Heavy haul railway train track correction method, device and equipment

Abstract

The invention relates to the field of train positioning, and provides a heavy haul railway train track correction method, device and equipment, and the method comprises the steps: obtaining Beidou satellite navigation system track data, and correcting the Beidou satellite navigation system track data; performing data fusion on the odometer data and the corrected Beidou satellite navigation system trajectory data to obtain posterior position estimation and posterior speed estimation; determining a central map point of the search area by combining a map database according to the posterior speed estimation and the standard deviation thereof; narrowing a map point search range taking the central map point as the center according to the design factor to obtain a map point set in the search range; selecting two map points closest to the posterior position estimation from the map point set in the search range; and performing trajectory mapping according to the two map points and the posterior position estimation to generate corrected trajectory data. According to the method, the defect that errors exist in heavy haul railway train track data in the prior art is overcome, and high-precision correction and mapping of the train track are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of train positioning, and in particular to a heavy-load railway train track correction method, device and equipment. Background Art

[0002] The main purpose of heavy-load railway trains is to transport large quantities of goods. Their load capacity is significantly higher than that of ordinary freight trains, and their train length may reach several kilometers. These trains are widely used in long-distance, large-scale logistics transportation of bulk goods such as coal, ore, steel, and grain. Due to their unique operating characteristics, heavy-load trains will generate large vibrations and shocks during operation, which poses a severe challenge to the accuracy of the train positioning system.

[0003] In the train positioning system, map matching technology plays a vital role. By matching the real-time position data of the train with the preset railway map, the precise position of the train can be determined, thereby achieving efficient scheduling and safety control. At present, map matching algorithms for railway lines have made certain progress. For example, some scholars have monotonically segmented the line and adopted different positioning methods for straight lines, gentle curves and circular curves to improve the positioning accuracy of trains in different line sections. In addition, there are studies that propose to convert the longitude and latitude data collected by the global navigation satellite system (GNSS) into plane coordinates through Gauss-Kruger projection, and then perform segment matching to simplify data processing and improve positioning efficiency. Another method is based on GPS and digital track line matching, generates digital track lines through fitting and interpolation algorithms, and combines track line positioning correction methods to obtain more accurate train position estimates.

[0004] Although the existing technology has improved the accuracy of train positioning to a certain extent, it still has significant defects. When directly using unprocessed raw sensor data for map matching, it will encounter serious noise interference problems, thereby reducing positioning accuracy. Summary of the invention

[0005] The present invention provides a heavy-haul railway train track correction method, device and equipment, which solves the defect of errors in heavy-haul railway train track data in the prior art and realizes high-precision correction and mapping of the train track.

[0006] The present invention provides a heavy-haul railway train track correction method, comprising the following steps: Obtaining Beidou satellite navigation system trajectory data, and correcting the Beidou satellite navigation system trajectory data; The odometer data and the corrected Beidou satellite navigation system trajectory data are fused to obtain a posteriori position estimate and a posteriori velocity estimate; Determine the center map point of the search area based on the a posteriori velocity estimate and its standard deviation in combination with a map database; narrow the map point search range centered on the center map point based on a design factor to obtain a set of map points within the search range; Selecting two map points closest to the posterior position estimate from the set of map points within the search range; Trajectory mapping is performed based on the two map points and the posterior position estimate to generate corrected trajectory data.

[0007] According to a heavy-load railway train track correction method provided by the present invention, the Beidou satellite navigation system track data is obtained and the Beidou satellite navigation system track data is corrected, specifically including: obtaining the original track curve segment of the train, generating Beidou satellite navigation system track data according to the original track curve segment, and the Beidou satellite navigation system track data is represented as a track point set formed by a group of track points; selecting the first point and the last point in the track point set as two end points to form an initial straight line segment; determining the vertical distance between the middle point and the initial straight line segment; the middle point is the other point on the original track curve segment except the two end points; determining whether there is a segmentation point according to the size of the vertical distance, if there is, determining a new sub-straight line segment according to the segmentation point, and recursively determining all segmentation points based on the new sub-straight line segment; performing a time constraint check on all segmentation points, and sequentially connecting the segmentation points that meet the time constraint to form a simplified representation of the original track curve segment to complete the correction of the Beidou satellite navigation system track data.

[0008] According to a heavy-haul railway train track correction method provided by the present invention, the method determines whether there is a segmentation point according to the size of the vertical distance, and if so, determines a new sub-straight line segment according to the segmentation point, and recursively determines all segmentation points based on the new sub-straight line segment, specifically including: comparing the vertical distance with a preset tolerance threshold; if the vertical distance is less than the preset tolerance threshold, using the initial straight line segment as a simplified representation of the original track curve segment; if the vertical distance is not less than the preset tolerance threshold, using the point corresponding to the maximum vertical distance as the segmentation point; dividing the original track curve segment into two sub-curve segments based on the segmentation point, and generating sub-straight line segments corresponding to the two sub-curve segments; continuing to obtain the maximum vertical distance of the track point relative to the sub-straight line segment, and judging whether the sub-curve segment needs to be further segmented according to the maximum vertical distance; and continuing to recursively process each sub-curve segment obtained by segmentation to determine all segmentation points.

[0009] According to a heavy-load railway train trajectory correction method provided by the present invention, the odometer data and the corrected Beidou satellite navigation system trajectory data are fused, specifically including: obtaining the timestamp range of the corrected Beidou satellite navigation system trajectory data and the odometer data, and determining the common time range of the two; linearly interpolating the odometer data to align the Beidou satellite navigation system trajectory data and the odometer data in the time dimension, and generating a fused data set with the same timestamp; dynamically adjusting the statistical characteristics of system noise and observation noise based on the fused data set through the Sage-Husa adaptive filtering algorithm, and outputting the fused data.

