Method and device for determining a vehicle pose of a vehicle

Description

[0001] The invention relates to a method and a device for determining a vehicle pose of a vehicle.

[0002] For collision-free driving of automated vehicles, knowledge of the current vehicle pose, consisting of its position and orientation, is required. To calculate this, camera and radar systems are typically used. These consist of several modules distributed throughout the vehicle. These enable excellent environmental detection and minimize errors in determining the vehicle pose.

[0003] However, this requires continuous processing and evaluation of the camera images and radar signals using suitable image and signal processing techniques. However, these are very computationally intensive and, with a required repetition rate of up to 10 Hz, require a processing unit with a powerful processor or processor cores.

[0004] To limit the associated increased costs, the vehicle pose can be determined by simply measuring the vehicle accelerations and yaw rates, followed by mathematical integration, given a known reference point (initial pose). The disadvantage of this less computationally intensive method, however, is an unavoidable drift in the vehicle pose calculated by integration, since (temperature-dependent) offsets from the acceleration and yaw rate sensors used are integrated with the actual useful signal, resulting in a continuously increasing error.

[0005] To prevent this error from becoming too large, the vehicle pose is often calculated not only by integrating the vehicle accelerations and yaw rates, but also by additionally considering the wheel rotation signals. This allows the increase in the error to be reduced to such an extent that a maximum permissible error limit is not exceeded within small to medium time intervals. This allows the current vehicle pose to be calculated with sufficient accuracy, for example, during parking maneuvers with typically short phases of forward and reverse maneuvering. However, this is only possible because it is possible to compensate for the temperature-dependent offsets of the acceleration and yaw rate sensors during repeated vehicle stops between the phases of forward and reverse maneuvering.

[0006] For longer journeys, however, there is no possibility of timely offset compensation, so that the error in calculating the vehicle pose exceeds a maximum permissible error limit, even when the signals for the wheel revolutions are also taken into account, due to the generally unknown tire slip.

[0007] US 2020 / 0192396 A1 discloses a driving control system for an autonomous vehicle. This system includes a 2D LIDAR sensor, a wheel speed sensor for detecting the vehicle's speed, and a yaw rate sensor for detecting the vehicle's angular velocity.The system further includes an error corrector for determining a straight-line situation using a LIDAR point detected by the 2D LIDAR sensor, extracting a straight lateral distance value according to the determination result, accumulating the LIDAR point according to the vehicle's trajectory detected by the wheel speed sensor and the yaw rate sensor, estimating an error between the accumulated point and the extracted straight line, and calculating and feeding back an offset correction parameter of the yaw rate sensor when the estimated error value is greater than a predetermined threshold to automatically correct an error parameter of the yaw rate sensor.

[0008] A dead reckoning unit is known from US 2008 / 0294342 A1. This unit calculates the vehicle position from a pitch angle and a yaw angle of dead reckoning sensors, a sensor installation pitch angle and a sensor installation yaw angle, and a travel distance calculated by a speed sensor, and calculates the vehicle speed from an acceleration signal. In a first cycle, a first correction unit calculates the vehicle speed from signals output by the speed sensor and corrects the pitch angle, the sensor installation pitch angle, and the sensor installation yaw angle based on the difference between the thus calculated vehicle speed and the vehicle speed calculated by the dead reckoning unit.In a second cycle, a second correction unit corrects the pitch angle, the pitch angle of the sensor installation, the yaw angle and the yaw angle of the sensor installation using the vehicle position and speed output by a GPS receiver and the vehicle position and speed output by the dead reckoning unit.

[0009] EP 1 096 230 A2 discloses an inertial navigation system for vehicles, also known as a dead reckoning system for vehicle navigation. The inertial navigation system can be used alone or in combination with other positioning means, such as GPS and map databases, to determine a vehicle's location. The dead reckoning system has several advantages over existing systems. First, it can be easily mounted on the chassis of any vehicle. Second, it does not require an interface to the vehicle's existing sensors. Third, the system includes logic to eliminate errors in position and heading determination caused by the angulation / rotation of the chassis and the inertial guidance sensors, which are caused by the inclination or tilt of the chassis with respect to an inertial / quasi-inertial reference frame, such as the Earth.The inertial guidance system comprises an inertial guidance sensor, a translation unit, and a logic unit. The inertial guidance sensor is suitable for coupling to the vehicle. The inertial guidance sensor detects the vehicle's motion in a non-inertial reference frame and generates a corresponding sensor signal. The translation unit is coupled to receive the sensor signal generated by the inertial guidance sensor. The translation unit translates the sensor signal into a quasi-inertial reference frame and generates a corresponding translated signal. The logic unit receives the translated signal generated by the translation unit and converts the translated signal into an estimated position and heading of the vehicle.

[0010] From US 2018 / 0 045 519 A1, a method for localization and mapping is known, in which an image is recorded with a camera mounted on a vehicle, the vehicle being associated with a global system location; a landmark represented in the image is identified with a landmark identification module of a computer system associated with the vehicle, the identified landmark having a geographical location of the landmark and a known parameter; extracting a set of landmark parameters from the image with a feature extraction module of the computer system; determining a relative position between the vehicle and the geographical location of the landmark in the computer system based on a comparison between the extracted set of landmark parameters and the known parameter; and updating the global system location in the computer system based on the relative position.

