Method and system for determining a corrected trajectory of a vehicle

By detecting the deviation between the vehicle trajectory section and the preferred line and performing average correction, the positioning offset problem in the prior art is solved, and the precise determination of the vehicle trajectory is achieved, and the cost and calculation complexity are reduced.

CN113165661BActive Publication Date: 2024-12-06VOLKSWAGEN AG
View PDF 18 Cites 0 Cited by

Patent Information

Application Number
CN201980083784.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2018-12-17
Filing Date
2019-11-25
Publication Date
2024-12-06
Estimated Expiration
2039-11-25

AI Technical Summary

Technical Problem

The prior art has positioning offset problems when determining vehicle trajectory, especially when using GNSS or map-based positioning systems, resulting in inaccurate position estimation, especially in urban areas and complex environments, and high-precision positioning. The system is costly and computationally intensive.

Method used

By detecting the comparison of the trajectory section of the vehicle with the preferred line, the positioning offset is determined, and the trajectory deviation is corrected by averaging over a longer period of time or distance, the trajectory section is detected using a global navigation satellite system or other positioning method, and Optimize with map data to achieve accurate matching of trajectories.

Benefits of technology

It is realized that the correction of mispositioning is simplified without adding high-value sensors and devices, and the accuracy of vehicle trajectory is improved, especially suitable for non-linear movements, reducing positioning costs and computing complexity.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113165661B_ABST
    Figure CN113165661B_ABST
Patent Text Reader

Abstract

In a method for determining a trajectory of a vehicle (1), map data about a traffic route (22) on which the vehicle (1) moves are provided, wherein the map data include a preferred route (24) for traveling on the traffic route (22). A trajectory segment of the vehicle (1) is detected. A positioning offset (28) is determined based on a comparison of the detected trajectory segment with the preferred route (24), and a correction trajectory segment is determined based on the positioning offset (28) and the detected trajectory segment. The positioning offset (28) is determined in such a way that the deviation between the preferred route (24) and the correction trajectory segment is optimized. A positioning system for determining a trajectory of a vehicle (1) comprises a map data providing unit (8) for providing map data about a traffic route (22) on which the vehicle (1) moves, wherein the map data include a preferred route (24) for traveling on the traffic route (22). The system further comprises: a positioning unit (5) for detecting the trajectory segment of the vehicle (1); and a computing unit (6) coupled to the positioning unit (5) and the map data providing unit (8). The calculation unit (6) is configured to: determine a positioning offset (28) based on a comparison of the detected trajectory segment with the preferred route (24); determine a correction trajectory segment based on the positioning offset (28) and the detected trajectory segment; and determine the positioning offset (28) such that a deviation between the preferred route (24) and the correction trajectory segment is optimized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a method for determining the trajectory of a vehicle. In this case, map data about a traffic route on which the vehicle moves are provided, wherein the map data include a preferred route for traveling on the traffic route. A trajectory segment of the vehicle is detected. The invention also relates to a positioning system for determining the trajectory of a vehicle. The system comprises a map data providing unit for providing map data about a traffic route on which the vehicle moves, wherein the map data include a preferred route for traveling on the traffic route. The system also comprises: a positioning unit for detecting the trajectory segment of the vehicle; and a computing unit, which is coupled to the positioning unit and the map data providing unit. Background Art

[0002] Modern vehicles often include a number of driver assistance systems, which assist in the control of the vehicle and allow the automation of driving functions to varying degrees. This can lead to fully autonomous control of the vehicle. For many driver assistance functions, highly accurate knowledge of the global vehicle position is required. The relative position with respect to a digital map with lane information is often also of particular interest, for example for determining the relative position of the vehicle with respect to the center of the lane. It has been shown that the detected position is often displaced with respect to a highly accurate map-based reference system.

[0003] The reasons for this depend on the positioning system used. The quality of the estimated position of a receiver of a global navigation satellite system (GNSS) can be affected, for example, by multipath effects, interference with satellite signals in the ionosphere or troposphere, or clock errors. Map-based positioning systems can, for example, have offsets due to insufficient calibration of the sensor devices used (e.g. lidar, camera, radar) or due to a lack of map material.

[0004] Thus, the position estimate of a positioning system often has an offset which is more or less constant within a certain time interval. This time interval depends on the specific positioning system and its environmental conditions. For example, for the global positioning system GPS (global positioning system) this is of the order of 30 to 90 seconds, but this time may also last longer in urban areas with dense buildings and thus poor signal reception. The reason for this is, for example, that only a certain number of navigation satellites can be realized within this time period.