[0010] According to a heavy-haul railway train track correction method provided by the present invention, the map point search range centered on the central map point is narrowed according to the design factor to obtain a map point set within the search range, specifically comprising: taking the central map point as the center, determining the boundary of the search range according to the design factor and the potential moving distance; the potential moving distance is determined according to the posterior velocity estimate and its standard deviation; determining a search area according to the boundary of the search range, and within the determined search area, screening all map points located within the search area from a map database to form a map point set.

[0011] According to a heavy-haul railway train trajectory correction method provided by the present invention, the trajectory mapping is performed according to the two map points and the posterior position estimate to generate the corrected trajectory data, which specifically includes: determining a unit vector according to the two map points; orthogonally projecting the posterior position estimate onto the trajectory line determined by the two map points according to the unit vector to obtain a projection result; using the projection result as a precise position estimate of the train on the railway track; and packaging the precise position estimate together with other relevant parameters to obtain the corrected trajectory data.

[0012] The present invention also provides a heavy-haul railway train track correction device, comprising the following modules: A data smoothing correction module is used to obtain Beidou satellite navigation system trajectory data and correct the Beidou satellite navigation system trajectory data; A data fusion module is used to fuse the odometer data with the corrected Beidou satellite navigation system trajectory data to obtain a posteriori position estimate and a posteriori velocity estimate; An area selection module is used to determine the center map point of the search area according to the a posteriori velocity estimate and its standard deviation in combination with a map database; narrow the map point search range centered on the center map point according to a design factor to obtain a set of map points within the search range; An optimal candidate road segment determination module, configured to select two map points closest to the posterior position estimate from the set of map points within the search range; The trajectory mapping module is used to perform trajectory mapping according to the two map points and the a posteriori position estimate to generate corrected trajectory data.

[0013] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the heavy-load railway train track correction method as described above is implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the heavy-load railway train track correction method as described in any one of the above is implemented.

[0015] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the heavy-load railway train track correction method as described in any one of the above is implemented.

[0016] The invention provides a heavy-duty railway train track correction method, device and equipment, which include the following beneficial effects: by correcting the Beidou satellite navigation system track data, the noise and deviation points in the original data are removed, and the key track features are retained, thereby improving the overall quality of the data and making the track data more accurate and reliable. By fusing the odometer data and the corrected Beidou satellite navigation system track data, the advantages of the two data sources are utilized, the uncertainty of a single data source is reduced, the accuracy of the posterior position estimation and the posterior velocity estimation is improved, and the robustness of the system is enhanced. By dynamically determining the central map point of the search area according to the posterior velocity estimation and its standard deviation in combination with the map database, and narrowing the search range according to the design factor, the search strategy can be adaptively adjusted under different speeds and environments, thereby improving the efficiency and accuracy of map matching. By selecting the two map points closest to the posterior position estimation from the set of map points within the narrowed search range, and using the orthogonal projection method for track mapping, the probability of mismatching is effectively reduced, and the reliability of the map matching result is improved. By correcting the data and dynamically adjusting the search range, the demand for data processing is significantly reduced, the calculation efficiency is improved, and the invention is more efficient in practical applications. Therefore, the present invention achieves high-precision correction and mapping of train tracks, and improves the accuracy and robustness of train positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a framework diagram of the ALIGN-OPMM algorithm provided by the present invention.

[0019] Figure 2 It is a flow chart of the heavy-haul railway train track correction method provided by the present invention.

[0020] Figure 3a The present invention provides Schematic diagram of the Douglas-Peucker algorithm when .

[0021] Figure 3b The present invention provides Schematic diagram of the Douglas-Peucker algorithm when .

[0022] Figure 4 This is an example diagram of the spatial constraints of the trajectory points provided by the present invention.

[0023] Figure 5 It is a structural schematic diagram of a heavy-load railway train track correction device provided by the present invention.

[0024] Figure 6 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. 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.

[0026] The present invention proposes a trajectory correction and orthogonal projection map matching ALIGN-OPMM algorithm framework for heavy-duty railway trains. Aiming at the noise interference and data redundancy problems of traditional map matching algorithms when processing large-scale raw sensor data, the algorithm is optimized and designed, which is particularly suitable for high-speed trains and the layout of mostly straight and smooth rails. First, the Douglas-Peucker algorithm with time constraints is used to correct the Beidou satellite navigation system data, and non-critical data points are eliminated to reduce noise interference and improve data processing efficiency. Subsequently, the odometer (Odometer, ODO) data and Beidou Satellite Navigation System (BeiDou Satellite Navigation System, BDS) data are fused through the Sage-Husa adaptive filter to enhance the robustness of the positioning system and reduce dependence on a single data source. Finally, an orthogonal projection map matching method based on the three-step method is proposed. This method takes into account the posterior velocity estimate and its standard deviation when determining the search area and performing map matching, and can dynamically adjust the search range and reduce the probability of mismatching. The algorithm significantly improves the accuracy and robustness of train positioning, especially in complex environments, and provides an effective solution for the precise positioning and efficient scheduling of heavy-load railway trains.

[0027] Combine the following Figure 1-Figure 6 Embodiments of the present invention are described in detail.

