• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The invention relates to a method for determining the position of a vehicle (1), in particular a rail vehicle or a line-bound vehicle, connected to a network, - wherein a database containing a number of network paths (12a, 12b, ...) is provided in advance representing the network within which the vehicle (1) moves, - driving is performed with this networked vehicle (1) and the movement path (2) of the vehicle (1) is determined, - wherein the movement path (2) is compared with the network paths (12a, 12b,...) stored in the database and checked for matches, - in the case where the movement path (2) with a network path (12a, 12b, at least partially coincident, this network path (12a, 12b, ...) is selected, and - on the basis of the determined match, the point assigned to the respectively last recorded point on the movement path (2) on the exten is determined Netzpfad (12a, 12b, ...) is determined and used as a position of the vehicle (1).
    公开号:AT516553A1
    申请号:T50886/2014
    申请日:2014-12-05
    公开日:2016-06-15
    发明作者:
    申请人:Ait Austrian Inst Technology；
    IPC主号:
    专利说明:

    The invention relates to a method for determining the position of a vehicle bound to a network. Such a vehicle is, for example, a rail vehicle such as a passenger or freight train or a trolley-bound vehicle such as a tram or an O-bus. Background of the invention is the location of rail vehicles or trolley-bound vehicles on a known network of possible paths. Another background of the invention is the detection of the future travel path of a network-bound vehicle and the determination of objects that are in the future travel path of such a vehicle.
    From the prior art, various methods are known with which the position of a vehicle can be generally determined. Furthermore, methods are known from the prior art, with which objects can be detected, which are located within a certain range with respect to the vehicle. All of these methods are particularly in curved Fahrweg courses of
    Rail vehicles with certain uncertainties afflict that complicate a targeted detection of objects in the travel path of a vehicle.
    The first object of the invention is therefore to determine the position of a rail-bound vehicle on a route network as safe as possible and error-free. Another object of the invention is to predict the future travel path of a vehicle located on a given network safely. A further object to be achieved by preferred aspects of the invention is to reliably detect objects which are located in the future travel path of the rail-bound vehicle.
    The invention solves the problem of accurately and safely determining the position of a vehicle by providing in advance a database containing a number of network paths representing the network within which the vehicle is moving, - a ride with this networked vehicle is made and the movement path of the vehicle is determined, - wherein the movement path is compared with the network paths stored in the database and examined for matches, - in the event that the movement path with a network path from the database at least partially matches, selected this network path is determined, and - based on the determined match of the last recorded point on the movement path associated point on the selected network path and is used as a position of the vehicle.
    Such a method can be carried out with little effort and also in positions, for example in tunnels or the like, where positioning by means of GPS or other satellite-based positioning systems is not possible. In order to determine the future travel path of a network-bound vehicle, it can be provided that, for determining the future travel path of a network-bound vehicle, a passage direction with respect to the network path is determined, which corresponds to the passage direction of the movement path on the basis of the determined match, and a partial path as the future travel path of the vehicle selected the selected network path, which follows in the direction of passage of the position of the vehicle.
    A particularly simple determination of network paths and to build a database for positioning a vehicle provides that the database is made available by one or more calibration trips are made with the vehicle along the predetermined routes within the network, and that while driving the Vehicle respectively recorded movement paths of the vehicle and added as network paths of the database.
    An advantageous development of the invention, which allows a particularly simple and effective detection of movement paths and matches of paths, provides that each point on the path in each case an orientation of the network-bound vehicle, in particular comprising three the orientation in the determination of the movement path or network path the vehicle relating linearly independent direction vectors, is associated with respect to the path and stored in the database, and in particular that the future travel path associated with the network path orientation information is assigned.
    A simple possibility for accelerating the method of determining the position and for reducing the calculation effort expended for determining the position provides that a route identifier is predefined both during the calibration runs and during the journeys, the route identifier being assigned to the network paths of the database as an additional attribute and at the
    Determination or search of the network path from the database or for the comparison of the network path with the movement path only located in the database network paths are used, which has been assigned the same route identifier.