[0005] In order to overcome or circumvent these difficulties, highly accurate GPS systems with accuracy in the centimeter range can be used, such as RTK-GPS systems or systems with ground correction data. However, such systems may lead to huge costs and are associated with high computational costs. Similar reference systems are also based on imaging sensor devices.

[0006] US 2017 / 0247032 A1 describes a method in which a corrected path is determined for a vehicle with a trailer in order to follow the course of a bend.

[0007] DE 102015214338 A1 describes a method in which the arrangement information of a vehicle is determined. For this purpose, a sensor detects the arrangement of lanes in the surroundings of the vehicle and compares the arrangement of lanes with map-based lanes. Based on these data, it is determined which lane the vehicle is located on.

[0008] DE 10 2016 213 782 A1 and DE 10 2016 213 783 A1 propose methods for determining the lateral position of a vehicle relative to a lane of a roadway. Here, an image capture unit detects lane markings and also determines whether these data are sufficient to determine the lateral position.

[0009] DE 102017006142 A1 describes a method for locating a vehicle within a lane, wherein the most likely position is estimated using a particle filter and an odometer model. A map with polygons can also be generated from a map with center lines and can be used to determine the stop probability.

[0010] DE 102010005293 A1 describes a method for lane path estimation using a sensor combination. Multiple sensors of a vehicle are monitored and particle points are generated, which are potential paths for the vehicle. In an iterative process, these particle points are used for navigation and are discarded after the vehicle has passed through them. Summary of the invention

[0011] The object of the present invention is to provide a method and a system of the type mentioned in the introduction which allow a particularly precise determination of the trajectory travelled by a vehicle using simple means.

[0012] According to the invention, this object is achieved by a method for determining a corrected trajectory of a vehicle and a system for determining a corrected trajectory of a vehicle. Advantageous embodiments and developments are described below.

[0013] The method according to the invention of the type mentioned at the outset is characterized in that a positioning offset is determined based on a comparison of the detected trajectory segment with a preferred path, and a correction trajectory segment is determined based on the positioning offset and the detected trajectory segment, wherein the positioning offset is determined such that the deviation between the preferred path and the correction trajectory segment is optimized.

[0014] Incorrect positioning can thus be corrected in an advantageously simple manner and without complex additional sensors and devices. The method is based on the assumption that the vehicle is actually moving substantially along a preferred route. The trajectory segment is considered as a whole, so that temporary deviations from the preferred route can be compensated by averaging over a longer period of time or a longer distance.

[0015] Thus, unlike known methods, the entire segment of the trajectory traveled by the vehicle is taken into account, rather than only performing a correction between the detected position and the actual position for a single position. In addition, in the case of the present invention, the course of the trajectory segment can also be used, for example in the case of non-linear movement, in order to determine the positioning offset particularly accurately.

[0016] In the method, map data are provided in a manner known per se, for example by a navigation system or a storage unit of the vehicle. These map data can also be provided by an external unit, such as an external server, from which the map data required for a specific geographical area are called up if the vehicle is approaching or moving within this area, or if the planned travel of the vehicle extends through this area.

[0017] These map data include in particular information about the geographical classification of traffic routes and about the lanes formed on the traffic routes. Here, these map data include a geometric description of the traffic routes, in particular a geometric description with information about preferred routes.

[0018] A preferred route for the purposes of the present invention is a route extending in the longitudinal direction of a traffic route. This route describes, for example, a trajectory along which a vehicle can move on the traffic route. In particular, it is lane center information or a center line or "centerline" included in the highly accurate maps that are usually used. In this case, the preferred route can extend, for example, in the geometric center of the lane along its longitudinal direction. It can also be a theoretical trajectory of a vehicle when driving on a traffic route, for example an ideal route for optimal driving on a traffic route.

[0019] In this case, the preferred route may already be included in the map data provided, for example, by a navigation system. Alternatively or additionally, preprocessing of the provided map data may be provided, wherein the preferred route is generated, for example, by a preprocessing unit of the vehicle. The preferred route may also be designed as a planned trajectory of the vehicle, in particular for autonomous control of the vehicle.

[0020] The trajectory segments of the vehicle are detected in a manner known per se, for example by means of a global navigation satellite system, such as GPS. The trajectory segments can also be detected based on other positioning methods, for example by landmark-based positioning. The trajectory segments can also be detected as planned trajectories of the vehicle, for example by a navigation unit or a unit for planning autonomous control of the vehicle. The detection of the trajectory segments can take place in a global or relative coordinate system.