[0028] Heavy-haul railway trains are designed to transport large amounts of cargo, usually with a significantly higher load capacity than ordinary freight trains. Heavy-haul trains are usually longer in length, which may reach several kilometers. Heavy-haul trains are widely used to transport bulk goods such as coal, ore, steel, and grain, and often carry out long-distance, large-scale logistics transportation. Due to their unique operating characteristics, heavy-haul trains will produce large vibrations and shocks during operation, which will seriously affect the measurement accuracy of the accelerometers and gyroscopes of the inertial navigation system (INS). Frequent vibrations and shocks may cause increased sensor output noise, which in turn affects the accuracy of the inertial navigation system. In addition, heavy-haul trains may experience large acceleration changes due to their weight and speed. Such changes will affect the accelerometer measurements of the INS, especially during acceleration and deceleration, which may cause inertial measurement errors.

[0029] Real-time and accurate acquisition of train location information is particularly important for train control systems. It is not only related to operational efficiency and safety, but also directly affects passengers' travel experience. The operation of trains will be affected by external conditions, such as weather, train operation conditions, train load, changes in track conditions, etc. In other words, the accuracy and continuity of train positioning results directly affect the safety of train travel. Map matching technology plays an important role in achieving precise train positioning. It determines the exact location of the train by matching the real-time location data of the train with the preset railway map. In order to obtain more accurate train positioning and achieve high-precision train safety and efficient scheduling, many scholars have proposed map matching algorithms for railways.

[0030] For the plane characteristics of railway lines, existing map matching algorithms include: 1. The line is monotonically segmented and a digital map database of the line plane and longitudinal section is established. Different methods are used for positioning for different line sections (straight lines, transition curves and circular curve sections): in straight line sections, the GPS position information is directly projected onto the actual line for train positioning; on transition curves, the train position is calculated based on the curvature of the measured current point and the curvature and total length of the transition curve; in circular curve sections, a method for continuous positioning of trains using the same tangent method is proposed. By comparing the tangent direction and GPS data, more accurate positioning can be achieved in circular curve sections. However, directly using unprocessed raw data for map matching positioning will encounter serious noise interference problems, thereby reducing positioning accuracy.

[0031] 2. A method of converting the longitude and latitude data collected by the global navigation satellite system (GNSS) into plane coordinates based on the Gauss-Kruger projection, and then performing segmented matching with the electronic map. First, the GNSS data is converted into plane coordinates through the Gauss-Kruger projection to simplify data processing. Then, the segmented linear track matching method is used to find the nearest projection point of the measured point on the broken line when matching the curve, quickly constrain the measured point to the track, and convert the two-dimensional coordinates into one-dimensional mileage values, reducing the complexity of the model and the amount of subsequent calculations. However, when all the track points in the curve section are uniformly mapped to the same track point, a certain error will occur.

[0032] 3. Train positioning method based on GPS and digital track line matching. The main method includes first generating digital track lines based on the fitting and interpolation algorithm of the train track line, and then correcting the GPS data through the track line positioning correction method to obtain a more accurate train position estimate. When performing map matching, the direct projection method and parallel distance method are used on the straight line to correct the position of the positioning point. This method is easy to calculate and shortens the calculation time for identifying the straight line. The article introduces three map matching methods, namely the projection correction method, the average distance method and the grid matching method, but does not discuss in detail the steps of applying these methods in actual train tracks and how to switch different map matching methods in real time.

[0033] In view of the shortcomings of the above-mentioned map matching algorithm, the present invention proposes the following scheme. The original sensor data is very large, so there will be a large amount of offset data in the data to generate noise interference, especially in the case of long-term tracking. Considering the high timeliness and anti-interference requirements of train positioning, it is crucial to correct the trajectory data during the map matching process. The key features of the curve are retained by data correction, which can enhance the accuracy of the map matching algorithm, because the feature points play a key role in the matching process. In this way, the potential risk of mismatching caused by unclear or redundant nodes on the path can be reduced. By removing non-critical data points through the correction algorithm, the influence of noise and error can be reduced, and the overall quality of the data can be improved. Further, the fusion correction of ODO and BDS data can enhance the robustness of the system to reduce the risk of relying on a single data source, such as BDS signal loss or odometer idling. Finally, considering the characteristics of high-speed train driving, continuous trajectory and the layout of most straight and smooth rails, in order to obtain more accurate position estimation in high-speed trains, the present invention selects orthogonal projection map matching.

[0034] In summary, this patent proposes the ALIGN-OPMM (trajectory alignment and orthogonal projection map matching algorithm) framework, which mainly makes the following contributions: 1) The present invention uses an improved Douglas-Peucker algorithm with time constraints to optimize BDS trajectory data, and then uses the Sage-Husa adaptive filter to give different weights to ODO and BDS according to their respective accuracy and reliability to optimize the positioning results. This method significantly reduces the data processing requirements, thereby improving the computational efficiency and providing an effective data preprocessing strategy for accurate map matching.

[0035] 2) The present invention proposes an Orthogonal Projection Map Matching (OPMM) algorithm based on a three-step method. The algorithm takes into account the posterior velocity estimate and its standard deviation when determining the search area and performing map matching, so that the search range can be dynamically adjusted and the probability of mismatching during the map matching process can be reduced. This increases the stability of the algorithm, especially in the case of high-speed movement or large environmental changes. At the same time, by gradually refining the map matching process, the OPMM algorithm can ensure the continuity and smoothness of the train trajectory, especially in complex or changing railway networks.