    In order to enable a georeferencing of the determined position, it can be provided that during the calibration drive each geographic location, in particular in the form of coordinates, is assigned to the individual points on the network paths and that during the journey at predetermined times, in particular when taking a At each point on the network path, a geographic location is made and assigned to the particular point on the network path. For an additional acceleration of the comparison between the movement path and the network path it can be provided that the geographical position of the vehicle is determined during the journey and only network paths from the database are used for the comparison, whose associated geographical localization information is within a predetermined environment around the vehicle located during the trip recorded geographical position.
    An additional acceleration of the finding of matching paths provides that, for comparison of a movement path with a network path, the partial path of the movement path representing the last traveled distance of a predetermined length is used and only this partial path is used as the movement path for the comparison with the network paths.
    An adaptive adaptation, which makes it possible to uniquely locate the rail-bound vehicle while at the same time requiring little computation, provides that if there is no match of a movement path with the network paths stored in the database or if there is a multiple match of a determined movement path with the network paths stored in the database the path used for comparing the paths is increased or decreased until there is a clear match of the path with a single network path stored in the database.
    The recording of movement paths and network paths can be carried out particularly simply by determining the position at discrete points in time when recording movement paths or network paths, which preferably have the same time interval at the respectively following point in time, wherein for each
    Waypoint respectively the relative position of the next following waypoint is specified.
    Particularly good matches can be achieved if for each waypoint, the relative rotation of the vehicle or change the orientation of the vehicle relative to the orientation of the vehicle is given in the respective preceding path.
    An advantageous development of the invention, which allows a particularly robust type of comparison based on a projected curvature vector function, provides that the movement paths and the network paths in the form of a curve function p (s) are given relative to a starting point, - that the path described so is projected onto a predetermined, in particular horizontally extending, plane - that the second derivative of the function pxy (s) representing the projection is then determined according to the distance s as the curvature vector function K (s), - that the curvature vector function K (s ) If necessary, a scalar, signed curvature function K (s) is determined, and - that compares a network path with a movement path of the curvature vector function or curvature function of the network path with the course of the curvature vector function or curvature function of the path is compared, in particular correlated, u nd - that the position of the network path and is determined that matches best with the movement path of the vehicle, - that this determined position of the current position of the networked vehicle is kept the same and - optionally that the sub-path of the found network path following this position as the future path of networked vehicle is viewed and kept available.
    A further simplification of the examination of movement and network paths for agreement provides that, at a point in time during the journey, each individual network path stored in the database is compared with the path determined by the networked vehicle, the network path in which the path traveled by the vehicle was found was used for the determination of the position s of the vehicle as well as for the determination of the future track course of the vehicle, and - that in further subsequent provisions of the position, and possibly the future route, the networked vehicle exclusively in this selected network path for matches of the Vehicle traveled path is searched with this network path, and - that in particular the remaining network paths are no longer used in the determination of the position and / or the future travel path, especially as long as the movement path in the selected th network path can be found.
    A simple determination of the local parameters of the path provides that the determination of the network paths and the determination of the movement path by means of a camera, wherein the camera image is supplied to a visual odometry method, with which the respective traversed by the networked vehicle travel path or network path is determined ,
    An easy way to determine the network paths provides that the determination of the network paths by a one-time extraction to be performed and suitable further processing of data from an existing map works done.
    For effective and simple detection of objects in the future travel path of the vehicle, it can be provided that a search for objects in the surroundings of the vehicle is carried out starting from the networked vehicle, this search being carried out in particular by means of a stereoscopic multi-camera arrangement, wherein a number of points or Objects are created with respect to the cameras, - wherein the determined in relation to the vehicle, determined future path with respect to the two cameras is determined and searched for detected objects in the area of the future path, and finding such objects in the area of A warning is issued in the future path that obstacles are in the future travel path of the vehicle.
    Advantageously, a program for carrying out a method according to the invention can be stored on a data carrier.
    The invention will be described in more detail below with reference to a non-limiting embodiment.
    Fig. 1a shows schematically the recording of a movement path by means of a vehicle. FIG. 1b shows the curvature function determined for the movement path. Fig. 2 shows a network comprising a number of network paths, each representing rails for a vehicle. Figures 3a and 3b show schematically the check for correspondence between a motion path and a network path. Fig. 3c shows the position of the vehicle in the network. 4 shows the determination of the future travel path. 5 shows a possible object recognition based on a future travel path determined according to the invention.
    Position determination in general
    In order to find the position of the vehicle 1, the steps shown below are performed in succession.
