-site mileage. The principle is as follows Figure 6 shown.

[0090] Preferably, the present invention can realize the alignment of the on-site mileage and the through mileage through a database, wherein the correspondence table of the on-site mileage and the through mileage can be maintained in advance in the database to establish a detailed mileage comparison table, and then the on-site mileage can be extracted through an associated search. The mileage comparison table can be shown in the following table.

[0091] id Actual mileage Mileage n <![CDATA[K n ]]> <![CDATA[G n ]]> n+1 <![CDATA[K n+1 ]]> <![CDATA[G n+1 ]]> n+2 <![CDATA[K n+2 ]]> <![CDATA[G n+2 ]]> … … …

[0092] Based on this, the present invention provides a method for correcting the mileage of multi-source detection data of urban rail comprehensive inspection vehicles based on the characteristics of pantograph-net pull-out value data, which has the advantages of simple operation process, small calculation amount, low difficulty in data acquisition, and good correction effect. The specific operation process can be as follows Figure 7 shown.

[0093] Example 2

[0094] This embodiment is a further improvement on Embodiment 1, and the repeated contents will not be repeated here.

[0095] The present invention also discloses a mileage correction system for multi-source detection data of an urban rail integrated inspection vehicle, in particular, a mileage correction system for multi-source detection data of an urban rail integrated inspection vehicle based on the data characteristics of the bow-net pull-out value, for executing the mileage correction method as described in Example 1.

[0096] Preferably, the mileage correction system may include: a detection module, which includes multiple detection devices for acquiring different types of detection data; a database for storing the detection data acquired by each detection device; and a processing module for extracting data from the database and performing data processing to complete mileage correction.

[0097] Preferably, the detection module may be configured with a bow-net pull-out value detection device for acquiring bow-net pull-out value detection data, wherein the bow-net pull-out value detection device may send the acquired bow-net pull-out value detection data to a database, so that the processing module may extract it from the database and complete the mileage correction using the mileage correction method as described in Example 1.

[0098] It should be noted that the above-mentioned specific embodiments are exemplary, and those skilled in the art can come up with various solutions inspired by the disclosure of the present invention, and these solutions also belong to the disclosure scope of the present invention and fall within the protection scope of the present invention. Those skilled in the art should understand that the present invention specification and its drawings are illustrative and do not constitute a limitation of the claims. The scope of protection of the present invention is defined by the claims and their equivalents. The present invention specification contains multiple inventive concepts, such as "preferably" or "according to a preferred embodiment", which means that the corresponding paragraph discloses an independent concept, and the applicant reserves the right to file a divisional application based on each inventive concept. Throughout the text, the features guided by "preferably" are only an optional method and should not be understood as a must-have setting. Therefore, the applicant reserves the right to abandon or delete the relevant preferred features at any time.

Claims

1. A method for correcting mileage of multi-source detection data of urban rail comprehensive detection vehicle, characterized in that: It includes: S1. Data feature identification: By analyzing the change characteristics of the bow-net pull-out value, the instantaneous mutation feature points at the anchor section joint are identified to obtain a series of feature points corresponding to the anchor section joint positions in the through mileage; S2. Alignment of the pull-out value data features with the through mileage: Through feature point alignment and resampling, the bow-net pull-out value data is matched with the through mileage to correct the mileage deviation and enhance the data positioning accuracy; S3. "Time-mileage" synchronization of multi-source data: Reconstruct the time series through linear interpolation to ensure that the resampled bow net pull-out value data is synchronized in time and mileage, and assign new time information to other professional detection data.

2. The method according to claim 1, characterized in that: The instantaneous mutation feature points at the anchor segment joints are obtained by drawing the "mileage-pull-out value" curve, wherein the slope between each data point and its previous data point in the "mileage-pull-out value" curve is calculated. When the slope corresponding to any data point exceeds the set slope threshold, the point is taken as a feature point and the mileage of the data is recorded as the identifier of the feature point.

3. The method according to claim 1 or 2, characterized in that: The identified feature points are extracted to form a data feature set, and the difference between the data feature set and the through-mileage anchor segment joint mileage set is calculated to obtain a phase difference set. The phase difference set is applied as a translation amount to each element of the original data feature set to obtain a translated data feature set, thereby achieving alignment of the bow-net pull-out value data feature information with the anchor segment joint in the through-mileage.

4. The method according to any one of claims 1 to 3, characterized in that: During the resampling process of the bow net pull-out value data, any two adjacent original data feature points are selected from the original data feature set, and the two original data feature points are mapped to the translated data feature set to obtain the corresponding two new data feature points, and sampling points are obtained at a preset sampling frequency within the original interval formed between the two original data feature points to form a pull-out value data set.

5. The method according to any one of claims 1 to 4, characterized in that: Sampling points are obtained in a resampling interval formed between two new data feature points in a manner maintaining the same sampling frequency, wherein when the original interval length is less than the corresponding resampling interval length, the number of sampling points in the resampling interval increases; when the original interval length is greater than the corresponding resampling interval length, the number of sampling points in the resampling interval decreases.

6. The method according to any one of claims 1 to 5, characterized in that: For any new sampling point within the resampling interval, its corresponding bow-net pull-out value can be calculated by linear interpolation, and the calculated bow-net pull-out value can be incorporated into the new pull-out value data set to achieve alignment of the pull-out value data with the through mileage.

7. The method according to any one of claims 1 to 6, characterized in that: For any new sampling point, first find the two closest feature points of the original data, and use linear interpolation to calculate the new time point so that the mileage value corresponding to the new sampling point is consistent with the new time point, thereby synchronizing the bow net pull-out value data and mileage data in the time dimension.

8. The method according to any one of claims 1 to 7, characterized in that: The various professional inspection data collected by the urban rail comprehensive inspection vehicle during operation can be stored in the database as multi-source data in the form of attached inspection mileage and inspection time, among which the multi-source data includes the pantograph-net pull-out value inspection data.

9. The method according to any one of claims 1 to 8, characterized in that: When the on-site mileage is inconsistent with the inspection mileage, the on-site mileage and the through mileage are aligned through the database. The correspondence table between the on-site mileage and the through mileage is maintained in the database in advance to establish a mileage comparison table, and then the on-site mileage is extracted through associated retrieval.

10. A multi-source detection data mileage correction system for urban rail comprehensive inspection vehicles, characterized in that: It includes: A detection module, which includes a plurality of detection devices for acquiring different types of detection data; A database, used to store the detection data obtained by each detection device; The processing module is used to extract data from the database and process the data to complete the mileage correction. Among them, the bow-net pull-out value detection device configured by the detection module can send the acquired bow-net pull-out value detection data to the database, so that the processing module can extract the bow-net pull-out value detection data from the database and complete the mileage correction using the mileage correction method described in any one of claims 1 to 9.

Images