[0036] Figure 1 ALIGN-OPMM algorithm framework diagram provided by the present invention. ALIGN-OPMM needs to process two types of data: (1) BDS trajectory data: BDS trajectory data containing noise, processed by a data smoothing and correction module. (2) ODO data: used for fusion with smoothed BDS data. (3) Electronic map data: contains precise location data obtained by pre-marking on the railway track. ALIGN-OPMM outputs a set of trajectory data that falls on the train track after correction. ALIGN-OPMM consists of five main modules to address these challenges, namely: data smoothing and correction module, data fusion module, area selection module, optimal candidate section module and trajectory mapping module, each module is as follows: Data smoothing correction module: solves the noise problem in BDS data. This module uses the Douglas-Peucker algorithm to remove non-critical points, thereby generating smoother and more accurate BDS trajectory data. The output correction data is sent to the data fusion module.

[0037] Data fusion module: solves the challenge of multi-source data fusion. The input is the corrected BDS trajectory data and ODO data. This module uses the Sage-Husa filter fusion algorithm to combine the two data sources, perform a posteriori position estimation, and send the results to the region selection module and the best candidate section module.

[0038] Area selection module: solves the problem of trajectory positioning accuracy. The input is the posterior velocity estimate and its standard deviation, combined with the map database to determine the center map point of the search area , and according to the design factor Narrow the search range of map points. Send a set of map points within a small range to the best candidate road segment module.

[0039] Best Candidate Section Module: It solves the preliminary preparation work of trajectory mapping. The input is the result of the posterior position estimation and the set of map points in a small range sent by the area selection module. This module selects the two map points closest to the posterior position estimation from the map point set and sends them and the posterior position estimation to the trajectory mapping module.

[0040] Trajectory Mapping Module: This module solves trajectory correction and map matching problems. The input is two map points provided by the best candidate road segment and the corrected posterior position estimate. This module uses the orthogonal projection method to perform trajectory mapping and generate the final corrected trajectory data.

[0041] Figure 2 FIG. 1 is one of the flow charts of the heavy-haul railway train track correction method provided by the present invention, such as Figure 2 As shown, the method comprises the following steps: S210: Obtain Beidou satellite navigation system trajectory data, and correct the Beidou satellite navigation system trajectory data.

[0042] According to a heavy-load railway train track correction method provided by the present invention, Beidou satellite navigation system track data is obtained, and the Beidou satellite navigation system track data is corrected, which specifically includes: obtaining the original track curve segment of the train, generating Beidou satellite navigation system track data according to the original track curve segment, and the Beidou satellite navigation system track data is represented as a track point set formed by a group of track points; selecting the first point and the last point in the track point set as two end points to form an initial straight line segment; determining the vertical distance between the middle point and the initial straight line segment; the middle point is the other point on the original track curve segment except the two end points; determining whether there is a segmentation point according to the size of the vertical distance, if there is, determining a new sub-straight line segment according to the segmentation point, and recursively determining all segmentation points based on the new sub-straight line segment; performing time constraint checks on all segmentation points, and sequentially connecting the segmentation points that meet the time constraints to form a simplified representation of the original track curve segment, so as to complete the correction of the Beidou satellite navigation system track data.

[0043] Specifically, in the implementation of high-speed train map matching algorithms, the data correction mechanism is particularly important, mainly because it can cope with the large-scale data streams generated by the rapid sampling of sensors. These sensors record the position and motion parameters of the train at a high frequency, thus generating a large number of data points, which contain a lot of redundant and non-critical information. Data correction reduces noise, identifies more accurate feature points, and optimizes data by intelligently filtering and eliminating these repeated invalid data.

[0044] The data smoothing correction module receives and reads the BDS trajectory data, which is usually in the form of a set of trajectory points. ,in For the The longitude and latitude coordinates of the points, is the timestamp, and n is the number of track points. First, perform a data integrity check to ensure that each point has a valid latitude, longitude and timestamp. Select the first point of the track ) and the last dot As the initial point pair, form the initial straight line segment . Setting Tolerances (maximum permissible deviation distance) and maximum permissible time interval . Then calculate the deviation of the trajectory point set to get the maximum deviation and the corresponding points Then perform a time-space constraint check to check whether the time order of all points complies with , further check the maximum distance difference, for any adjacent point , calculate the Euclidean distance and ensure The following is the specific algorithm flow.

[0045] According to a heavy-duty railway train track correction method provided by the present invention, whether there is a segmentation point is determined according to the size of the vertical distance. If so, a new sub-straight line segment is determined according to the segmentation point, and all segmentation points are recursively determined based on the new sub-straight line segment, specifically including: comparing the vertical distance with a preset tolerance threshold; if the vertical distance is less than the preset tolerance threshold, the initial straight line segment is used as a simplified representation of the original track curve segment; if the vertical distance is not less than the preset tolerance threshold, the point corresponding to the maximum vertical distance is used as the segmentation point; the original track curve segment is divided into two sub-curve segments based on the segmentation point, and sub-straight line segments corresponding to the two sub-curve segments are generated; the maximum vertical distance of the track point relative to the sub-straight line segment is continuously obtained, and whether the sub-curve segment needs to be further segmented is determined according to the maximum vertical distance; and all segmentation points are determined by recursive processing of each sub-curve segment obtained by segmentation.

[0046] Specifically, the tracks on which heavy-load trains run usually pass through remote areas, mountainous areas and complex terrains. The BDS signal coverage in these areas is poor, and BDS data acquisition will be affected by environmental factors, causing data point deviation.

[0047] Therefore, the present invention uses the Douglas-Peucker algorithm (DP algorithm for short) to process BDS data. The Douglas-Peucker algorithm has the characteristics of simplicity and high efficiency, which can help remove these small random deviations and retain smoother trajectories. It can effectively retain key trajectory information while reducing data complexity and processing burden.