    In this context, the term path means a data representation of an actual distance existing in the real world. A path can be specified in different ways, for example by a number of coordinates of points lying on the respectively represented route. A network is understood to mean a number of paths that can be interconnected.
    This exemplary embodiment comprises the determination of the path traveled by the vehicle 1, hereinafter referred to as movement path 2, which provides a database of existing paths, hereinafter referred to as network paths 12a, 12b,... (FIG. 2), and a comparison of a movement path 2 with a number of network paths 12a, 12b, ... from the database. Due to the comparison, agreement can be determined and the position of the vehicle 1 in the network 10 can be found.
    Determination and processing of motion paths
    In Fig. 1a, a movement path 2 is shown, which is traveled by the vehicle 1. The movement path 2 of the vehicle 1 can be determined concretely in different ways, wherein the determination of the movement path 2 by means of visual odometry is particularly advantageous.
    In this case, a visual odometry method based on a stereo camera system is used while driving along a rail track within a known route network. A possible implementation of the method is explained in more detail in (Nister et al, Visual Odometry, CVPR, 2004). In this method, based on stereo camera images taken by cameras located on the vehicle 1, disparities are determined based on corresponding image features. Due to these disparities, the three-dimensional position of individual object points is determined in relation to the camera system and thus to the vehicle. By means of correspondence formation within consecutive image pairs it is determined how the position of recognized object points has changed. From this, the travel path of the vehicle 1 can be determined in reverse and represented as a movement path 2.
    In the example shown in FIG. 1a, the vehicle 1 moves along a rail track and is at the waypoint s0 at time t0, at time U after traveling a distance Si, etc. The movement path 2 of the vehicle 1 can take the form of a curve function p (see FIG ), which indicates the respective position of the vehicle 1 in each case as a function of the traveled distance s. Even if the curve function p (s) could in principle also be given as a function of the time t, the representation as a function of the traveled distance s has the advantage that the comparison of movement paths 2 and network paths 12a, 12b is made without specifying the respective speed can be.
    The curve function p (s) comprising the identified motion data of the vehicle 1 are fundamentally in the form of three-dimensional data. In this embodiment of the invention, these data are projected onto a local, in particular horizontal, reference plane to which the last known position and orientation of the vehicle relates. By means of this step, drift effects in the vertical direction, which can result from unavoidable inaccuracies in the determination of the individual motion vectors, are eliminated. In this way, due to the resulting and juxtaposed motion increments, a corrected track layout is determined. As the vehicle 1 progresses, a metric is continuously calculated to serve as the basis for later localization. In the present exemplary embodiment, this is the local curvature K of the travel path in relation to the path s. Concretely, a short section of a few meters is transformed into the reference coordinate system of the last known vehicle position s. The resulting deviations y, (s) in the direction of the local lateral
    Vehicle axis are approximated by means of a polynomial function qi (s), whose curvature factor is then used as a metric for the corresponding section of the route. This method is repeated as soon as the vehicle moves on by a certain distance increment. Due to the fact that a local alignment of the route section takes place orientation-related dependencies of the method can be avoided. From the individual local curvatures K determined, a curvature course K (s) can be created via the movement path 2. By way of example, the curvature curve K (s,) from the polynomial function q can be determined by the following rule: K (Si) = -d2q, / ds2 (¾)
    In Fig. 1b, the curvature K (s) is given, which indicates for each waypoint on the movement path 2 of the vehicle 1 and the relative rotation of the vehicle 1 or change the orientation of the vehicle 1 relative to the orientation of the vehicle 1 in the respective preceding movement path 12a , For example, it can be seen that the vehicle 1 makes a stronger clockwise turn to the waypoint s3, the curve K (s3) accordingly has a positive rash. Accordingly, the vehicle 1 turns to the waypoint s4 to the left. From Fig. 1b can be a negative rash in the curvature K (s) recognize. In this way, it is possible to determine a characteristic profile in the form of the curvature course K (s) for each movement path 2 traversed by the vehicle 1, which profile can be distinguished from other movement paths 2 or network paths 12a, 12b likewise traversed by the vehicle 1. Due to this characteristic profile, represented by the curvature course K (s), it is possible to compare a movement path 2, which has been traversed during the movement of a vehicle 1, with a multiplicity of previously determined or traversed network paths 12a, 12b,.