[0021] The determined positioning offset corresponds in particular to an offset vector, which represents a vector from the measured position to the actual position. In this case, these positions can be determined in a global coordinate system or can be described in a local coordinate system relative to a specific reference point, usually depending on the method used for determining the position. By means of the positioning offset, the detected trajectory segment can be corrected so that it matches the preferred route as well as possible and is thus as close as possible to the actual traveled trajectory. In this case, an optimization problem is solved, in which in particular the deviation between the preferred route and the trajectory segment corrected according to the positioning offset is reduced to a minimum. In order to solve this optimization problem, statistical methods known per se can be used.

[0022] The positioning offset determined in the method can be output, for example by transmission to a driver assistance system or another device. The offset of the position measurement relative to the map data determined using this positioning offset can now be used to correct the detected track section of the vehicle. This can achieve that the detected track lies substantially or on average over a certain time or length region on the preferred route. Such a track is considered a plausible track and can be used subsequently.

[0023] In the method, the positioning offset can be determined in the vehicle coordinate system or other relative coordinate systems and / or in the global coordinate system. If there is information about the expected positioning error, such as about its type or form of expression, the system can be limited to this method, for example, in order to determine the positioning offset only in the global coordinate system. This can make sense, for example, in the case of receivers of global navigation satellite systems. For map-based positioning systems, for example, the relative positioning error can be determined in the global coordinate system.

[0024] In one embodiment of the method according to the invention, the detected trajectory segment includes a plurality of temporally consecutive position measurements, in particular a plurality of temporally consecutive position measurements in a global coordinate system. The movement of the vehicle is thus advantageously described by discrete position measurements. The plurality of measurements permits a reliable statistical analysis of the data in order to precisely determine the course of travel of the vehicle despite possible fluctuations.

[0025] In the method, the detected track segment is compared with the preferred route in order to determine the positioning offset in order to correct the position measurement for the track segment. The method is based in particular on the fact that the vehicle always moves close to the preferred route on average when traveling along the track segment. Smaller deviations to the left or right are compensated in the case of averaging over time or space intervals or other statistical analyses. In order to be able to use this assumption as a starting point, sufficient data about the track must be available for the statistical analysis.

[0026] In another embodiment, the detected trajectory segments have position measurements during a specific time interval, within a specific driving route and / or a specific driving route length. Thus, the detected preferred route advantageously has a specific spatial or temporal length, which is defined, for example, by a specific number of measurements, detections within a specific time period or along a specific route length.

[0027] Thus, data detected within a suitable length of the travel route or within a suitable length of the time interval are provided, which are compared with the preferred route. For example, it can be assumed that in the case of averaging over at least 20 seconds, preferably at least 30 seconds, the deviation has been significantly reduced. In another example, measurements within a certain length of the track segment can be considered, for example, measurements within at least 150m, preferably at least 250m. If, for example, the prerequisite is detection within a certain time period and a minimum spatial length for the available detection at the same time, the length of the track segment can also be determined by a combination of spatial and temporal lengths.

[0028] In one embodiment, the trajectory segments are detected by means of a sliding window algorithm, in particular a so-called “Sliding Window.” Advantageously, the trajectory segment of the last travel is thus taken into account in order to determine the position offset.

[0029] In the case of a sliding window algorithm, it can be provided, for example, that the positioning offset is determined for a specific point in time or a specific position, wherein the detected trajectory segment only includes data previously detected within a specific time period or length range. In this way, the detected trajectory segment is continuously "shifted" during the movement of the vehicle or over time. For example, the time or spatial length used to detect the trajectory segment is limited. When executing this method, data outside this window are not taken into account, so that, for example, deviations from the preferred route only affect the calculation within a limited time period. On the contrary, the width of the window is selected so that a sufficient database is provided for statistical analysis.

[0030] The width of the sliding window within which the track segment is detected corresponds in particular to the time or spatial length of the detected track segment. This time or spatial length can be fixed or dynamically determined. For example, different widths can be selected depending on the driving situation or speed of the vehicle.

[0031] In one embodiment of the method, the detected trajectory segment is smoothed when comparing the trajectory segment with the preferred route. Thus, for example, statistical fluctuations of the trajectory are no longer taken into account. Alternatively or additionally, a plurality of deviations of the detected trajectory segment from the preferred route segment can be determined and averaged. Alternatively or additionally, other statistical methods can be used, for example statistical methods with weights, in which deviations from the preferred route are taken into account more strongly or less strongly depending on the amplitude of these deviations.

[0032] In one embodiment of the invention, when comparing the detected trajectory segment with the preferred route, a preferred route segment is determined, which corresponds to the segment of the traffic route traveled by the vehicle. Thus, the detected trajectory segment is advantageously compared with a matching segment of the preferred route.