[0048] like Figure 3a and Figure 3b As shown in Figure 1, the core of the Douglas-Peucker algorithm is to evaluate the perpendicular distance of the line segments formed by the point-to-point sequence based on a set tolerance threshold. If the distance of a point is less than the tolerance threshold, the point is considered redundant and removed from the trajectory. The process starts from the two endpoints of the trajectory and is recursively applied to each segment on the trajectory until the distance between all retained points is greater than or equal to the tolerance threshold.

[0049] Through this recursive segmentation and evaluation process, the Douglas-Peucker algorithm not only compresses the data volume, but also ensures that the overall shape and key polyline features of the trajectory are preserved. This strategy of balancing the details of the trajectory and the amount of data makes the algorithm particularly suitable for BDS data processing, especially in situations where data needs to be reduced to improve processing speed and storage efficiency, while maintaining the accuracy and integrity of the trajectory.

[0050] The specific steps of the original DP algorithm are as follows: 1. Chord creation: The simplification process is started by drawing a straight line segment, called a "chord", between the starting point A and the end point B of the trajectory curve.

[0051] 2. Calculate the vertical distance: traverse all the middle points on the curve and calculate the vertical distance from each point to AB. Determine the point C farthest from the chord. The maximum distance is expressed as .

[0052] 3. Threshold comparison: judgment . Is it less than the predefined tolerance threshold? ,if , then accept straight line AB as a suitable approximation of the curve segment and complete its simplification.

[0053] 4. Split and recursion: If , then the point As the segmentation point, the original curve AB is divided into two sub-segments AC and CB. Steps 1 to 3 are recursively applied to each sub-segment.

[0054] 5. Polyline reconstruction: After the recursive processing of all curve segments is completed, a polyline is constructed by sequentially connecting the segmentation points. The resulting polyline is used as a simplified representation of the original trajectory curve.

[0055] However, the original algorithm itself does not consider time series information. It only focuses on the distribution of points in space, so it cannot be directly used to process time characteristics. For example, it is impossible to determine whether a point at time k should be behind a point at time k-1. Therefore, the present invention adopts the Douglas-Peucker algorithm with added time constraints, which adds time characteristics to ensure the consistency of time sequence.

[0056] Trajectory points should obey the constraints of physical motion, that is, they should be subject to speed and direction constraints. Figure 4 As shown, The position of the BDS received at any time Should be The received position at the moment within a reasonable space.

[0057] Speed ​​Constraints: The distance traveled by a train is related to the duration, and given a certain time span, the train is physically limited to where it can go. Figure 2 It is described as the area within the dotted circle, from any point p on the circle to the center Maximum distance It's speed max × ,in yes and The timestamp difference between them.

[0058] Direction constraint: A position data S should follow the direction of the train and cannot jump back to the back of F. This area is Figure 2 It is described as the area outside the shadow represented by the diagonal line, where the calculation method is that the position S at time k deviates from its previous position F within a certain angle (default 45°).

[0059] The Douglas-Peucker algorithm with added time constraints reduces storage requirements and processing time by effectively simplifying trajectory data while maintaining the basic geometry of the trajectory. The algorithm can reduce the noise introduced by BDS signal interference or multipath effects by removing points that do not significantly change the trajectory shape, thereby improving data quality. This simplification not only speeds up subsequent processing steps such as data fusion and map matching, but also ensures that important geographic features are retained during the data simplification process.

[0060] S220, fusing the odometer data with the corrected BeiDou satellite navigation system trajectory data to obtain a posteriori position estimate and a posteriori velocity estimate.

[0061] According to a heavy-load railway train trajectory correction method provided by the present invention, odometer data and corrected Beidou satellite navigation system trajectory data are fused, specifically including: obtaining the timestamp range of the corrected Beidou satellite navigation system trajectory data and the odometer data, and determining the common time range of the two; linearly interpolating the odometer data to align the Beidou satellite navigation system trajectory data and the odometer data in the time dimension, and generating a fused data set with the same timestamp; dynamically adjusting the statistical characteristics of system noise and observation noise based on the fused data set through the Sage-Husa adaptive filtering algorithm, and outputting the fused data.

[0062] Specifically, in the trajectory correction and orthogonal projection map matching algorithm of heavy-duty railway trains, data fusion is one of the key steps, aiming to improve the accuracy and robustness of the positioning system. Data fusion mainly combines sensor data from different sources to obtain more reliable location information. Kalman filtering is a well-known variable optimal estimation algorithm, which is widely used in many fields such as track calculation, target tracking and navigation systems. However, in the case of inaccurate system model or non-Gaussian distribution of noise statistical characteristics, the performance of Kalman filter may decline or even diverge. In order to improve this divergence problem, the Sage-Husa adaptive filtering algorithm is used in the present invention to fuse BDS (Beidou Satellite Navigation System) and ODO (odometer) data.

[0063] 2.1. Time synchronization.

[0064] The preprocessing stage of sensor data fusion begins with the alignment (synchronization) of timestamp data. The processed BeiDou Navigation Satellite System (BDS) and odometer (ODO) data are time synchronized using a linear interpolation method.

[0065] (1) Timestamp correction: Determine the timestamp range of BDS data and ODO data and find the common time range of the two.

[0066] (2) Linear interpolation: Linear interpolation is performed on the ODO data to align its timestamp with the BDS data. For example, suppose the ODO data is at time and Missing data points between , then interpolation is performed using the following formula:

[0067] Similarly, interpolation is performed on the BDS data to generate a fused dataset with the same timestamp.