    Database of network paths
    Fig. 2 shows schematically a network 10 comprising a number of network paths 12a, 12b, ..., on which vehicles 1 can run. The individual network paths 12a, 12b,... Of the network 10 are stored in a database. This database can be created in different ways.
    One possibility is to determine the network paths 12a, 12b,... By means of the method described above analogously to the movement path 2, wherein for each
    Net path 12a, 12b, ... each have a curvature function or curvature vector function is available. The database is made available by making one or more calibration trips with the vehicle 1 along the predetermined routes within the route network. During the journey of the vehicle 1, the paths of the vehicle are respectively recorded. These are available as curve function p (s).
    Another possibility is to feed existing surveying data of existing routes of the route network as network paths 12a, 12b,... Into the database. Such survey data is regularly available as a series of coordinate values corresponding to the curve function p (s).
    In accordance with the procedure described above, the curvature function or the curvature vector function are determined in both procedures from the curve function p (s). The curvature function or curvature vector function is assigned to the respective network path 12a, 12b and added to the database. For the method of determining the position of a vehicle 1 shown below, it is only necessary that such a database with network paths is available, regardless of how the individual network paths 12a, 12b were determined. It is not necessary that the specific nature of the relationship of the individual network paths 12a, 12b is not necessarily stored in the database.
    Comparison of paths and position determination
    An advantageous possibility for locating ascertained movement paths 2 in network paths 12a, 12b,... Is described in more detail in connection with FIGS. 3a and 3b. For the comparison of movement paths 2 and network paths 12a, 12b,... Shown below, the curve of the curvature vector function K (s) or curvature function K (s) of the respective network path 12a, 12b is compared with the profile of the curvature vector function K (s) or curvature function K (s) of the movement path 2 compared, in particular correlated.
    With the aid of a pattern recognition method from the signal processing, it is determined whether and at which position within the network paths 12a, 12b,... Of the network there is a match with the movement path 2. This method, which is based on the mathematical method of normalized cross-correlation, is for example implemented in the free program library OpenCV and is documented at: http://docs.opencv.org/modules/imgproc/doc/object_detection.html highlight=matchtempla te #matchtemplate (OpenCV - Object detection). For the movement path 2 (FIG. 3 a) as well as for each of the network paths 12 a, 12 b,... (FIG. 3 b) there is in each case a curvature vector function K (s) or curvature function K (s). In the event that the movement path 2 with a network path 12a, 12b, ... from the database at least partially matches, this network path 12a is selected for further localization of the position of the vehicle 1.
    Subsequently, with the aid of the pattern recognition method it is determined at which position the mutually coinciding regions of the movement path 2 and of the selected network path 12a coincide with one another. The position for which the respective match is maximum is regarded as the instantaneous position of the vehicle 1. In the present case, the recorded movement path 2 of the vehicle 1 best matches the path area between sx and sy of the selected network path 12a. The positions sx and sy of the beginning and the end of the movement path are shown in Fig. 3c with respect to the selected network path 12a. On the basis of the determined match, the point assigned to the respectively last recorded point on the movement path 2 on the selected network path 12a is determined and used as the position sy of the vehicle 1.
    When localizing the current vehicle position within a route network or even a single route, ambiguities in the determination of the optimal match position can occur. For this reason, the coincidence range of the last traveled route section can be successively increased to thereby achieve an unambiguous position determination. If, for some reason, this is not sufficient to avoid ambiguity, filtering methods may be used that relate to the consistency of successive localization steps. Should the vehicle perform movements, such as jumps, that do not appear plausible, these localization hypotheses will be discarded. The remaining optimal solution for the current position of the vehicle 1 is subsequently used to determine the route ahead of the vehicle 1. For the comparison, the entire movement path of the vehicle 1 need not be used. Rather, it is sufficient that for comparison of a movement path 2 with a network path 12a, 12b of the movement path 2 in each case the last traveled distance of a predetermined length is used and this is used for the comparison with the network paths 12a, 12b.
    In the present embodiment of the invention, in order to save computing time, it is provided that at a time during the journey each individual network path stored in the database is compared with the movement path 2 determined by the vehicle 1, wherein the network path 12a in which the network path traveled by the vehicle 1 Movement path 2 was found was used for the determination of the position s of the vehicle.
