JPWO2019154720A5 - - Google Patents

Description

The invention relates in particular to the use of railway vehicles for their construction, their maintenance, their maintenance, their repair, their renewal or their demolition, for auscultation of railway tracks or for work on railway tracks. It also provides early localization of points of interest on railroad tracks and appropriate interest for subsequent intervention of intervention tools on railroad tracks that are carried and steered by machines running on railroad tracks. It also relates to demarcating regions. It also relates to such localization transposition for subsequent use, in particular for later auscultation of the track by an auscultation device or for later intervention on the track by an interventional tool.

PRIOR ART Document US4986189 describes an intervention machine for the maintenance or repair of railroad tracks, which is intended to operate while the machine is moving forward on the railroad tracks in the direction of travel of the work. and intervention tools. A measurement beam is located in front of the machine to predict the presence of obstacles and to allow automated command of tool positioning. The measurement beam is arranged horizontally and vertically with respect to the direction of the track and has sensors aligned to detect the lateral positioning of the rail. Another sensor built by a camera monitors obstacles on the track. The measuring beam also carries an odometer. The sensor signal is sent to a circuit for commanding the tool with a delay that is a function of the odometer signal and the predetermined distance between the measurement beam and the tool. It is therefore possible to arrange the measurement beams at a distance from the tool without the risk of interference with them. However, this mode of operation assumes that a certain distance between the measuring beam and the tool is precisely known. As long as the desired accuracy of tool positioning is centimeters, it is necessary to provide a very rigid common chassis for supporting the interventional tools and the measuring beams in order to create a common frame of reference. It also assumes that the perpendicularity of the measurement beam to the line is correct. Additionally, mileage errors linked to touch probe wheel sliding or angular inscription, etc. relative to the rail neutral line are cumulative. Finally, the system does not manage curved tracks.

The paper "A High-Performance Inspection and Maintenance System of Track using Continuous Scan Image" by Masato Ukai and Nobuhiko Nagahara, 11th WCRR 2016 describes an early analysis system for railway tracks aimed at later intervention by maintenance vehicles. . The analysis system is mounted on a dedicated rail vehicle capable of circulating the track at a maximum speed of 45 km/h. The analysis system utilizes a linear camera coupled to the odometer and mounted on the vehicle across the track to synchronize the odometer pulses with the linear shots by the linear camera. The system allows the construction of continuous bitmap images (2D) of railway tracks. Analysis of bitmap images, which is not performed in real time, detects objects with a given signature, in particular obstacles, to identify areas of the track where intervention is possible and "prohibited" areas where automated intervention is not possible. ” area can be determined. Periodically resetting the odometer signal by recognizing predetermined markers placed on the track and whose absolute position is known, in order to compensate for odometer measurement errors, especially on curves It is proposed to The analysis system makes it possible to generate intervention programs that can later be executed by maintenance machine tools carried by maintenance vehicles that circulate on railroad tracks. However, the use of later measurements by maintenance vehicles requires the latter to have available its own means of positioning relative to the track. Furthermore, potential perpendicularity defects between the linear camera and the track are not taken into account.

The present invention, in particular, overcomes the drawbacks of the prior art and enables accurate early localization of points of interest or lines on a railroad track, where appropriate by intervention machines traveling on the railroad track. The aim is to propose a means of delimiting a region of interest for later intervention of a transported and driven tool.

To this end, a method for locating a railroad track according to the first aspect of the invention, comprising at least one linear camera facing the railroad track and one or more odometers. The railroad locating system is carried out by a railroad locating system equipped with a railroad locating system traveling in the direction of travel on the railroad track and the locating method performs the following actions:
- using the odometer(s) to repeatedly acquire progress data of the railroad localization system on the railroad track in the direction of travel;
- repeatedly acquiring instantaneous linear optical data along the instantaneous measurement line with a linear camera pointing at the railway track;
- constructing a bitmap image of the area of the railroad track surface by processing at least the instantaneous linear optical data and, where appropriate, the progress data;
- identifying at least one spatial index marker of a predetermined signature in the constructed bitmap image by processing the constructed bitmap image; determining the positioning of the spatial index marker relative to the curvilinear abscissa and the railroad track reference line;
- identifying a point of interest or line in the constructed bitmap image and determining the coordinates of the point of interest or line in a two-dimensional localization reference frame linked to the spatial index marker and the reference line; include.

Linear cameras offer the advantage of being very insensitive to parasitic movements, especially vibrations experienced by a localization system traveling on a railroad track, and this advantage is due to the fact that the localization system is fixed to an intervention machine on the railroad track. or may be further emphasized when hooked. Furthermore, the formation time of pixels for linear images is short compared to matrix techniques. Thus, positioning latency and uncertainty that can be caused by long exposures and excessive build times are reduced.

A dataset is constructed by defining the coordinates of points of interest or lines using the spatial index markers and the reference line as a local frame of reference, which is later exploited by the on-board transposition system itself, and the spatial index markers and a baseline can be identified.

Linear optical data acquired at a given instant corresponds to a line in the bitmap image. The construction of bitmap images from linear optical data is straightforward, as there is no overlap problem between successive images that would be inherent in acquisition with a matrix camera. The curve is also "naturally" modified with a bitmap image constructed from the instantaneous linear data of the linear camera. This allows for a simplified, nevertheless operator-appropriate representation.

According to one embodiment, acquisition of instantaneous linear optical data is triggered by receipt of progress data.

If the spatial resolution of the odometer is high, it is possible to trigger the acquisition line of linear optical data with each pulse of the odometer, or every non-zero integer N pulses, thereby: It is possible to have a constant spatial step for continuous acquisition of lines of linear optical data.

According to one embodiment, the instantaneous linear optical data and progression data are acquired in a synchronized manner.

In particular, if the spatial resolution of the odometer is below the desired resolution for constructing the bitmap image, for example observing the time interval T between two successive pulses I−1 and I of the odometer, , dividing this time interval by a predetermined non-zero integer N, it is possible to trigger the linear optical data acquisition line at a constant time interval T/N between odometer pulses I and I+1. . The number of measurement lines is counted up to pulse I+1, and a posteriori the step of spatial acquisition of linear optical data in the direction of travel between pulses I and I+1 is deduced therefrom. In practice, the variability of the travel speed of the localization system is low on the observed scale, and the assumption of constant speed between two consecutive pulses of the odometer leads to negligible deformation of the bitmap image. More generally, whatever the algorithm holds, it is possible to trigger the linear camera, for example by an electronic card or a dedicated software block for generating the pulse train.

According to one embodiment, the progression data and the instantaneous linear optical data are time stamped and a bitmap image is preferably constructed as a function of the time stamps. Therefore, from the spatial resolution of the odometer and the knowledge of the time interval observed between two pulses of the odometer and between two measurement lines of the linear camera, the potential A spatial step can be determined that is variable to .