[0033] The preferred route section is determined in a manner known per se. For example, it can be assumed that the vehicle is located at the nearest position of the preferred route. The preferred route section can also be determined based on the route planned for the vehicle, as long as it can be assumed that the vehicle actually moves along this route.

[0034] In another embodiment, when determining the preferred route section, it is determined on which lane of the traffic route the vehicle is located. This advantageously makes it possible to determine particularly precisely which preferred route of the map data the detected trajectory section should be compared with.

[0035] The current lane can be determined in a manner known per se, for example from video data from a camera, by means of other sensors such as lidar, radar or from information from a navigation system which specifies the lane to be used. Alternatively or additionally, an odometer method can be performed or information about the choice of lane, such as the actuation of a turn signal indicator, can be evaluated.

[0036] In one embodiment, the lateral position of the vehicle relative to the traffic route is detected when the trajectory segment is detected. In particular, the so-called "lateral deviation" of the vehicle is determined. This advantageously makes it possible to check particularly precisely whether the detected trajectory segment actually runs along the preferred route and whether and to what extent the actually driven trajectory deviates from the preferred route. This allows a refinement of the measurement by comparing the measured trajectory segment with the trajectory actually driven on the traffic route.

[0037] Here, too, methods known per se can be used, for example by means of a camera, a radar or lidar scanner, a local device for positioning or other sensors.

[0038] The assumption that the vehicle moves substantially along a trajectory segment may be foreseeably violated in certain situations, such as when the vehicle performs a lane change, cuts into a curve or travels in roundabout traffic. However, since the duration of such events is usually very short, these events often only affect the average value insignificantly. However, measures can also be provided to avoid distortion of the measured values, for example by not detecting the data of this trajectory segment in the corresponding situation or by not taking into account the traveled segment in which the problematic situation occurs in the comparison with the preferred route.

[0039] In one embodiment of the method, when a track segment is detected, a usable area of ​​the track segment is determined, based on which the comparison with the preferred route is performed. The detected track segments can thus advantageously be limited to areas in which the assumptions on which the method is based can be considered to be satisfied.

[0040] For example, a filter can be implemented by which the method is activated only in such areas in which it can be safely assumed that the vehicle is actually moving along the preferred route. Then, the trajectory segments detected in such areas can be used as a reference and for comparison with the preferred route.

[0041] In another embodiment, the available area is determined based on the vehicle's intention to deviate. In this case, in particular, the intention to perform a driving maneuver such as changing lanes, overtaking or avoiding is detected, or the end of such a driving maneuver is detected. As a result, the available area can be advantageously determined particularly reliably.

[0042] In one embodiment, the available area is determined as a function of the course of the traffic route, for example as a function of the position of a bend. This advantageously allows particularly reliable predictions of events and driving maneuvers, in which deviations from the preferred route are taken into account. This is the case, for example, at bends, at intersections, at construction sites, at roundabouts, at exits or at locations where parked vehicles must be avoided.

[0043] In one embodiment, the usable area is determined using a machine learning method, thereby making it possible to advantageously carry out a particularly reliable determination of the usable area.

[0044] The machine learning method can be implemented in a manner known per se. In particular, training is performed based on data detected in the past, in which, for example, deviations from a preferred route are detected based on specific characteristics of the traffic route or other current conditions. For example, in this way it is possible to learn at which locations an avoidance is required with respect to vehicles parked at the side of the road, which leads to deviations from the preferred route. The influence of the weather on the deviations can also be determined.

[0045] In another embodiment, a measurement accuracy is determined for detecting the track segment and the length of the detected track segment is determined as a function of the measurement accuracy. The track segment can thus advantageously be detected in such a way that the measurements allow statistically significant information to be derived about the vehicle travel.

[0046] The measurement inaccuracy can be determined, for example, based on the standard deviation or other measures for the uncertainty of the position measurement. The device for detecting the track segment can also provide the data required for this. For example, a receiver of a global navigation satellite system can provide information about how many navigation satellites are currently visible, how well these navigation satellites are received, and what influence this has on the accuracy of the detected track segment.

[0047] In particular, the greater the uncertainty associated with the detection, the longer the trajectory segments are detected. Conversely, in the case of lower measurement uncertainties, shorter trajectory segments can be detected and used for the method.

[0048] In one embodiment, a change in the measurement uncertainty for detecting the initial trajectory is detected and the detection of the trajectory segment is started or ended as a function of this change. This advantageously makes it possible to detect when the conditions for detecting the trajectory segment and, if applicable, the type of measurement error caused thereby, have changed.