    If a network path 12a has already been successfully identified as coinciding with the movement path 2, then in the following determinations the position of the networked vehicle 1 only needs to be searched in this selected network path 12a for matches of the movement path 2 traveled by the vehicle 1 with this network path 12a. The remaining network paths 12b, ... need not be used in the determination of the position of the vehicle until the movement path with the selected network path no longer matches. Future trip
    In order to be able to effectively detect possible obstacles for a vehicle 1, it is advantageous to know the future travel path 3 of the vehicle 1. While prior art methods extract the future route 3 directly from the images obtained on the vehicle 1, the future route 3 of the vehicle 1 can be easily read on network paths 12a, 12b,..., Where no intersections exist Location of the available network paths 12a, 12b, ... are determined. If several options are available for a crossover or switch, a number of future network paths 12a, 12b,... Can be examined for the presence of obstacles as a precaution. Alternatively, even with knowledge of the actually selected future network path 12a, 12b,..., Only this can be examined for objects.
    In order to determine the future travel path 3 of a vehicle 1, a passage direction D, shown as arrow (FIG. 4 a), which corresponds to the passage direction D of the movement path 2 on the basis of the determined network path 12 a, is defined. As a future travel path 3 of the vehicle 1, a partial path 3 of the selected network path 12 a is selected, which follows in the direction of passage D of the position sy of the vehicle 1. This partial path 3 of the selected network path 12a following the position sy of the vehicle 1 can be regarded as a future path 3 of the vehicle 1 and kept available.
    Object recognition in the future travel route
    Fig. 5 shows the area around the future travel path of the rail vehicle in detail. The identification of the future route is based on the fact that both the position and the orientation, including the direction of travel and possibly inclinations, of the vehicle 1 are known within a certain route section. After the position sy has been determined unambiguously by means of the described correlation method and with the aid of various filter methods, the orientation of the vehicle in relation to the route is now determined.
    This is possible by means of a two-dimensional matching method, the alignment of a short section at the end of the already traveled travel path to the corresponding section of the stored network path 12a, in which it was located, is aligned. The least squares method is used to determine the spatial transformation x necessary to transform the following known waypoints from the network path into the vehicle's local coordinate system. By not only aligning the orientations of the identified locating points, but applying a best-fit matching over a short span, slight deviations of the current orientation of the vehicle 1 compared to the orientation of the network path can be identified. In particular for an object and obstacle detection, the knowledge of the specific orientation of the future travel path 3 in relation to the camera system mounted on the vehicle is relevant.
    Starting from the vehicle 1, a search for objects 21, 22, 23, 24 in the vicinity of the vehicle 1 is made. This search can be carried out in particular by means of a stereoscopic multi-camera arrangement. A
    Number of objects 21, 22, 23, 24 is detected in this area and related to the future route 3 in relation.
    As shown in FIG. 5, some objects 21, 22, 23, 24 are located in the region of the vehicle. Due to the knowledge of the position of the travel path 1 (sy,..., Sz) relative to the vehicle 1 or relative to the vehicle The object detection means arranged in the vehicle 1 can determine which of the objects 21, 22, 23, 24 is located in the future travel path 3 of the vehicle 1. In this way it can be stated that the object 22 constitutes an obstacle for the vehicle 1, while the remaining objects 21, 23, 24 are located outside the future travel path 3. If necessary, it can also be ascertained on the basis of a movement analysis of the objects 21, 22, 23, 24 which of the objects 21, 22, 23, 24 are to move on the future travel path 3 of the vehicle 1 and whether collisions are imminent.
    Upon finding such objects in the area of the future travel path 3, a warning is issued that obstacles 22 are located in the future travel path 3 of the vehicle 1.
    Further improvements through geolocation and attribution
    The database of the present embodiment of the invention can be improved by assigning geo-localization data and / or network path attributes to the individual network paths 12a, 12b, .... With geo-localization data, individual points of the network paths 12a, 12b, ... can be uniquely determined with respect to the earth's surface. The assignment of network path attributes allows a unique identification of network paths 12a, 12b, ....
    The network paths 12a, 12b,... Of the database are each assigned a route identifier as an additional attribute. In the determination or search of the network path 12a, 12b,... In the database or for the comparison of the network path 12a, 12b,... With a movement path 2, only the network path 12a, 12b in the database is used the respective route identifier has been assigned. In this way, the search of the position of the vehicle 1 or of the future travel path 3 can be drastically restricted.