According to one embodiment, the points of interest or lines constitute the boundaries of the vertices of the region of interest, preferably the quadrilateral that constitutes the region of interest. Preferably, the localization system is able to identify in the bitmap image intersections between sleepers or neutral lines of sleepers on the one hand and rails or neutral rail lines of railway tracks on the other, these intersections being of interest. Form at least part of a point or line.

The coordinates of the point of interest may be Cartesian coordinates of a reference frame with the X axis as the reference line, the Y axis as the axis perpendicular to the reference line, and the origin at the spatial index marker . In this hypothesis, the coordinates include the distance to the reference line measured perpendicular to the reference line and the distance to the spatial index marker measured parallel to the reference line.

The reference line may be locally determined by at least two points of data, or by points of data and a steering vector, or by any other equivalent means. According to one embodiment, the reference line is the neutral line of one of the rails of the railway track or a line constructed from the neutral line of the rails of the railway track, for example the center line between two rails of the railway track. be. In practice, it is possible to determine the neutral line of the rail, for example, by locating the edge of the rail on a bitmap image constructed from instantaneous linear data. It is also possible to resort to an orientation matrix camera, as explained below.

Preferentially, an orientation device of the railroad localization system iteratively determines angular orientation data of the railroad localization system with respect to the reference line, and coordinates of points of interest or lines are determined as a function of the orientation data. With knowledge of the angle between the linear camera and the reference line in the projection plane of the bitmap image, it is possible to modify the coordinates of the interest points and spatial index markers. In fact, if there is a non-perpendicular angle between the linear camera and the reference line, the linear camera's apparent distance between the point of interest and the rail will be greater than the actual distance measured perpendicular to the reference line. . Naturally, the distance between two adjacent pixels of the linear camera is known and specified by the calibration parameters.

According to an embodiment at least one first touch probe for detecting orientation of the localization system with respect to a first rail of the railway track, wherein the orientation device of the localization system constitutes the first orientation rail and preferably comprising a second touch probe for detecting orientation of the localization system with respect to a second rail of the railroad track, constituting a second orientation rail.

According to another embodiment, the orienting device of the localization system is disposed facing a first one of the rails of the railway track, constituting the first orienting rail. An orientation matrix camera is provided, the orientation device taking shots with the first orientation matrix camera and processing the shots to detect the orientation of the first orientation rail relative to the first orientation matrix camera target in the shots. a second orientation matrix camera disposed facing a second rail of the railway track, preferably constituting a second orientation rail, the orientation device taking shots with the second orientation matrix camera; and processing the shot to detect the orientation of the second orientation rail relative to the target of the second orientation matrix camera within the shot. A shot of the linearly oriented camera(s) also allows the reference line to be determined. Various options are provided for approximating and adjusting the data from the linear and orientation camera(s). According to a first method, it is possible to calibrate the railway localization system such that the calibration parameters characterize its relative positioning with an orientation device constituted by a linear camera and an orientation camera(s). is. From these calibration data, it is possible to transpose the rail guide lines detected in the orientation camera image into a bitmap image constructed from the instantaneous linear optical data. This first method can be implemented without detecting the rail neutral line, or more generally the reference line, in the bitmap image constructed from the instantaneous linear optical data. Conversely, it is possible to exploit only the angular orientation data provided by the orientation device, and the reference line is calculated only from the data of the bitmap image constructed from the instantaneous linear data.

Where appropriate, correction may be made by inertial measurements performed at the midpoint of the beam, or by optical velocimetry techniques (which means taking close matrix images with an orientation camera). These approaches can be used to select a suitable image (low specific specific rotational power) and also determine the average angle of inscription of the beam with respect to the rail.

Where appropriate, it is possible to mix cameras and touch probes on the same rail or on two different rails. Therefore, it is possible to perform an interpolation between the orientation data delivered by the two orientation matrix cameras to estimate the orientation of the linear camera. The presence of at least one sensor ( touch probe or matrix camera) above each rail of the railway track allows the rail to be detected on a section of railway track, especially when the localization system travels along a track device, e.g. a point. It is also possible to offset what is not there. It is also possible to choose a particular rail orientation in the curve, for example a rail located inside the curve.

Preferably the orientation data is time-stamped.

According to one embodiment, the localization system distinguishes regions of the railroad track surface comprising railroad ties and inter-sleeper regions, and the orientation device provides orientation data only once for each of the inter-sleeper regions. Only deliver. Since the variation in orientation between two sleepers is small, it is best to limit the amount of data processed.

Preferably the orientation data is used to determine the reference line.

A reproduction of the bitmap image can be provided on the display screen of the localization system to allow human intervention on the identified points of interest or regions of interest. Preferably, visual identification on the display screen of the interest point or line is provided.

Providing activation and/or deactivation of at least some of the points of interest or lines or regions of interest, or identification of the regions of interest as potential interventions or forbidden regions, following input on the human-machine input interface Is possible.

According to another aspect of the invention, the invention is a positioning method performed by a measurement assembly comprising a localization system and a transposition system, the localization system comprising a linear camera and an odometer. , wherein the transposition system comprises one or more transposition matrix cameras positioned behind the linear cameras of the localization system at a distance and in the direction of travel, the method being performed by the localization system, wherein and then performed by the transposition system, the following actions:
- acquiring a set of one or more transposed matrix images in the spatial reference frame of the transposed system using the transposed matrix camera(s);
- identifying spatial index markers in a set of one or more transposed matrix images as a function of progression data acquired by the odometer; and data of the spatial index markers and baselines in the spatial reference frame of the transposed system determining the coordinates of the property;
- the transposed coordinates of the point of interest or line in the spatial reference frame of the transposed system and of the point of interest or line in the localization frame of reference as a function of the spatial index marker and the coordinates of the data characteristic of the reference line in the spatial reference frame of the transposed system; A method comprising calculating coordinates and a transposing step.

According to another aspect of the invention, the invention comprises an interventional tool that travels along a railroad track and is mounted to an interventional chassis by a measurement assembly that includes a localization system and a transposition system. a method of driving an interventional machine, wherein the localization system comprises at least one linear camera and an odometer, and the displacement system moves at a distance in the direction of travel behind the linear camera of the localization system. comprising at least one transposed matrix camera positioned, the transposed matrix camera being integral with the interventional chassis of the interventional tool, the method comprising the positioning procedure as described above and the coordinates of the points of interest or lines in the transposed reference frame; an interventional procedure including positioning of an interventional tool as a function.

The interventional tool can be of any type, such as a tamping tool or a ramming tool.

Methods according to this aspect of the invention may include various embodiments that reproduce the characteristics of all or some of the embodiments described in relation to the first aspect of the invention.

According to another aspect of the invention, the invention relates to a rail vehicle comprising a localization system comprising a linear camera and an odometer, the localization system incorporating one of the aforementioned methods as an alternative One or the other can be implemented.