[0049] In the event of a change in the measurement uncertainty, it can be ascertained whether the measured position or estimate has a greater or lesser statistical uncertainty. Furthermore, a change in the measurement uncertainty can lead to a change in the systematic offset between the measured trajectory and the actual trajectory. Thus, in particular, such a change can be ascertained when the deviation of the detected position from the actual position changes abruptly and thus has to be corrected using a further positioning offset. Such a change can occur, for example, when the number of navigation satellites visible at a point in time changes.

[0050] Changes in the measurement uncertainty can be detected in different ways, for example based on data provided by a unit for detecting the track segment. In addition, with respect to the detected track segment, it can be checked whether sudden changes in position occur, in particular when these changes in position are identified as unreasonable in the physical model. For example, if a sudden change in position of the vehicle is measured despite the vehicle not moving or moving at a lower speed or if the direction of the change in position does not match other movements of the vehicle, a detection can be performed. In particular, an odometer is used for this purpose in order to detect the actual movement of the vehicle or to determine the plausibility of the detected change in position depending on the type of vehicle.

[0051] For example, the time derivative of the detected vehicle movement along the trajectory segment can be determined. If a certain threshold value of the time-dependent position change is exceeded, this can be interpreted as a "jump" and evaluated as an indication for a positioning offset to be re-determined. In the case of a detected trajectory segment, a reset can be performed, in particular in the case of a sliding window method, so that the detection of the trajectory segment is restarted.

[0052] In a further embodiment of the method, an undercalibration of a unit for detecting a track segment can be determined and / or characterized based on the determined positioning offset. Such undercalibration often leads to systematic errors in the detection of track segments, which can be detected by the method according to the invention. In addition, the type or nature of the positioning offset can make it possible to infer which component is incorrectly calibrated or in what way the calibration must be corrected. It can be provided that an incorrect calibration is corrected based on the determined positioning offset or that error data are output, based on which, for example, a correction of the calibration can be carried out.

[0053] In another embodiment, the determination of the positioning offset is carried out by an external unit, for example by an external server. This advantageously allows the use of computing power of the vehicle to be economical. The external unit is connected to the vehicle in a data-technical manner and receives the detected track segments. In addition, map data can be transmitted from the vehicle to the external unit, or the external unit can in turn provide map data for the vehicle. The external unit can now determine the positioning offset and, if necessary, transmit it to the vehicle.

[0054] In another embodiment, alternatively or additionally, the positioning offset includes a correction of the orientation of the vehicle. Thus, the correction can be applied not only to the trajectory traveled by the vehicle or the position of the vehicle, but also to the orientation of the vehicle in space. The combination of position and orientation is usually referred to as "attitude" and is often determined together. Systematic errors that may also occur here, especially systematic errors in certain time periods, can also be corrected using the method according to the invention. In this regard, it is particularly taken into account that the orientation of the vehicle is usually related to the movement orientation, that is, the vehicle is oriented in the direction of the movement during a straight-line movement.

[0055] A positioning system according to the invention of the type mentioned at the outset is characterized in that the computing unit is configured to: determine a positioning offset based on a comparison of a detected trajectory segment with a preferred route; determine a corrected trajectory segment based on the positioning offset and the detected trajectory segment; and determine the positioning offset so that the deviation between the preferred route and the corrected trajectory segment is optimized.

[0056] The device according to the invention is especially designed to implement the method according to the invention described above. Therefore, the device has the same advantages as the method according to the invention.

[0057] The embodiment of the positioning system according to the invention comprises a detection unit coupled to a computing unit for detecting surroundings data in the surroundings of the vehicle. The computing unit is also configured to determine a usable area of ​​the trajectory segment based on the detected surroundings data, based on which a comparison with the preferred route is performed. This advantageously allows the positioning deviation to be determined particularly reliably, since, for example, unusable areas in which the actual position of the vehicle deviates significantly from the preferred route can be excluded from the analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Now, the present invention will be explained based on embodiments with reference to the accompanying drawings.

[0059] Figure 1 A vehicle having an embodiment of a positioning system according to the present invention is shown;

[0060] Figure 2 showing the vehicle as it travels along a traffic route; and

[0061] Figure 3 An exemplary embodiment of the method according to the invention is shown. DETAILED DESCRIPTION

[0062] refer to Figure 1 An embodiment of a positioning system according to the invention is described.

[0063] The vehicle 1 comprises a positioning unit 5, which in this embodiment is based on a global navigation satellite system, in particular a global positioning system GPS. The vehicle also comprises a detection unit 2, which in this embodiment comprises a camera 3 and a preprocessing unit 4. In other embodiments, alternatively or additionally, the detection unit 2 may comprise any sensor device, such as a lidar or radar scanner, a time-of-flight camera, a stereo camera, an infrared camera, an ultrasonic sensor or other sensors.