    Furthermore, additionally or alternatively, a geolocalization of the network paths 12a, 12b,... Can also be undertaken, which is used as an additional clue in the detection of the position of the vehicle 1. In the case of the calibration runs carried out to create the network paths 12a, 12b,..., The individual points on the network paths 12a, 12b,... Are each assigned a geographical localization. Alternatively, it is also possible to extract the geographic location information from other sources and insert it into the database. During the travel of the vehicle 1, a geographical localization is performed at predetermined times, in particular when each point is recorded on the movement path 2. For the comparison, only network paths 12a, 12b,... From the database are used whose associated geographical localization information is located within a predetermined environment around the point recorded during the journey. By this measure, the number of possible comparisons of movement and network paths 2, 12a, 12b, ... can be drastically reduced.
    权利要求:
    Claims (16)
    [1]
    claims:
    A method for determining the position of a network-bound vehicle (1), in particular rail vehicle or trolley-bound vehicle, - wherein a database is provided in advance, which contains a number of network paths (12a, 12b, ...), the represent the network within which the vehicle (1) moves, - with this network-bound vehicle (1) a journey is made and the movement path (2) of the vehicle (1) is determined, - wherein the movement path (2) with the in the database stored network paths (12a, 12b, ...) is compared and examined for matches, - wherein in the event that the movement path (2) with a network path (12a, 12b, ...) from the database at least partially is selected, this network path (12a, 12b, ...) is selected, and - on the basis of the determined match, the point on the selected path (12a) assigned to the respectively last recorded point on the movement path (2) 12b, ...) is determined and used as the position of the vehicle (1).
    [2]
    2. The method of claim 1, wherein for determining the future travel path (3) of a network-bound vehicle (1), a run-through direction (D) with respect to the network path (12a, 12b, ...) is determined, due to determined match of the passage direction (D) of the movement path (2) and as the future travel path (3) of the vehicle (1) a partial path of the selected network path (12a, 12b, ...) is selected in the direction of passage (D) of the position (sy) of the vehicle (1) follows.
    [3]
    3. The method according to any one of the preceding claims, characterized in that the database is made available by one or more calibration trips are made with the vehicle (1) along the predetermined routes within the network, and that during the journey of the vehicle (1 ) are respectively recorded movement paths (2) of the vehicle (1) and added as network paths (12a, 12b, ...) of the database.
    [4]
    4. The method according to any one of the preceding claims, characterized in that in the determination of the movement path (2) or network path (12a, 12b, ...) each point on the path (2,12a, 12b, ...) each have a Orientation of the networked vehicle (1), in particular comprising three linearly independent directional vectors relating to the orientation of the vehicle (1), associated with the path and stored in the database, and in particular that the future travel path (3) 12a, 12b, ...) associated with orientation information.
    [5]
    5. The method according to any one of the preceding claims, characterized in that in each case a route identifier is predetermined in both the calibration trips and during the trips, wherein the route identifier the network paths (12a, 12b, ...) is assigned to the database as an additional attribute and in the determination or search of the network path (12a, 12b, ...) from the database or for the comparison of the network path (12a, 12b, ...) with the movement path (2) only in the database located network paths (12a , 12b, ...) to which the same route identifier has been assigned.
    [6]
    6. The method according to any one of the preceding claims, characterized in that during the calibration drive the individual points on the network paths (12a, 12b, ...) in each case a geographical location, in particular in the form of coordinates assigned, and that while driving to predetermined time points, in particular when each point on the network path (12a, 12b, ...) is recorded, a geographical localization is undertaken and assigned to the respective point on the network path, and - during the journey, the geographical position of the vehicle (1) and only network paths (12a, 12b, ...) from the database are used for the comparison whose associated geographical localization information is within a given environment around the geographical position recorded during the journey.
    [7]
    7. The method according to any one of the preceding claims, characterized in that for comparison of a movement path (2) with a network path (12a, 12b, ...) in each case that of the last traveled distance of a predetermined length preceding partial path of the movement path (2) used and only this partial path is used as the movement path (2) for the comparison with the network paths (12a, 12b,...).