According to another aspect of the invention, the invention is a method of locating a railroad track, performed by a railroad locating system traveling in a direction of travel on the railroad track, the method comprising: action:
- repeatedly acquiring instantaneous linear optical data along the instantaneous measurement line with the linear camera of the railway localization system facing the railway track;
- repeatedly obtaining orientation data of the railroad localization system with respect to the reference line of the railroad track using the railroad localization system's orientation device;
- constructing a potentially distorted bitmap image of the surface area of the railroad track by processing at least the instantaneous linear optical data;
- identifying interest points or lines in a potentially distorted bitmap image;
- determining the corrected coordinates of the interest point or line as a function of the interest point or line in the frame of reference of the potentially distorted bitmap image and the potentially distorted coordinates of the orientation data. Regarding.

Linear cameras offer the advantage of being very insensitive to the vibrations experienced by the locating system traveling on the railroad track, which is advantageous if the locating system is fixed to an intervention machine on the railroad track or hooked. can be further emphasized when

Instantaneous linear optical data acquired at a given instant corresponds to a line in the bitmap image. As the localization system progresses, the instantaneous measurement line is displaced relative to the railroad track.

With knowledge of the angle between the linear camera and the reference line in the projection plane of the bitmap image, it is possible to modify the coordinates of the interest points and spatial index markers. In fact, if there is a non-perpendicular angle between the linear camera and the reference line, the linear camera's apparent distance between the point of interest and the rail will be greater than the actual distance measured perpendicular to the reference line. .

According to one embodiment, to obtain orientation data, a first touch probe of the orientation device detects orientation of the localization system with respect to a first rail of the railway track, which constitutes the first orientation rail. Preferably, a second touch probe of the orientation device shall detect the orientation of the localization system with respect to a second rail of the railway track, which constitutes the second orientation rail.

According to another embodiment, a first orientation device arranged facing a first one of the rails of the railway track, constituting the first orientation rail, for obtaining orientation data takes a shot processed by the orientation device and detects in the shot the orientation of the first orientation rail relative to the target of the first orientation matrix camera and preferably the second orientation rail A second orientation matrix camera arranged facing a second one of the rails of the railroad track, constituting a Let us detect the orientation of the second orientation rail with respect to the target of the matrix camera.

Where appropriate, it is possible to mix cameras and touch probes on the same rail or on two different rails. Therefore, it is possible to perform interpolation between the orientation data delivered by the two orientation matrix cameras to estimate the orientation of the linear camera. The presence of at least one sensor ( touch probe or matrix camera) on each rail of the railway track allows the rail to be detected on a section of railway track, especially when the localization system travels along a track device, e.g. a point. It is also possible to offset what is not there. It is also possible to choose a particular rail orientation in the curve, for example a rail located inside the curve.

Preferably the orientation data is time-stamped.

According to one embodiment, the localization system detects an area of a surface of a railroad track with sleepers and an inter-sleeper area of the surface of the track, and an orientation device generates orientation data once for each inter-sleeper area. is only delivered. Since the directional variation between the two sleepers is small, it is best to limit the amount of data to process.

Preferably, the reference line is the neutral line of one of the rails of the railway line or a line constructed from the neutral line of the rails of the railway line. Preferably, the orientation data is used to construct a reference line from the rail neutral line of the railroad track.

According to a particularly advantageous embodiment, the one or more odometers of the railroad localization system repeatedly acquire progress data of the railroad localization system on the railroad tracks in the direction of travel.

According to a first implementation of this embodiment, acquisition of instantaneous linear optical data is triggered by receipt of progress data.

If the spatial resolution of the odometer is high, it is possible to trigger a line of linear optical data acquisition on each pulse of the odometer, or on every non-zero integer N pulses, thereby , making it possible to have a constant spatial step for continuous acquisition of lines of linear optical data.

According to a second implementation of this embodiment, instantaneous linear optical data and progression data are acquired in a synchronized manner.

In particular, if the spatial resolution of the odometer is less than the desired resolution for constructing the bitmap image, the time interval T between two consecutive pulses I−1 and I of the odometer is observed to determine the predetermined It is possible to divide this time interval by a non-zero integer N of , and trigger a line of linear optical data acquisition at a constant time interval T/N between odometer pulses I and I+1. The number of lines of measurement up to pulse I+1 is counted, and a posteriori the step of spatial acquisition of linear optical data in the direction of travel between pulses I and I+1 is deduced therefrom. In practice, the variation of the travel speed of the localization system is small on the observed scale, and the assumption of constant speed between two consecutive pulses of the odometer leads to negligible deformation of the bitmap image.

According to one embodiment, the progression data and the instantaneous linear optical data are time stamped and a bitmap image is preferably constructed as a function of the time stamps. Therefore, from the spatial resolution of the odometer and the knowledge of the time interval observed between two pulses of the odometer and between two measurement lines of the linear camera, the potential A spatial step can be determined that is variable to .

According to a particularly advantageous embodiment, the localization method comprises:
- by processing the potentially distorted bitmap image, at least one spatial index marker of a predetermined signature is identified in the potentially distorted bitmap image;
- by processing the progress and orientation data, the curvilinear abscissa of the spatial index marker relative to the reference line of the railroad track is determined;
- The modified coordinates of the point of interest or line are determined in the spatial index marker and the local 2D localization reference frame linked to the reference line.

A dataset is constructed by defining the coordinates of points of interest or lines using the spatial index markers and the reference line as a local frame of reference, which is later exploited by the on-board transposition system itself, and the spatial index markers and a baseline can be identified.

According to one embodiment, the points of interest or lines constitute the boundaries of the vertices of the region of interest, preferably the quadrilateral that constitutes the region of interest. Preferably, the localization system is capable of identifying, in the bitmap image, points of intersection between sleepers or sleeper neutral lines on the one hand and rails or rail neutral lines of railroad tracks on the other hand, these points of intersection being Forming at least part of the interest point or line.

The coordinates of the point of interest may be Cartesian coordinates of a reference frame with the X axis as the reference line, the Y axis as the axis perpendicular to the reference line, and the origin at the spatial index marker . In this hypothesis, the coordinates include the distance to the reference line measured perpendicular to the reference line and the distance to the spatial index marker measured parallel to the reference line.

A reproduction of the bitmap image can be provided on the display screen of the localization system to allow human intervention on the identified points of interest or regions of interest. Preferably, visual identification on the display screen of the interest point or line is provided.

Providing activation and/or deactivation of at least some of the points of interest or lines or regions of interest, or identification of the regions of interest as potential interventions or forbidden regions, following input on the human-machine input interface Is possible.

According to another aspect of the invention, the invention relates to a railway vehicle comprising a linear camera, at least one matrix camera and a localization system preferably comprising an odometer, the localization system comprising a method as described above can be performed with one or the other of the alternatives.

According to another aspect of the invention, the invention relates to a localization system with a linear camera and an odometer, and a rail vehicle with an orientation device, which may in particular be a touch probe or a matrix camera. This vehicle can be autonomous or hooked to an intervention vehicle that supports an intervention tool or track auscultation unit.

According to another aspect of the invention, the invention provides a localization system positioned in a first portion of a vehicle and a vehicle positioned behind the first portion in the direction of travel of the vehicle at a distance. For rail vehicles equipped with a transposition system positioned in the second part of the position determination system comprises a linear camera and an odometer, and potentially an orientation device, which may in particular comprise a touch probe or a matrix camera, The transpose system comprises at least one transposed matrix camera.