[0064] The vehicle 1 further comprises a computing unit 6 which comprises a navigation unit 8 and is coupled to the detection unit 2 , the positioning unit 5 and the driver assistance system 7 .

[0065] refer to Figure 2 and 3 Exemplary embodiments of the method according to the invention have been described. The above-described exemplary embodiments of the locating system according to the invention are used as a starting point.

[0066] The vehicle 1 is traveling on the traffic route 22 in the direction indicated by the arrow 25. Two lanes are formed on the traffic route 22, which are defined by the side boundaries of the traffic route 22 drawn in solid lines and the dashed center strip 23. The vehicle 1 is located on the lane on the right in the driving direction 25. A preferred route 24 is also drawn, which in this embodiment runs essentially centrally along the lane of the vehicle 1.

[0067] In this exemplary embodiment, it is assumed that the vehicle 1 is controlled along an actual trajectory which runs essentially centrally on the traffic route 22 and thus corresponds largely to the preferred route 24. Thus, in this exemplary embodiment, the trajectory of the actual travel of the vehicle 1 differs from the preferred route 24 only by slight statistical deviations in the transverse direction.

[0068] In the method, map data are provided in vehicle 1 by navigation unit 8 , ie, the navigation unit acts as map data providing unit 8 , which map data include information about the course of traffic route 22 , its lane divisions, the course of center line 23 and the course of preferred route 24 .

[0069] In other exemplary embodiments, it may be provided that, alternatively or additionally, map data are provided by an external unit, which map data supplement or alternatively provide this information. In this case, the external unit is detected via a data technology connection, for example upon request of the vehicle or by automatically transmitting the relevant map data to the vehicle 1. In particular, the per se known “streaming” of the map data relevant to the vehicle 1 is carried out via the Internet.

[0070] While the vehicle 1 is traveling along the traffic route 22— Figure 2 and Figure 3 The actual trajectory corresponds essentially to the preferred route 24 and is therefore not shown separately - the geographical position 26 of the vehicle 1 is determined at intervals by the positioning unit 5 of the vehicle 1 and transmitted to the calculation unit 6. In this exemplary embodiment, this determination is performed with the aid of GPS in global coordinates. In other exemplary embodiments, alternatively or additionally, other positioning can be performed, for example with the aid of marker-based positioning, and the position can also be determined in a relative coordinate system.

[0071] As in Figure 2 and 3 It can be seen that the detected positions 26 of the vehicle 1 do not follow the trajectory or preferred route 24 actually traveled by the vehicle 1, but are offset therefrom. Furthermore, the detected positions 26 have a statistical distribution which takes into account that the actual trajectory of the vehicle 1 does not correspond exactly to the preferred route 24, as would be expected, for example, for a rail vehicle. However, in the case of the actual trajectory, the deviations from the preferred route 24 are designed such that they occur only briefly compared to the length of the entire trajectory segment shown.

[0072] The calculation unit 6 determines a trajectory segment based on the detected positions 26, which in this embodiment includes all the positions 26 shown. In this embodiment, the length of the trajectory segment is defined as a collection of position measurements that are detected within a specific time interval, here 30 seconds. In other embodiments, another time interval can be provided, such as 1 minute, or the length of the trajectory segment can be defined based on its spatial length and include, for example, a driving route of 200m or 400m.

[0073] The detection of the trajectory segments is carried out with the aid of a "sliding window", wherein the detected trajectory segments are continuously shifted over time and with the continuous movement of the vehicle 1. The temporal and / or spatial limitation of the length of the trajectory segments can be referred to as the "width" of the sliding window. This width is constant in this embodiment, but it can also be configured dynamically in other embodiments, for example as a function of the driving situation, speed, measurement inaccuracies provided by the positioning unit 5 or user input.

[0074] The calculation unit 6 determines the position offset 28 of the detected trajectory segment by determining the segment of the preferred route 24 that corresponds to the detected trajectory segment. In the exemplary embodiment shown, this is essentially the entire preferred route 24 shown. This is based on the assumption that the trajectory of the vehicle 1 actually driven on the traffic route 22 can be well approximated by the preferred route 24. In order to determine the relevant segment of the preferred route 24, in this exemplary embodiment, the detection unit 2 is also used to determine which lane of the traffic route 22 the vehicle 1 is located on.