[0068] 2.2 Sage-Husa filter fusion

[0069] The Sage-Husa filter performs recursive estimation using observed data and adjusts and corrects the statistical characteristics of system noise and observation noise in real time through a time-varying statistical estimator. This dynamic adaptation not only reduces the deviation of the model, but also suppresses the filter divergence, thereby improving the overall accuracy of the filtering process.

[0070] The prediction process of the Sage-Husa filter is consistent with the prediction process of the Kalman filter.

[0071]

[0072] It's time The prior state estimate of is the state transition matrix, yes The posterior state estimate at time t, is the prediction covariance matrix, is the transpose of the state transfer matrix F, is the prior covariance estimate, is the process noise covariance matrix.

[0073] However, compared to the Kalman filter, the Sage-Husa filter has an additional step in the correction phase to adjust for the statistical properties of the noise.

[0074]

[0075] in, is the residual, i.e. the observed value The difference between the predicted value and represents the observed value at the current moment, , represents the observation noise covariance matrix, represents the identity matrix, represents the Kalman gain at the previous moment, For the forgetting factor, , the value of b is usually between 0.955 and 0.995.

[0076] Kalman gain calculation:

[0077] Status Update:

[0078] Covariance update:

[0079] Through the above steps, the Sage-Husa adaptive filter can effectively fuse BDS and ODO data to generate more accurate location information. This fusion method not only improves the system's anti-noise ability, but also dynamically adapts to environmental changes to ensure the accuracy and continuity of train positioning. This module will pass the fused trajectory data to the area selection module and the best candidate section module.

[0080] S230, determining the central map point of the search area based on the posterior velocity estimate and its standard deviation in combination with the map database; narrowing the search range of map points centered on the central map point based on the design factor to obtain a set of map points within the search range.

[0081] According to a heavy-haul railway train track correction method provided by the present invention, a map point search range centered on a central map point is narrowed according to a design factor to obtain a map point set within the search range, specifically comprising: taking the central map point as the center, determining the boundary of the search range according to the design factor and the potential moving distance; determining the potential moving distance according to a posterior velocity estimate and its standard deviation; determining a search area according to the boundary of the search range, and within the determined search area, screening all map points located within the search area from a map database to form a map point set.

[0082] Specifically, the core of the map matching algorithm is to enhance the accuracy and stability of positioning by constraining the potential location of the mobile entity to a small area of ​​the map. This method is achieved by comparing and matching the actual location of the mobile entity with an approximate point or a set of approximate points on the map. The map matching strategy can be further developed into point-to-curve matching, which not only involves the correspondence between points, but also includes finding the curve segment closest to the estimated location on the map. This method takes into account the geometry and spatial layout of the road in more depth, thereby improving the accuracy of the matching process.

[0083] Taking into account the characteristics of high-speed train travel, the layout of most straight and smooth rails, and the high timeliness requirement for positioning, in order to obtain more accurate position estimates in high-speed trains, an OPMM algorithm based on a three-step method is proposed. The present invention uses an orthogonal projection map matching algorithm based on a three-step method. The train orthogonal projection mapping matching (OPMM) algorithm uses map points on a railway map, which are spaced one meter (1m) along the railway track and define the position information in the ECEF geodetic coordinate system. In the first step, the posterior velocity estimate and its standard deviation after data fusion are used to select the search area for map matching. In the second step, two map points one meter apart that are closest to the posterior position estimate are selected in the search area. Finally, the OPMM algorithm is used to calculate the position of the train on the track. This algorithm aims to optimize the train positioning process by effectively compressing trajectory data and adopting an orthogonal projection method to improve the accuracy and efficiency of positioning.

[0084] 3.1. Select the search area.

[0085] After obtaining the sensor data processed by the Sage-Husa filter, the posterior position estimate is obtained. However, in general, this estimate may not accurately reflect the actual position on the railway track due to measurement errors. In order to improve the accuracy of the position estimate, the present invention adopts a strategy of projecting the posterior position estimate onto the actual track of the train. This process first involves defining a search area for map points to determine the unit vector of the projection. The key to determining the search area is to find the center map point The point is based on the train's time interval Specifically, considering the posterior velocity estimate and its standard deviation , the present invention can calculate the central map point by the following formula (10)(11)(12) .

[0086]

[0087] in, and are the two closest map points at k−1, l is the index of the map point where the train is currently located, is a function that rounds the element x to the nearest integer to infinity, For the railway track Map points.

[0088] The unit vector between two map points For vector projection in The present invention can be used or The substitution formula (11) , however their performance is very similar, but using It can reduce search computing power.

[0089] Next, The search area centered on and As shown in formula (13), the size of the search area depends on the design factor selected by the user. .

[0090]

[0091] Note that the total number of map points in the search area is equal to .

[0092] S240: Select two map points closest to the posterior position estimate from the set of map points within the search range.

[0093] Specifically, 3.2, confirm the best candidate road section.

[0094] In this step, a method is described for determining the a posteriori position estimate The two closest map points are recorded as and The method aims to calculate First, the distance between the posterior position estimate and each map point in the search area is calculated to identify the closest point. , the process is shown in equation (14). Then, evaluate Adjacent map points to determine the point with the shortest distance , as described in equation (15). Finally, based on and Determine and The two nearest map points, i.e. and , as defined in equation (17). This pair of map points will be used in the next projection calculation as the unit vector to determine the projection direction.

[0095] For each map point in the search area, calculate its The distances are as follows: For all arrive ,calculate

[0096] Find out The two nearest map points

[0097] This computational process ensures that the map point closest to the posterior position estimate can be accurately identified, providing an accurate benchmark for orthogonal projection mapping.