According to a preferred embodiment, the rail vehicle is a machine for building or maintaining railroad tracks, and at a distance, in the direction of travel of the vehicle, to a third portion of the vehicle located behind the second portion. Further comprising at least one tool disposed on the railroad track for intervention.

The invention also enables determining the positioning of one or more sets of tools carried by the rail intervention vehicle from previously acquired data, particularly regarding the positioning of points of interest or lines in the localization frame of reference. The purpose is to propose a means to

To this end, according to another aspect of the invention, a method is proposed for commanding a set of one or more tools attached to a rail intervention vehicle traveling on a railroad track in its direction of travel. and performed by a transposition system comprising a transposition chassis mounted on a rail intervention vehicle and one or more transposition matrix cameras fixed to the transposition chassis, the method comprising the following actions:
- Receipt of data characterizing the curvilinear abscissa of a spatial index marker of known signature, as well as the spatial index marker relative to a railroad track reference line and a point of interest or in a two-dimensional localization reference frame linked to the spatial index marker and the reference line positioning the coordinates of the line,
- Acquisition with the transposed matrix camera(s) of a set of one or more transposed matrix images in a spatial reference frame of the transposed system, fixed with respect to the transposed chassis,
- Acquisition of progress data of the transposition system relative to the railroad track using an odometer,
- identification of spatial index markers in a set of one or more transposed matrix images as a function of progression data and curvilinear abscissa data;
- determination of data characteristics of spatial index markers and reference lines in the spatial reference frame of the transposed chassis;
- the transposed coordinates of the point of interest or line in the spatial reference frame of the transposed system and the coordinates of the point of interest or line in the localization frame of reference as the coordinates of the spatial index marker and the data characteristic of the reference line in the spatial reference frame of the transposed system; calculation of, including

The transposing chassis is preferably fixed relative to the main chassis of the rail vehicle and may, where appropriate, be formed with this main chassis alone. A rail vehicle preferably comprises a number of undercarriages that support a main chassis and run on railroad tracks.

The coordinates of the point of interest may be Cartesian coordinates of a reference frame with the X axis as the reference line, the Y axis as the axis perpendicular to the reference line, and the origin at the spatial index marker . In this hypothesis, the coordinates include the distance to the reference line measured perpendicular to the reference line and the distance to the spatial index marker measured parallel to the reference line.

Preferably, the method includes positioning the set of one or more tools as a function of the coordinates of the point of interest or line in the transposed reference frame and positioning data of the set of one or more tools relative to the transposed chassis. Further includes procedures.

The set of one or more tools can be of any type, such as tamping tools, ramming tools, or bolting tools.

According to one embodiment, the set of one or more tools is movable with respect to the transposed chassis, and the intervention procedure comprises positioning data of the set of one or more tools with respect to the transposed chassis by the position measuring device. including the acquisition of

According to a particularly preferred embodiment for the interventional tool, the interventional procedure includes commands for displacing one or more sets of tools with respect to the displacement chassis.

According to one embodiment, the points of interest or lines constitute boundaries or properties of vertices or sides of a region of interest, preferably a quadrilateral forming the region of interest. Preferably, receiving coordinates of points of interest or lines in a two-dimensional localization frame of reference linked to spatial index markers and reference lines results in receiving data for identifying regions of interest as potential intervention or exclusion regions. Positioning of a set of one or more tools is uniquely performed, including where the region of interest is a potential intervention region.

According to one embodiment, the reference line is the neutral line of one of the rails of the railway line or a line constructed from the neutral line of the rails of the railway line.

According to one embodiment, determining the data characteristics of the spatial index markers and the reference line includes a transposition system orientation device that repeatedly determines angular orientation data of the railroad localization system with respect to the reference line.

In particular, the orientation device of the localization system comprises at least one first touch probe for detecting the orientation of the localization system with respect to the first rail of the railway track, constituting the first orientation rail, preferably , a second touch probe for detecting the orientation of the localization system with respect to a second rail of the railroad track, constituting the second orientation rail.

Also, the orientation device of the transposition system may comprise at least one first orientation matrix camera arranged facing a first rail of the railway track, constituting the first orientation rail. The first orientation matrix camera is preferably constituted by a first of the transposed camera(s). The orientation device takes a shot with the first orientation matrix camera and processes the shot to detect in the shot the orientation of the first orientation rail with respect to the target of the first orientation matrix camera.

The orientation device of the transposition system preferably comprises a second orientation matrix camera arranged facing a second rail of the railway track, constituting a second orientation rail. The second orientation matrix camera is preferably constituted by a second one of the transposed camera(s). The orientation device takes a shot with the second orientation matrix camera and processes the shot to detect in the shot the direction of the second orientation rail relative to the target of the second orientation matrix camera.

Where appropriate, it is possible to mix cameras and touch probes on the same rail or on two different rails. Therefore, it is possible to perform interpolation between the orientation data delivered by the two orientation matrix cameras to estimate the orientation of the transposed chassis. The presence of at least one sensor ( touch probe or matrix camera) above each rail of the railway track ensures that sections of the railway track are rail-free, especially when the transposition system proceeds along track devices, e.g. points. It is also possible to cancel out It is also possible to choose a particular rail orientation in the curve, for example a rail located inside the curve.

According to another aspect of the invention, the invention provides a set of one or more tools for intervention on a railroad track and a transposition chassis supported by a rail intervention vehicle and one or more tools fixed to the transposition chassis. It relates to a railway intervention vehicle with a transposition system with multiple transposition matrix cameras.

Various aspects of the invention can of course be combined as well as various embodiments.

Other features and advantages of the present invention will become apparent upon reading the following description with reference to the accompanying drawings in which the following description is made.
1 is a schematic side view of a rail intervention machine equipped with a localization system and a transposition system for implementing methods according to embodiments of the invention; FIG. 2 is a schematic top view of certain elements of the rail intervention machine of FIG. 1; FIG. 2 is a schematic diagram of the vehicle localization system of FIG. 1; FIG. 2 is a schematic diagram of a displacement system of the vehicle of FIG. 1; FIG. Figure 4 is a schematic diagram of an area of railroad track directly in line with the localization system of Figure 3;

For greater clarity, identical or similar elements are designated with the same reference symbols in all figures.

Figures 1 and 2 show a rail intervention machine 1 consisting of a first rail locating vehicle 2 and a second tow rail intervention vehicle 3, which again has a traction function. The railroad locating vehicle 2 comprises a main chassis 4 supported by several undercarriages 5 spaced from each other in the longitudinal direction of the railroad vehicle 2, while the railroad intervention vehicle 3 is a railroad locating vehicle with a main chassis 6 at one end. 2 fasteners 7 and supported at the other end by an undercarriage 8. Mounted on the main chassis 6 of the rail intervention vehicle 3 is a shuttle 9 which runs on the track via a chassis 10 and supports a set of one or more intervention tools 11 . A set of one or more actuators (not shown) allows displacement of the shuttle 9 with respect to the main chassis 6 parallel to the longitudinal direction of the main chassis 6 . The rail intervention machine 1 is positioned on the rail locating vehicle 2 and on the shuttle 9 of the rail intervention vehicle 3, at a distance, in the direction of travel 100 of the intervention machine 1, behind the locating system. and a set of one or more intervention tools 11 on the railroad track 22 are displaced at a distance, in the direction of travel 100, on the shuttle 9 of the railroad intervention vehicle 3 It is arranged behind the system 16 .