[0075] Then, a comparison is performed, in the case of which the calculation unit 6 determines the deviation of the detected positions 26 relative to the preferred route 24 or the determined preferred route section. In the present embodiment, this is the geometric distance of the detected positions 26 relative to the respective nearest point of the preferred route 24; however, other measures can be used in other embodiments. An optimization is performed, in the case of which the distance of the detected positions 26 from the preferred route 24 is reduced to a minimum. In the case of this optimization, a positioning offset 28 is determined, which corresponds to the offset vector 28 and describes the spatial displacement of the detected positions 26 so that these detected positions are optimally displaced with respect to the smallest possible deviation from the preferred route.

[0076] In other exemplary embodiments, statistical methods can be used for the detected trajectory sections in order to, for example, perform smoothing or averaging and compensate for short-term deviations from the smoothed trajectory; this can facilitate the comparison since low-frequency deviations no longer have to be taken into account.

[0077] Depending on the positioning offset 28 determined in this way, the computing unit 6 corrects the detected position 26 of the trajectory segment in order to reach a corrected position 27. Figure 3 Detected positions 26 are shown shifted by a positioning offset 28 toward a corrected position 27. Corrected position 27 follows preferred path 24 significantly better than detected positions 26, ie a single positioning offset 28 is determined for correcting all detected positions 26 of a detected trajectory segment.

[0078] In another embodiment, the rotation by a specific angle can also be determined according to the same optimization task. This allows not only a correction for the translation correction of the detected position 26 relative to the correction position 27, but also a rotation correction.

[0079] exist Figure 3In the case shown in FIG. 2 , it can be seen that the corrected position 27 deviates significantly from the preferred route 24 in the region of a bend in the preferred route 24. This can occur, for example, when the vehicle 1 “cuts into” a bend and is therefore not travelling on the preferred route 24. If such an event has only a short duration, so that only a few detected positions 26 are detected, and the vehicle 1 deviates from the preferred route 24, then the determination of the positioning offset 28 has only a minor effect. Therefore, in this exemplary embodiment, the detected positions 26 are not filtered in order to exclude these positions.

[0080] However, in the case of events lasting longer or with greater deviations, it is provided that the computing unit 6 filters out data from the detected data in the case of actual deviations of the trajectory traveled by the vehicle 1 from the preferred route 24. For example, the computing unit 6 identifies situations in which deviations from the preferred route 24 occur, such as cutting in at a bend, changing lanes, avoiding or overtaking maneuvers, based on the data of the detection unit 2. The positions 26 detected in this area are then excluded from the determination of the positioning offset 28.

[0081] In this case, the calculation unit 6 determines the usable area of ​​the detected trajectory segment, wherein the usable area must also meet the condition of a certain length of the trajectory segment that can be used for the method. Those detected positions 26 that are detected in an unusable area, for example in the area of ​​a bend, are discarded. In addition, it is detected based on the data of the detection unit 2 whether there is an intention of the driver of the vehicle 1 to deviate from the preferred route 24, such as whether the flashing lights are activated or whether a rapid change of direction is made. In addition, information about deviations from the preferred route 24 or a corresponding intention of the driver can be detected by a driver assistance system 27, such as an overtaking assistant.

[0082] In other embodiments, machine learning methods can be used to distinguish usable areas from unusable areas of the detected trajectory segments. In this case, the system is trained in a manner known per se based on historical data and can automatically identify areas in which deviations from the preferred route 24 can be expected. Such areas may, for example, have vehicles parked at the edge of the traffic route 22 or specific devices for regulating traffic, which need to be avoided.

[0083] If, in addition, a lane change or another change of preferred route 24 is detected, then, for different subsets of detected positions 26 , deviations of these detected positions from different preferred routes 24 are determined by optimizing position offset 28 .

[0084] In another embodiment, the detection unit 2 of the vehicle 1 is also used in order to determine the position of the vehicle 1 on the traffic route 22, in particular in the transverse direction, and to take into account where the vehicle 1 is actually located relative to the preferred route 24 when determining the deviation of the measured position 26 from the preferred route 24. In this way, the positioning offset 28 can be determined more accurately, because the deviation from the actual position can now be determined without having to accept the assumption that the position of the vehicle 1 is on the preferred route 24.

[0085] In this embodiment, it is also provided that if a change in the positioning offset 28 is detected, the detection of the trajectory segment using the sliding window can be interrupted. This change is detected based on the data of the positioning unit 5, which outputs, for example, a change in the measurement uncertainty. In addition, the change is detected when the detected trajectory segment has a sudden position change that is recognized as unreasonable by the system. For example, if the conditions for detecting the vehicle position change, a jump-like position change may be detected. The position change can be checked: whether the position change is physically possible and / or whether the position change is consistent with the odometer data or other data of the vehicle 1. If this is not the case, there are conditions in which the trajectory segment is completely redetected. Therefore, it involves a "reset" in the case of detecting the trajectory segment using the sliding window method, and the trajectory segment is completely redetected again.