    [8]
    8. The method according to any one of the preceding claims, characterized in that in the absence of coincidence of a movement path (2) with the network paths stored in the database (12a, 12b, ...) or with multiple match of a determined movement path (2) with in the database stored network paths (12a, 12b, ...), the distance used for the comparison of the paths is increased or decreased until a clear match of the path with a single stored in the database network path (12a, 12b, ... ) is present.
    [9]
    9. The method according to any one of the preceding claims, characterized in that in the recording of movement paths (2) or network paths (12a, 12b, ...), a position determination is carried out at discrete times, preferably at the next following time have the same time interval , where for each waypoint respectively the relative position of the next following waypoint is given.
    [10]
    10. The method according to claim 9, characterized in that for each waypoint, the relative rotation of the vehicle (1) or change in the orientation of the vehicle (1) relative to the orientation of the vehicle (1) is given in the respective preceding path.
    [11]
    11. The method according to any one of the preceding claims, characterized in that the movement paths (2) and the network paths (12a, 12b, ...) in the form of a curve function p (s) are given relative to a starting point, - that the so-described Path is possibly projected onto a predetermined, in particular horizontally extending, plane - that subsequently the second derivative of the projection representing function pxy (s) is determined according to the distance s as the curvature vector function K (s), - that from the thus determined curvature vector function K (S) if necessary, a scalar, signed curvature function K (s) is determined, and - that for comparison of a network path (12a, 12b, ...) with a movement path (2) the course of the curvature vector function or curvature function of the network path (12a, 12b , ...) is compared with the course of the curvature vector function or curvature function of the path (2), in particular correlated, and - that position (sy) of the network path (12a, 12b, ...) and is determined, which coincides with the movement path (2) of the vehicle (1) best, - that this determined position (sy) of the current position of the networked vehicle (1 ) is maintained and - if necessary, that the sub-path of the found network path (12a, 12b, ...) following this position (sy) is viewed and made available as a future path (3) of the network-bound vehicle (1).
    [12]
    12. The method according to any one of the preceding claims, characterized in that - at a time during the journey, each individual network path stored in the database is compared with the path determined by the networked vehicle (1), wherein the network path in which the vehicle ( 1) traveled path has been found for the determination of the position s of the vehicle (1) and for the determination of the future track course of the vehicle (1) is used, and - that in further subsequent determinations of the position, and possibly the future travel path (3) , the networked vehicle (1) exclusively in this selected network path (12a, 12b, ...) is searched for matches of the movement path (2) traveled by the vehicle (1) with this network path (12a, 12b, ...), and - That in particular wherein the other network paths (12a, 12b, ...) in the determination of the position and / or the future travel path (3) no longer used In particular, as long as the movement path in the selected network path (12a, 12b, ...) can be found.
    [13]
    13. The method according to any one of the preceding claims, characterized in that the determination of the network paths (12a, 12b, ...) and the determination of the movement path (2) by means of a camera, wherein the camera image is supplied to a visual odometry method, with which the respective path of travel (2) covered by the networked vehicle (1) or network path (12a, 12b, ...) is determined.
    [14]
    14. The method according to any one of the preceding claims, characterized in that the determination of the network paths (12a, 12b, ...) by a one-time extraction to be carried out and suitable further processing of data from an existing map works done.
    [15]
    15. The method according to any one of the preceding claims, characterized in that starting from the networked vehicle (1) a search for objects (21,22, 23, 24) in the vicinity of the vehicle (1) is made, said search in particular by means of a stereoscopic multi-camera arrangement, wherein a number of points or objects are created with respect to the cameras, wherein the determined future path (3) present in relation to the vehicle (1) is determined with respect to the two cameras and detected Objects in the area of the future path (3) is searched, and when finding such objects (22) in the area of the future path (3) a warning is issued that there are obstacles in the future travel path (3) of the vehicle (1).