Location system 12 comprises a linear camera 26 and an odometer 28 connected to processing unit 30 . Linear camera 26 may, where appropriate, consist of several sensor units aligned along the same line of measurement. The linear camera 26 is mounted on a locating chassis 32 mounted under the main chassis 4 of the railroad locating vehicle 2 . An odometer 28 is mounted on one of the undercarriages 11.1.

As shown in FIG. 3 , the localization system 12 further comprises a human-machine interface 34 arranged in the control cabin 36 of the interventional machine 1 . This human-machine interface 34 comprises a screen 38 and a human-machine input interface 40, which can be integrated into the screen if it is a touch screen or if it is constituted by a keyboard or control levers, for example.

The transposition system 16, shown schematically in FIG. 4, comprises at least one transposition matrix camera 44A and a command unit 42 connected to a controller 46 for commanding a set of one or more tools 11. The transposed matrix camera 44A is now supported by the chassis 45 of the shuttle 9 of the rail intervention vehicle 3 . Where appropriate, command units 42, 30 of transposition system 16 and localization system 12 may be combined in the same unit. The set of one or more tools 11 can be of any type, in particular tamping tools, ramming tools or bolting tools.

By construction, the rails of the railroad track 22 locally define a horizontal or inclined reference plane according to the slope of the track. As long as the railroad locating vehicle 2 runs on the railroad track 22, the chassis 4 of the locating device is considered parallel to this reference plane, which is the railroad track 22 and the set of one or more tools 11. constitutes an acceptable approximation to the localization needs of commands in . The photosensitive cells of linear camera 26 are oriented along a direction perpendicular to the reference plane.

In the reference plane, as shown in FIG. 5, the orientation of the measurement line 50 defined by the linear camera 26 is measured in the track reference plane by an angle .delta. In the first embodiment, the neutral line of one of the rails 22A, 22B of the railroad track 22 is taken as the reference rail 22A. This orientation is considered unknown. This can vary as a function of the radius of curvature of the railroad track 22 , the positioning of the linear camera 26 with respect to the railroad locating vehicle 2 chassis, and as a function of the yaw of the railroad locating vehicle 2 with respect to the railroad track 22 .

At each pulse of odometer 28, the curvilinear abscissa of position location system 12 along datum line 122A is determined.

At a given instant, the linear camera 26 captures instantaneous linear optical data that constitutes a line of measurements covering the full width of the railroad track 22, including the width of the sleepers where appropriate. This input is repeated, with successive lines of measurement, making it possible to construct a two-dimensional bitmap image with steps that are a function of the distance traveled between the two measurements.

According to a first embodiment, the odometer 28 provides a pulse each time a known base distance moves in the direction of travel 100 of the railroad locating vehicle 2 on the railroad track 22, and these successive pulses are , are used to trigger the linear camera 26 . The spatial distance between two consecutive lines in the bitmap image is known and constant.

According to another embodiment, linear camera 26 is triggered at time intervals determined by dividing the observed time interval between two successive pulses of odometer 28 . Thus, the time T elapsed between odometer pulses I-1 and I is observed, and this time T is divided by a predetermined non-zero integer N to separate odometer pulses I and I+1. The period T/N is used as the time interval between two triggers of the linear camera. A posteriori, the trigger is observed between pulses I and I+1 of the odometer and the spatial step between the two triggers of the linear camera between pulses I and I+1 is deduced therefrom. If the travel speed is estimated to vary little on this scale, it can be hypothesized that the step is constant between two successive pulses. Other hypotheses consider, for example, that the velocity variation (acceleration or deceleration) is constant between the two pulses, resulting in a linearly variable spatial step between the two pulses I and I+1. can be assumed.

According to another embodiment, the linear camera 26 is triggered at regular intervals and the visual image data is time-stamped. The odometer 28 pulses are also time-stamped and the path traveled between two consecutive lines of the bitmap image, which may vary, can be determined by interpolation.

According to this latter embodiment alternative, the linear camera 26 is triggered at intervals that are not necessarily regular, thereby densifying the measurements of regions of interest where the increased accuracy may prove useful. , making them farther apart in the region of no interest, allowing optimization of the amount of data.

The bitmap image resulting from the continuous input of linear camera 26 is distorted due to the angle δ between linear camera 26 and the normal to reference line 122A. If a Cartesian reference frame is considered in which the X axis is parallel to the reference line 122A and the Y axis is perpendicular to the X axis in the reference plane, the railroad tracks 22 simultaneously measured by the linear cameras 26 at a distance from each other. appears to have the same abscissa in the distorted bitmap image, but in fact, in the previously defined Cartesian frame of reference , the distance and angle δ between the two points It is observed to have a difference proportional to the sine. Furthermore, two points P1, P3 of the railroad track 22 having the same ordinate in the previously defined Cartesian reference frame appear separated from each other in the distorted bitmap image, their apparent distance being the cosine of the angle δ is a function of

By estimating the angle δ, it is possible to "correct" the distorted image.

A first estimate of the angle δ is obtained by comparing the distance measured between two known points of interest at a given line of measurement of the linear camera with the known spacing between these points. obtain. Thus, on a given line of measurement, it is possible to evaluate the distance measured between two points, one centered on the reference rail 22A and the other centered on the other rail 22B of the railroad track 22. be. The ratio of the rail E spacing to the measured distance D is equal in absolute value to the cosine of the angle δ.

However, apart from the fact that this estimate of the angle δ cannot determine its sign, the accuracy of the estimate is low for small values of the angle δ and the derivative of the cosine function has a value close to zero. It is also assumed that the actual spacing between rails is constant and known to the desired accuracy.

By marking in a distorted bitmap image a given object that exists on the railroad track and whose contour or specific dimensions are known, and the apparent contour or dimensions measured on the distorted bitmap is referred to as the known contour or A second estimation method can be implemented from the matrix image by comparing with the image with the actual dimensions. However, in practice, there are no objects on the railroad track spaced close enough to serve as a comparison. In other words, using this method, the estimates of angle δ may be too far apart from each other.

To estimate the angle δ, we preferably rely on an orientation device 52 connected to the localization chassis 32 .

According to one embodiment, orienting device 52 comprises a touch probe fixed to the localization chassis.