[0098] S250 , performing trajectory mapping according to the two map points and the posterior position estimation to generate corrected trajectory data.

[0099] According to a heavy-haul railway train track correction method provided by the present invention, track mapping is performed according to two map points and a posteriori position estimate to generate corrected track data, specifically comprising: determining a unit vector according to the two map points; orthogonally projecting the a posteriori position estimate onto a track line determined by the two map points according to the unit vector to obtain a projection result; using the projection result as an accurate position estimate of the train on the railway track; and encapsulating the accurate position estimate together with other relevant parameters to obtain the corrected track data.

[0100] Specifically, 3.3, correct trajectory.

[0101] Get position estimate via orthogonal projection mapping is the key output of this algorithm. The process starts with two mapping points determined from equation (17) and , thus constructing the unit vector This unit vector not only indicates the direction of the railway track, but also provides a basis for subsequent projection calculations.

[0102] Specifically, the posterior position estimate is orthogonally projected onto the vector This process is accurately described by equation (18). This equation quantifies The length of the projection onto the track vector determines the exact location of the projected point on the railway track. This calculation not only focuses on the magnitude of the projection, but also accurately locates the position of the train on the track.

[0103] Then, equation (19) is used to calculate the final projection position estimate This involves not only Projecting onto the trajectory, you also need to add a reference point , to ensure the accuracy and practicality of the projection position. In this way, It accurately reflects the actual position of the train on the railway track, thereby greatly improving the accuracy and reliability of position estimation.

[0104]

[0105] In this way, It accurately reflects the actual position of the train on the railway track, thereby greatly improving the accuracy and reliability of position estimation.

[0106] In view of the data noise and processing efficiency problems encountered by existing map matching algorithms in train positioning, the present invention proposes an ALIGN-OPMM algorithm framework that integrates data correction, fusion and map matching. Through BDS data correction, the present invention effectively reduces the noise and deviation points in the data, removes redundant data, and significantly improves the overall quality and processing speed of the data. The data fusion step enhances the system's adaptability to changes in various sensor data, especially when the BDS signal is unstable or lost, it can ensure the continuity and accuracy of positioning. The orthogonal projection map matching algorithm (OPMM) considers the posterior velocity estimate and its standard deviation to dynamically adjust the search area, effectively reducing mismatches, especially in complex environments. The present invention not only improves the accuracy of map matching, but also optimizes the computational efficiency, and is suitable for use in the precise positioning system of heavy-duty railway trains.

[0107] The heavy-haul railway train track correction device provided by the present invention is described below. The heavy-haul railway train track correction device described below and the heavy-haul railway train track correction method described above can be referred to each other.

[0108] like Figure 5 The present invention provides a heavy-duty railway train track correction device, comprising: The data smoothing correction module 510 is used to obtain the trajectory data of the Beidou satellite navigation system and correct the trajectory data of the Beidou satellite navigation system; A data fusion module 520 is used to fuse the odometer data with the corrected Beidou satellite navigation system trajectory data to obtain a posteriori position estimate and a posteriori velocity estimate; The area selection module 530 is used to determine the center map point of the search area based on the posterior velocity estimate and its standard deviation in combination with the map database; narrow the map point search range centered on the center map point according to the design factor to obtain a set of map points within the search range; The best candidate road segment determination module 540 is used to select two map points closest to the posterior position estimate from the set of map points within the search range; The trajectory mapping module 550 is used to perform trajectory mapping according to two map points and the posterior position estimate to generate corrected trajectory data.

[0109] Figure 6 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 6As shown, the electronic device may include: a processor 610, a communication interface 620, a memory 630 and a communication bus 640, wherein the processor 610, the communication interface 620 and the memory 630 communicate with each other through the communication bus 640. The processor 610 may call the logic instructions in the memory 630 to execute the heavy-haul railway train track correction method, which includes: obtaining Beidou satellite navigation system track data and correcting the Beidou satellite navigation system track data; fusing the odometer data and the corrected Beidou satellite navigation system track data to obtain a posteriori position estimate and a posteriori speed estimate; determining the center map point of the search area according to the posteriori speed estimate and its standard deviation in combination with the map database; reducing the map point search range centered on the center map point according to the design factor to obtain a map point set within the search range; selecting two map points closest to the posteriori position estimate from the map point set within the search range; performing track mapping according to the two map points and the posteriori position estimate to generate corrected track data.

[0110] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0111] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the heavy-load railway train trajectory correction method provided by the above-mentioned methods, which includes: obtaining Beidou satellite navigation system trajectory data and correcting the Beidou satellite navigation system trajectory data; fusing the odometer data and the corrected Beidou satellite navigation system trajectory data to obtain a posteriori position estimate and a posteriori speed estimate; determining the center map point of the search area based on the posteriori speed estimate and its standard deviation in combination with a map database; narrowing the search range of map points centered on the center map point based on a design factor to obtain a set of map points within the search range; selecting two map points closest to the posterior position estimate from the set of map points within the search range; performing trajectory mapping based on the two map points and the posterior position estimate to generate corrected trajectory data.

[0112] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the heavy-load railway train trajectory correction method provided by the above-mentioned methods, the method comprising: obtaining Beidou satellite navigation system trajectory data, and correcting the Beidou satellite navigation system trajectory data; fusing the odometer data and the corrected Beidou satellite navigation system trajectory data to obtain a posteriori position estimate and a posteriori speed estimate; determining the center map point of the search area based on the posteriori speed estimate and its standard deviation in combination with a map database; narrowing the search range of map points centered on the center map point based on a design factor to obtain a set of map points within the search range; selecting two map points closest to the posterior position estimate from the set of map points within the search range; performing trajectory mapping based on the two map points and the posterior position estimate to generate corrected trajectory data.