    [16]
    16. Disk on which a program for performing a method according to one of the preceding claims is stored.
    类似技术:
    公开号 | 公开日 | 专利标题
    DE102010006828B4|2021-07-08|Method for the automatic creation of a model of the surroundings of a vehicle as well as driver assistance system and vehicle
    EP3271231B1|2018-12-26|Method and device for monitoring a target trajectory to be travelled by a vehicle for absence of collisions
    DE112018002143T5|2020-01-16|SYSTEMS AND METHODS FOR COMPRESSING TRAFFIC DATA
    EP3584663B1|2022-03-09|Method for automatic transverse guidance of a follow vehicle in a vehicle platoon
    DE102015203016B4|2019-08-01|Method and device for optical self-localization of a motor vehicle in an environment
    DE102012014397B4|2016-05-19|Method for determining a position of a vehicle and vehicle
    DE102017206847B4|2020-08-13|Method for the automatic creation and updating of a data record for an autonomous vehicle
    DE112011105829T5|2014-08-07|Autonomous driving system
    DE102016205434A1|2017-06-01|Method and system for creating a lane-accurate occupancy map for lanes
    DE102018208182A1|2019-11-28|Method and device for carrying out at least one safety-enhancing measure for a vehicle
    DE102016213782A1|2018-02-01|A method, apparatus and computer readable storage medium having instructions for determining the lateral position of a vehicle relative to the lanes of a lane
    DE102016220308A1|2018-04-19|System and method for generating digital road models from aerial or satellite imagery and vehicle-acquired data
    EP3227159B1|2018-12-05|Method for determining the position of a vehicle
    DE102014212866A1|2016-01-07|Method for determining a parking space from a number of measuring points
    EP3161808A1|2017-05-03|Method for processing measurement data of a vehicle in order to determine the start of a search for a parking space and computer program product
    DE102016213817B4|2019-03-07|A method, apparatus and computer readable storage medium having instructions for determining the lateral position of a vehicle relative to the lanes of a lane
    DE102020100749A1|2020-07-16|DETERMINATION OF ROUTES FOR VEHICLES
    DE102010033372A1|2012-02-09|Method for determining position of rail vehicle, involves detecting surroundings of rail vehicle by a sensor, and creating model of environment of rail vehicle
    EP3707468B1|2021-12-08|System and method for navigating within a track network
    EP2995971A1|2016-03-16|Method and system for positioning
    WO2012163632A1|2012-12-06|Method for determining the position of moving objects
    DE102017210131A1|2018-12-20|Method, computer program product and rail vehicle, in particular rail vehicle, for lane detection in rail traffic, in particular for track identification in rail transport
    WO2019052867A1|2019-03-21|Method for determining a course of driving lanes of a road network, and server device for carrying out the method
    DE102016212786A1|2018-01-18|Driver assistance system for an ego vehicle on a path of a road network and method for operating the driver assistance system
    DE102020208378B3|2021-09-23|Traffic light lane assignment from swarm data
    同族专利:
    公开号 | 公开日
    EP3227159A1|2017-10-11|
    EP3227159B1|2018-12-05|
    AT516553B1|2018-02-15|
    WO2016086249A1|2016-06-09|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    DE102004029543A1|2004-04-07|2005-10-27|Daimlerchrysler Ag|Vehicle position determination procedure smooths the digital map profile and measured values along possible path using low pass filters|
    DE102006007788A1|2006-02-20|2007-08-30|Siemens Ag|Computer-assisted driverless railway train monitoring system, to show its travel behavior, has train-mounted sensors and track position markers for position data to be compared with a stored model|
    DE102006062061A1|2006-12-29|2008-07-03|Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V.|Device, method, and computer program for determining a position on the basis of a camera image|
    US20090105893A1|2007-10-18|2009-04-23|Wabtec Holding Corp.|System and Method to Determine Train Location in a Track Network|EP3456606A1|2017-09-15|2019-03-20|Aktiebolaget SKF|Position determination method and system|US6374184B1|1999-09-10|2002-04-16|Ge-Harris Railway Electronics, Llc|Methods and apparatus for determining that a train has changed paths|DE102017104202A1|2017-03-01|2018-09-06|Knorr-Bremse Systeme für Schienenfahrzeuge GmbH|Method for the continuous adaptation and extension of track data for a rail vehicle|
    法律状态:
    优先权:
    申请号 | 申请日 | 专利标题
    ATA50886/2014A|AT516553B1|2014-12-05|2014-12-05|Method for determining the position of a vehicle|ATA50886/2014A| AT516553B1|2014-12-05|2014-12-05|Method for determining the position of a vehicle|
    PCT/AT2015/050250| WO2016086249A1|2014-12-05|2015-10-08|Method for determining the position of a vehicle|
    EP15793679.0A| EP3227159B1|2014-12-05|2015-10-08|Method for determining the position of a vehicle|
    [返回顶部]