According to another embodiment, the orienting device 52 is fixed to the locating chassis 32 and perpendicular to the reference plane of the railroad track 22 facing one of the two rails 22A, 22B, e.g. the reference rail 22A. at least one first orientation matrix camera 54A having a viewing direction of . The orientation bitmap image delivered by the orientation matrix camera allows, through image processing, the orientation of the reference line 122A to be identified and determined in the orientation bitmap image, thereby determining the orientation of the chassis 32. , thus providing direct access to the angle δ of the linear camera 26 .

In order to limit the computational power required to process the orientation images of the first orientation matrix camera 54A, when necessary, especially when it is estimated, or when it is determined that the angle δ is variable, It is possible to perform only the measurement of the angle δ.

In practice, it can be seen that the variation in the observed angle δ is small in the inter-sleeper space. It is therefore advantageously possible to use the odometer 28 to trigger the first orientation matrix camera 54A each time a predetermined distance is traveled on the reference rail 22A, this distance preferably being between sleepers. Equal to space. Furthermore, using the image processing operations performed on the distorted bitmap image delivered by the linear camera 26, each time a new inter-sleeper space is detected on the distorted bitmap image, the first orientation matrix camera It is also possible to trigger a shot by 54A.

It is also possible to choose to trigger the first orientation matrix camera 54A when it is determined that the angle δ has likely been corrected thanks to further measurement data. For this purpose, for example, an accelerometer positioned on the localization chassis 32 can be used. It is also possible to use the variations in the above-mentioned distance D measured between rails 22A, 22B by linear camera 26. FIG. It is of course possible to combine trigger modes, for example by combining systematic triggering by the odometer 28 and additional triggering as a function of measured or monitored data.

Since the amplitude and sine of the angle δ are known, it is possible to "fix" the distorted bitmap image estimated from successive shots of the linear camera 26 and distance traveled measurements delivered by the odometer 28. It is possible.

It is also possible to limit this correction to a few points of interest identified on the distorted image over a given inter-sleeper space. Furthermore, this limits the computational power required unless the distortion observed on the distorted image prevents locating the interest points by processing the image on the uncorrected distorted image. is the preferred solution for

In fact, assuming that the measurement of angle δ has been pre-triggered, on the distorted bitmap image there will be spatial index markers of a given signature, e.g. An overlying bolt center 56 or any other predefined track component, such as a fastening element or point core, is located. This localization is performed, in particular, by comparing the distorted bitmap image with various predetermined shapes of predefined railroad components distorted according to the angle δ, or by using an artificial neural network or, more generally, a received pre-training, in particular by artificial intelligence units with deep learning, for example by pixel-by-pixel pixel image segmentation methods (type SegNet), or object detection methods (type RFCN). Once this spatial index marker 56 is identified, its apparent distance to the reference line 122A is read on the distorted image, and the actual distance between the spatial index marker 56 and the reference line 122A as a function of the angle .delta. calculated and measured perpendicular to the reference line 122A. This then uses spatial index markers 56 and reference lines 122A (which are assumed to be straight on the scale of the inter-sleeper space) to establish a local two-dimensional localization reference frame of the considered inter-sleeper space. give. To get an idea, this frame of reference maps the projection O of the spatial index marker on the reference line 122A perpendicular to the reference line 122A for the origin and the direction of travel 100 of the vehicle 2 for the X axis. and for the Y-axis perpendicular to the X-axis, pass through the origin O (and pass through the spatial index marker 56).

A region of interest bounded by the lateral edges 58, 60 of two consecutive sleepers 62, 64 and the inner edges 66, 68 of the two rails 22A, 22B is then focused (its potential curvature is can be ignored). On the track, this area constitutes a quadrilateral that can be defined by points of intersection A, B, C, D between the edges of the sleepers 58, 60 and the inner edges 66, 68 of the two rails 22A, 22B. In a distorted bitmap image, the image of the quadrilateral itself is distorted, but is still locatable at its vertices. Then it suffices to identify the coordinates of the vertices in the distorted image and apply the necessary correction as a function of the angle δ to obtain the coordinates of the vertices A, B, C, D in the localization frame of reference. is.

In practice, the contrast of the image is not always sufficient to determine the edges of the sleepers and rails in the immediate vicinity of the points of intersection A, B, C, D on the distorted bitmap image. According to the alternative, therefore, the region of interest is defined by neutral lines 162, 164 constructed for each of the sleepers 62, 64 and neutral lines 122A, 122B constructed for each of the rails 22A, 22B. It is preferably defined as a quadrilateral bounded by the intersection points A', B', C', D' between them. A neutral line is constructed by processing the image over the inter-sleeper space.

The focus is also on identifying potential obstacles in a rectangular region of interest <A, B, C, D> or <A', B', C', D'>. The presence or absence of such obstacles can identify a region of interest as a potential intervention region (allowed region) or a potential exclusion region (forbidden region). Independent of obstacle detection, the dimensions of the region of interest are used to qualify the region of interest as a potential intervention region (if inter-sleeper space is sufficient to allow later intervention) or a forbidden region can.

The above operations (calculating angle δ, locating spatial index marker 56 and calculating its distance to reference line 122A, quadrilateral of interest point <A,B,C,D> or <A',B',C ', D'> and calculation of their coordinates in the localization frame of reference O, x, y defined by the spatial index marker 56 and the reference line 122A that locates the potential obstruction) to each sleeper. It is run periodically for the interspace, initiated by the odometer 28 or by recognition of the edge of the sleeper in the distorted bitmap image. In practice, an order number is assigned to each cycle and each inter-sleeper space.

These processing operations are performed by shape recognition algorithms. The human-machine interface 34 will disable them (if auto-recognition is considered enabled by default) or enable them (at least in learning mode, if the reliability of shape recognition is as long as the level is insufficient). For this purpose, the operator can see a distorted image displayed on the control screen 38 of the human-machine interface 34 . A quadrilateral <A, B, C, D> or <A', B', C', D'> is realized superimposed on the screen, for example by a rectangle, and the identified potential obstacles are: Where appropriate, they are marked according to predetermined visual conventions (arrows, outlines, etc.). When in verification mode, the operator clicks with the pointer in the region of interest to confirm its status. Conversely, when in service mode, the operator can also click with the pointer in the region of interest to indicate its status. Of course, many alternatives can be provided for the interaction between the operator and the location system 12 according to desired ergonomics and objectives.

At the end of the localization procedure described above, in the local reference frame <O, x, y>, for each localized inter-sleeper space, the coordinates of the point of interest A, B, C, D or A', B', C' , D′ are available and delimit regions of interest partitions [A,B,C,D] or [A′,B′,C′,D′] that are qualified as allowed or forbidden. It is also possible to make other data available, such as railroad ties.

These data are transmitted to the control unit 42 of the transposition system 16 so that, following successive advances of the rail intervention vehicle 3, the transposition system 16 is positioned in the predetermined space previously located by the localization system 12. The moment it is positioned at height, it can benefit from it. Due to the curvature of the railroad tracks 22, when the matrix camera 44A of the transposition system 16 is positioned above the inter-sleeper space, the positioning of the shuttle 9 relative to a given inter-sleeper space is similar to that of the linear camera 26. It differs from the positioning previously taken by the chassis 4 of the localization system 12 when positioned above the inter-sleeper space.