[0086] Reference numerals list

[0087] 1 Vehicle

[0088] 2 Detection unit

[0089] 3 Camera

[0090] 4 Preprocessing unit

[0091] 5 Positioning unit

[0092] 6 Computational Unit

[0093] 7 Driver Assistance Systems

[0094] 8 Map data providing unit, navigation unit

[0095] 22 Traffic routes and roads

[0096] 23. Middle Zone

[0097] 24 Preferred routes

[0098] 25 Arrow, direction

[0099] 26 Detected positions

[0100] 27 Correction Position

[0101] 28 Positioning offset, offset vector.

Claims

1. A method for determining a trajectory of a vehicle (1), wherein providing map data about a traffic route (22) on which the vehicle (1) moves, wherein the map data includes a preferred route (24) for traveling on the traffic route (22); and detecting a trajectory segment of the vehicle (1), It is characterized in that determining a positioning offset (28) based on a comparison of the detected trajectory segment with the preferred route (24); and determining a modified trajectory segment based on the positioning offset (28) and the detected trajectory segment, wherein the positioning offset (28) is determined such that a deviation between the preferred route (24) and the modified trajectory segment is optimized, Wherein, when the track segment is detected, an available area of ​​the track segment is determined, and a comparison with the preferred route is performed based on the available area. The available area is determined based on the deviation intention of the vehicle (1), or the available area is determined based on the direction of the traffic route (22).

2. The method according to claim 1, It is characterized in that The detected trajectory segment includes a plurality of temporally consecutive position measurements.

3. The method according to claim 2, It is characterized in that The detected trajectory segment has position measurements during a specific time interval, within a specific driving route and / or within a specific driving route length.

4. The method according to any one of claims 1 to 3, It is characterized in that The trajectory segments are detected with the aid of a sliding window algorithm.

5. The method according to any one of claims 1 to 3, It is characterized in that By comparing the detected trajectory segment with the preferred route (24), a preferred route segment is determined, which corresponds to the segment of the traffic route (22) on which the vehicle (1) travels.

6. The method according to claim 5, It is characterized in that When the preferred route section is determined, it is determined on which lane of the traffic route (22) the vehicle (1) is located.

7. The method according to any one of claims 1 to 3, It is characterized in that When detecting the trajectory section, a lateral position of the vehicle (1) relative to the traffic route (22) is detected.

8. The method according to any one of claims 1 to 3, It is characterized in that The available area is determined according to a machine learning method.

9. The method according to any one of claims 1 to 3, It is characterized in that A measuring accuracy for detecting the trajectory segment is determined and a length of the detected trajectory segment is determined as a function of the measuring accuracy.

10. The method according to any one of claims 1 to 3, It is characterized in that A change in the measurement uncertainty for detecting the initial trajectory is detected, and the detection of the trajectory segment is started or ended as a function of the change.

11. A positioning system for determining a trajectory of a vehicle (1), the positioning system comprising: a map data providing unit (8) for providing map data on a traffic route (22) on which the vehicle (1) moves, wherein the map data includes a preferred route (24) for traveling on the traffic route (22); A positioning unit (5) for detecting a track segment of the vehicle (1); a calculation unit (6), the calculation unit being coupled to the positioning unit (5) and the map data providing unit (8); and a detection unit (2) coupled to the computing unit (6) for detecting surrounding data in the surroundings of the vehicle (1), It is characterized in that The calculation unit (6) is configured as follows: determining a positioning offset (28) based on a comparison of the detected trajectory segment with the preferred route (24); determining a corrected trajectory segment based on the positioning offset (28) and the detected trajectory segment; The positioning offset (28) is determined so that the deviation between the preferred route (24) and the modified trajectory section is optimized; and determining a usable area of ​​the trajectory segment based on the detected surrounding environment data, and performing a comparison with the preferred route based on the usable area, The available area is determined based on the deviation intention of the vehicle (1), or the available area is determined based on the direction of the traffic route (22).

Citation Information

Patent Citations

  • System and method for track path estimation using a sensor association

    DE102010005293A1

  • Determination of order information for a vehicle

    DE102015214338A1

  • Method, apparatus and computer-readable storage medium with instructions for determining the lateral position of a vehicle relative to the lanes of a roadway

    DE102016213782A1

  • Method, device, and computer-readable storage medium with instructions for determining the lateral position of a vehicle relative to the lanes of a road

    DE102016213783A1

  • Method for localizing a vehicle within a traffic lane

    DE102017006142A1