[0113] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0114] Through the description of the above implementation modes, those skilled in the art can clearly understand that each implementation mode can be implemented by means of software plus a necessary general hardware platform, or of course by hardware. Based on such an understanding, the above technical solution can essentially or in other words be embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A heavy-haul railway train track correction method, characterized in that: include: Obtaining Beidou satellite navigation system trajectory data, and correcting the Beidou satellite navigation system trajectory data; The odometer data and the corrected Beidou satellite navigation system trajectory data are fused to obtain a posteriori position estimate and a posteriori velocity estimate; Determining a center map point of a search area based on the a posteriori velocity estimate and its standard deviation in combination with a map database; narrowing the map point search range centered on the central map point according to the design factor to obtain a set of map points within the search range; Selecting two map points closest to the posterior position estimate from the set of map points within the search range; Trajectory mapping is performed based on the two map points and the posterior position estimate to generate corrected trajectory data.

2. The heavy-haul railway train track correction method according to claim 1, characterized in that: The obtaining of Beidou satellite navigation system trajectory data and correcting the Beidou satellite navigation system trajectory data specifically includes: Acquire an original track curve segment of the train, and generate Beidou satellite navigation system track data according to the original track curve segment, wherein the Beidou satellite navigation system track data is represented as a track point set formed by a group of track points; Select the first point and the last point in the trajectory point set as two end points to form an initial straight line segment; Determine the vertical distance between the middle point and the initial straight line segment; the middle point is the other point on the original trajectory curve segment except the two end points; Determine whether there is a segmentation point according to the size of the vertical distance, and if so, determine a new sub-straight line segment according to the segmentation point, and recursively determine all segmentation points based on the new sub-straight line segment; A time constraint check is performed on all segmentation points, and segmentation points that meet the time constraint are sequentially connected to form a simplified representation of the original trajectory curve segment to complete the correction of the Beidou satellite navigation system trajectory data.

3. The heavy-haul railway train track correction method according to claim 2, characterized in that: The determining whether there is a segmentation point according to the size of the vertical distance, and if there is, determining a new sub-straight line segment according to the segmentation point, and recursively determining all segmentation points based on the new sub-straight line segment, specifically includes: comparing the vertical distance to a preset tolerance threshold; If the vertical distance is less than the preset tolerance threshold, the initial straight line segment is used as a simplified representation of the original trajectory curve segment; If the vertical distance is not less than the preset tolerance threshold, the point corresponding to the maximum vertical distance is used as the segmentation point; The original trajectory curve segment is divided into two sub-curve segments based on the segmentation point, and sub-straight line segments corresponding to the two sub-curve segments are generated; Continue to obtain the maximum vertical distance of the trajectory point relative to the sub-straight line segment, and determine whether the sub-curve segment needs to be further segmented according to the maximum vertical distance; Continue to recursively process each sub-curve segment obtained by segmentation to determine all segmentation points.

4. The heavy-haul railway train track correction method according to claim 1, characterized in that: The odometer data and the corrected Beidou satellite navigation system trajectory data are fused, specifically including: Obtain the timestamp range of the corrected Beidou satellite navigation system trajectory data and odometer data, and determine the common time range of the two; Linearly interpolating the odometer data to align the Beidou satellite navigation system trajectory data and the odometer data in the time dimension, generating a fused data set with the same timestamp; The statistical characteristics of system noise and observation noise are dynamically adjusted based on the fused data set through the Sage-Husa adaptive filtering algorithm, and the fused data is output.

5. The heavy-haul railway train track correction method according to claim 1, characterized in that: The step of narrowing the map point search range centered on the central map point according to the design factor to obtain a set of map points within the search range specifically includes: Taking the central map point as the center, determining the boundaries of the search range according to the design factor and the potential moving distance; the potential moving distance is determined according to the a posteriori velocity estimate and its standard deviation; A search area is determined according to the boundary of the search range, and within the determined search area, all map points within the search area are screened from a map database to form a map point set.

6. The heavy-haul railway train track correction method according to claim 1, characterized in that: The performing trajectory mapping according to the two map points and the a posteriori position estimation to generate corrected trajectory data specifically includes: determining a unit vector based on the two map points; orthogonally projecting the posterior position estimate onto a trajectory line determined by the two map points according to the unit vector to obtain a projection result; Using the projection result as an accurate position estimate of the train on the railway track; The precise position estimate and other relevant parameters are packaged together to obtain corrected trajectory data.

7. A heavy-haul railway train track correction device, characterized in that: include: A data smoothing correction module is used to obtain Beidou satellite navigation system trajectory data and correct the Beidou satellite navigation system trajectory data; A data fusion module is used to fuse the odometer data with the corrected Beidou satellite navigation system trajectory data to obtain a posteriori position estimate and a posteriori velocity estimate; an area selection module, for determining a center map point of a search area based on the a posteriori velocity estimate and its standard deviation in combination with a map database; narrowing the map point search range centered on the central map point according to the design factor to obtain a set of map points within the search range; An optimal candidate road segment determination module, configured to select two map points closest to the posterior position estimate from the set of map points within the search range; The trajectory mapping module is used to perform trajectory mapping according to the two map points and the a posteriori position estimate to generate corrected trajectory data.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the heavy-haul railway train trajectory correction method as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the heavy-haul railway train track correction method as claimed in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the heavy-haul railway train track correction method as claimed in any one of claims 1 to 6 is implemented.

Images