The transpose system 16 provides a translation between coordinates determined by the localization system 12 in a localization frame of reference <O, x, y> linked to a given inter-sleeper space, and in particular of a set of one or more tools 11. It is intended to allow the transposition, ie the modification of the Cartesian frame of reference, to coordinates that can be used specifically at the level of the transposition system 16 for commands.

To this end, matrix camera 44A of displacement system 16 is disposed facing reference rail 22A and has a field width sufficient to capture both reference rail 22A and spatial index marker 56; was selected near the reference rail 22A.

The transposition system 16 must first be able to determine what localization data is due to the inter-sleeper space seen by the matrix camera 44A of the transposition system 16 at any given moment.

However, knowledge of the geometry of the intervention machine 1 does not affect the odometer 28 of the localization system 12 and the transposition system 16 as long as the shuttle 9 is assumed to be movable with respect to the main chassis 6 of the rail intervention vehicle 3 . Even a rough estimate of the distance to the matrix camera 44A may prove insufficient. Additional measurements are therefore required which may be provided by the position sensor of the shuttle 9 relative to the main chassis 6 or by the optional odometer 70 integral with the undercarriage 10 of the shuttle 2 .

The combined measurements of the position sensor of the shuttle 9 and the odometer 28 of the localization system 12, or alternatively the odometer 70 measurements of the transposition system 16, have a sufficient level of confidence that any It is possible to determine whether the localization data is due to the inter-sleeper space seen by matrix camera 44A of transposition system 16 at a given moment.

The transposed matrix camera 44A is connected to a transposed frame of reference fixed relative to the shuttle 9 . By processing the bitmap image of the transposed matrix camera 44, the transpose unit 42 identifies the reference rail 22A, constructs the neutral line 122A that makes up the reference line, and determines its orientation in the bitmap image, thereby , the degree of orientation γ of the localized reference frame O, x, y with respect to the transposed reference frame . Transpose unit 42 also identifies spatial index marker 56 and constructs a projection of spatial index marker 56 onto reference rail 22A perpendicular to the latter, which corresponds to localization reference frame <O, x, y>. Define the coordinates of the origin O and its transposed reference frame . Transpose unit 42 may then transpose the coordinates of points of interest A, B, C, D that localization system 12 transmitted to it in localization frame of reference <O, x, y> to the transposed frame of reference.

Based on this, it is possible to transmit the transposed coordinates of the region of interest and its identification (as a potential intervention region) to the controller 46 for commanding the set of one or more tools 11 . The controller 46 then generates commands that enable the set of one or more tools 11 to intervene or not intervene in the regions of interest A, B, C, D thus delimited. Where appropriate, there may be one or more degrees of freedom of movement of the set of one or more tools 20 relative to the transposition chassis 45 that supports the transposition camera(s) 44A, 44B. The command is to rotate the set of one or more tools 20 about an axis perpendicular to the reference plane, and one or lateral translation of a set of multiple tools 20 . Commands may also include raising or lowering a set of one or more tools 20 by identifying the area of interest as a potential intervention area or prohibited area.

Of course, the examples shown in the figures and discussed above are given for illustrative and non-limiting purposes only. It is expressly provided that the different embodiments shown can be combined between them to propose other embodiments.

The reference line chosen in the localization step is not necessarily the line along which the estimation of the angle δ is performed. If appropriate, it is possible to select the neutral line 122B of rail 22B as a reference line and measure the angle δ with respect to rail 22A.

According to the alternative, the reference lines constructed by the localization system 12 are virtual in the sense that they are not linked to specific rails 122A, 122B. It can be, for example, a railroad track centerline 222 constructed from the neutral lines 122A and 122B of the rails 22A, 22B of the railroad track 22 . Thus singularities such as interruptions of one of the rails at the level of several track gears are eliminated.

By processing the distorted bitmap image, by searching point by point for each measurement line of the linear camera 26 at the middle of the segment delimited by the centers of the two rails 22A, 22B, or preferably first , the neutral line 122A, 122B of each rail 22A, 22B, and then a line located at an intermediate distance from the neutral line 122A, 122B.

With this hypothesis of the virtual reference line constructed and used by the localization system 12, the transpose system 16 must also be able to reconstruct the virtual reference line. To this end, the displacement system preferably comprises a second displacement matrix camera 44B disposed above and facing the second rail 22B.

A second transposed matrix camera 44B is fixed to the transposed chassis 45 of shuttle 9 so that the relative position of the two transposed matrix cameras 44A, 44B is known and calibrated. Each of the transposed matrix cameras 44A, 44B can only display the rails 22A, 22B positioned above it due to its narrow field width, but only the rails 22A, 22B positioned above it can be displayed. By determining the positioning of the neutral line 122A of one rail 22A, by determining the positioning of the neutral line 122B of the second rail 22B in the bitmap image of the second displaced camera 44B, and the two neutral lines The center line 222 between the neutral lines 122A, 122B is positioned by calculating the center of the straight line segment between 122A, 122B from these measurements of the distance between the two displaced cameras 44A, 44B and the calibration data. It is possible to decide

According to an alternative of the locating system 12, the latter is fixed to the chassis 4 of the locating vehicle 4 and faces another rail 22B and has a viewing direction perpendicular to the reference plane of the railroad track 22. An orientation matrix camera 54B is provided. A second orientation matrix camera 54B may be used to provide a second value of the angle δ measured with respect to the second rail 22B. Various algorithms may be implemented to take advantage of these data. For example, one or It is possible to assign a reliable index to each bitmap image delivered by another orientation matrix camera. It is also possible to combine the measurements performed to calculate the "average" angle δ. As long as the distance between the two orientation matrix cameras 54A, 54B is known and calibrated, they can also be used to determine the virtual reference line 222 described above. Finally, two independent measurement and calculation lines can be envisaged. One uses a first orientation matrix camera 54A and a first transposed matrix camera 44A for a first reference line 122A, and the other uses a second orientation matrix camera 54B and a second transposed matrix for a first reference line 122B. Use camera 44B. If appropriate, it is also possible to provide two odometers at the level of the monitoring system. One for each rail 22A, 22B, each dedicated to one of the two measurement and calculation lines.

In the localization phase, the identification of points of interest is not limited to identification of the vertices A, B, C, D or A', B', C', D' of a potential intervening quadrilateral. Other types of points of interest can be identified, such as the coordinates of the center of the screw head for screw tightening or modification. The intervening region does not necessarily have to be quadrilateral, but can be any polygon. Instead of or in addition to the points of interest A, B, C, D, it is also possible to seek to identify lines of interest, eg, straight lines that constitute the neutral lines 162, 164 of the sleepers 62, 64. Each is located at the corresponding center of the sleeper 62, 64 or of the straight line 163 forming the line between the sleepers 62, 64, which is the axis of symmetry of the preceding straight line 162, 164 located mid-distance between the two sleepers 62, 64. be. In the local frame of reference <O, x, y>, the coordinates of such a line can be, for example, the coordinates of two points belonging to it, or the coordinates of a point and an angle of a line.

Although the above description focused on inter-sleeper spaces, it is also possible to analyze sections of railroad tracks with sleepers in a localization system to detect points of interest therein that specifically require intervention. .

The processing of the odometer 28, linear camera 26 and orienting device 52 data by the localization system 12 is done in real time or with a very slight delay and belongs to the same continuously advancing intervention machine 10 transposition system 16. may use it. The distance between the locating system 12 located in the front region 14 of the locating vehicle 2 on the one hand in the direction of travel 100 and the transposition system 16 located at a distance behind the vehicle on the other hand is, in particular, for the operator allows enabling or disabling points of interest A, B, C, D or certifying them.

Alternatively, the interventional machine 10 can be advanced in a direction for localization and then retracted during the second pass to command the set of transposition and one or more tools. It is possible. During displacement, the vehicle may travel in the opposite direction or in the same direction as the direction of travel during localization.

The railway intervention machine 10 in question may consist of one or more vehicles that are indirectly joined between them. Thus, where appropriate, the localization system 12, the displacement system 16 and the set of one or more tools 11 may be on a single, same vehicle. According to another embodiment, the localization system 12 may be mounted on a tank traveling on tracks and articulated at the front by a rolling unit carrying a displacement system 16 and a set of one or more tools 11. .

It is also possible to envisage that the localization system 12 is mounted on an autonomous localization vehicle independently of the intervention machine 10 carrying the set of one or more tools 11 . In this latter hypothesis, it is advantageous to provide the locating vehicle with an absolute positioning unit, e.g. a GPS unit, to allow each sleeper interval to be assigned a coordinate in the absolute positioning reference frame , the latter being the two consecutive Separate the inter-sleeper spaces.

Where appropriate, one or more sets of tools 20 may be fixed relative to the main chassis 6 of the intervention vehicle 3 . Next, approximate the positioning of the set of one or more tools 20 from the unique data of the odometer 28 of the localization system 12 to determine which, at any given moment, is the localization data associated with the displacement. It is possible to determine whether

Although the description focuses specifically on the use of interventional tool command transposition procedures, transposition also benefits from manipulating accurate auscultation of the track by an auscultation instrument carried by a chassis remote from the localization chassis. can receive

The human-machine interface 34 may be located remotely with respect to the vehicle 10, for example at a remote control post.

Claims (12)

A method for commanding a set of one or more tools mounted on a rail intervention vehicle traveling in a direction of travel on a railroad track , performed by a transposition system , said transposition system comprising: a chassis attached to a rail intervention vehicle ; and one or more transposition cameras fixed to the chassis of the transposition system , the one or more transposition cameras comprising: one or more matrix cameras, the method comprising the following actions:
- Receipt of data characterizing the curvilinear abscissa of a spatial index marker of a predetermined signature and the positioning of said spatial index marker relative to said railroad track reference line, and 2 linked to said spatial index marker and said reference line; receiving coordinates of a point of interest or line in a dimensional localization frame of reference;
- acquisition with said one or more transposed cameras of a set of one or more bitmap images in a spatial reference frame of said transposed system, fixed with respect to said chassis ;
- acquisition of progress data of said transposition system with respect to said railroad track using an odometer ;
- identifying said spatial index markers in said set of one or more bitmap images as a function of said progression data and curvilinear abscissa data;
- determining data characteristics of the spatial index marker and the reference line in the spatial frame of reference of the chassis ;
- calculating the transposed coordinates of the point of interest or line in the spatial reference frame of the transposed system as a function of the spatial index marker and the data properties of the reference line in the spatial reference frame; calculating the coordinates of the point of interest or line in a dimensional localization frame of reference;
- an interventional procedure comprising positioning the set of one or more tools as a function of the transposed coordinates of the point of interest or line in the spatial frame of reference of the transposed system;
method including. wherein the set of one or more tools is movable with respect to the chassis of the displacement system , and wherein the intervention procedure is performed by positioning and measuring the one or more tools relative to the chassis of the displacement system; 2. A method according to claim 1 , comprising acquiring positioning data for a set of tools of . 3. The method of claim 2 , wherein the intervention procedure includes commands to displace the set of one or more tools with respect to the chassis of the displacement system . 4. A method according to any one of claims 1 to 3 , characterized in that said points of interest or lines constitute boundaries or features of vertices or sides of a region of interest, preferably a quadrilateral forming said region of interest. How . Receipt of coordinates of points of interest or lines in a two-dimensional localization frame of reference linked to the spatial index marker and the reference line receives identification data of the region of interest as a potential intervention or forbidden region. 5. The method of claim 4 , wherein said positioning of said set of one or more tools is uniquely performed when said region of interest is a potential intervention region. law. The method according to any one of claims 1 to 5 , characterized in that said reference line is a neutral line of one rail of said railway line or a line constructed from neutral lines of two rails of said railway line. A method according to any one of paragraphs. further comprising positioning performed by a railroad locating system comprising at least one linear camera facing said railroad track and one or more odometers, said railroad locating system traveling on said railroad track; Direction and Locate following actions:
- using the one or more odometers to repeatedly acquire progress data of the railroad localization system on the railroad track in the direction of travel;
- repeatedly acquiring instantaneous linear optical data along an instantaneous measurement line with said linear camera facing said railroad track;
- constructing a bitmap image of a region of the surface of said railroad track by processing at least said instantaneous linear optical data;
- by processing the bitmap image of the area of the surface of the railroad track, identifying in the bitmap image the spatial index marker of a predetermined signature, and processing at least the progress data; determining the positioning of the spatial index marker relative to the curvilinear abscissa of the spatial index marker and the reference line of the railroad track;
- identifying the points or lines of interest in the bitmap image of the area of the surface of the railroad track; and determining coordinates of points or lines. Determining data characteristics of the spatial index marker and the reference line is repeated angular orientation data of the railroad localization system with respect to the reference line using an orientation device of the railroad localization system. 8. A method according to claim 7 , characterized in that it comprises determining 4. The railroad localization system's orientation device comprises at least one first touch probe for detecting an orientation of the railroad localization system with respect to a first rail of the railroad track. 8. The method according to 8. 10. The method of claim 9, wherein the orientation device of the railroad localization system comprises a second touch probe for detecting an orientation of the railroad localization system with respect to a second rail of the railroad track. The orientation device of the transposition system comprises at least one first orientation camera disposed facing a first rail of the railroad track, the first orientation camera being a matrix camera. and the orienting device takes a shot with the first orienting camera , processes the shot, and directs the first laser to a target of the first orienting camera . 9. The method of claim 8, wherein the orientation of the le is detected in the shot. The orientation device of the transposition system comprises a second orientation camera disposed facing a second rail of the railroad track, the second orientation camera being a matrix camera, the orientation device comprising: taking a shot with the second oriented camera and processing the shot to detect in the shot the orientation of the second rail with respect to the target of the second oriented camera; 12. The method of claim 11.