[0011] From EP 3 462 132 A2 a method for determining a current position of a motor vehicle in an area surrounding the motor vehicle is known, wherein inertial sensor data from an inertial sensor device of the motor vehicle and odometry data from an odometry sensor device of the motor vehicle are received by a control device of the motor vehicle and wherein a movement vector describing a movement of the motor vehicle is determined on the basis of the inertial sensor data of the motor vehicle, support data describing the movement of the motor vehicle are determined on the basis of the odometry data, the movement vector is corrected on the basis of the support data and the current position of the motor vehicle is determined as a function of the corrected movement vector.

[0012] The invention is based on the object of improving a method and a device for determining a vehicle pose of a vehicle, in particular with regard to a drift correction.

[0013] The object is achieved according to the invention by a method having the features of patent claim 1 and a device having the features of patent claim 10. Advantageous embodiments of the invention emerge from the subclaims.

[0014] In particular, a method for determining a vehicle pose of a vehicle is provided, wherein the vehicle pose is estimated by means of integration based on sensor data from at least acceleration sensors and yaw rate sensors and a known initial pose, wherein the estimated vehicle pose is corrected with the aid of at least one reference vehicle pose, wherein time intervals between the corrections (ie between the correction times) are determined taking into account a current driving situation of the vehicle.

[0015] Furthermore, in particular, a device for determining a vehicle pose of a vehicle is created, comprising acceleration sensors, yaw rate sensors, and a data processing device, wherein the data processing device is configured to estimate a vehicle pose by means of integration based on sensor data from at least the acceleration sensors and the yaw rate sensors and a known initial pose, to obtain at least one reference vehicle pose and to correct the estimated vehicle pose using the at least one reference vehicle pose, and to define time intervals between the corrections (ie between the correction times) taking into account a current driving situation of the vehicle.

[0016] The method and the device make it possible to specify intervals between corrections (i.e. between the correction times) at which the estimated vehicle pose is corrected using the at least one reference vehicle pose, taking into account the current driving situation. This allows, in particular, a variable selection of the time intervals so that corrections to a drift can be made depending on the (driving) situation. This ensures, in particular, that corrections are only made when necessary. Computing power can therefore be saved, so that a computing device can be made smaller and / or energy consumption can be reduced. Depending on the current driving situation, an error that continuously grows due to integration can grow to different degrees.Accordingly, the method and device can be used to react to the speed of the increasing error in a situation-dependent and thus situation-appropriate manner by making more or less frequent corrections. When considering the situations in isolation, an error generally grows more slowly when driving straight ahead than when cornering. Even when cornering with different radii of curvature, the rate at which the error grows generally varies. In particular, tire slip in the lateral direction varies with different curve radii. This leads to different errors in the yaw angle and, consequently, different errors in the calculation (e.g. the x and y components) of the vehicle position. Here, too, corrections can be made at intervals selected depending on the situation using the reference vehicle pose.In principle, it can also be provided that, depending on the situation, a constant time interval is selected between the correction times, at least until the next driving situation.

[0017] A vehicle pose includes, in particular, a vehicle position and a vehicle orientation. The vehicle position can, for example, include coordinates of a global coordinate system. The vehicle orientation includes at least a yaw angle, in particular with respect to a global coordinate system. The vehicle pose is estimated using sensor data from at least acceleration sensors and yaw rate sensors. In other words, the vehicle pose is estimated, in particular, using an inertial navigation system.

[0018] An initial pose refers, in particular, to the last known or estimated vehicle pose. The initial pose forms, in particular, a reference pose or a starting pose for the integration. Starting from the initial pose, the next vehicle pose is then determined or calculated using integration based on sensor data from at least the acceleration sensors and the yaw rate sensors. It may be provided that yaw rate sensors or yaw rate signals for the vehicle's tires are also taken into account when estimating the vehicle pose.

[0019] A reference vehicle pose (reference pose of the vehicle) refers, in particular, to a vehicle pose determined using different sensors and / or in a different way than the estimated vehicle pose. The reference vehicle pose is or will be determined in particular in such a way that its accuracy and / or reliability is greater than that of the estimated vehicle pose. In particular, it is provided that the reference vehicle pose is determined only for the correction times in order to minimize the computing power required to provide the reference vehicle pose.

[0020] A current driving situation refers, in particular, to a situation and / or scenario in which the vehicle is currently located, i.e., at the time the vehicle pose is estimated. The current driving situation can, in particular, include at least one manifestation of one of the following characteristics: a traffic scenario (e.g., driving straight ahead, cornering, parking, exiting a parking space, open sky, in a parking garage, etc.), vehicle parameters (speed, acceleration, etc.), and / or weather conditions (temperature, rain, slippery conditions, etc.).

[0021] The time intervals between corrections are determined taking the current driving situation into account. In particular, look-up tables can be used for this purpose. In these look-up tables the time intervals and / or calculation rules for calculating the times are stored for each driving situation so that these can be retrieved from the look-up table as needed. The time intervals and / or the times (or a corresponding calculation rule for calculating the times) for different driving situations can be determined, for example, with the help of empirical test series and / or simulations. In particular, it can be provided that the time intervals are selected or set in such a way that an error always remains below a predetermined limit. In particular, it is provided that the reference vehicle pose is then determined or has to be determined only for the correction times.

[0022] The method and the device are used in particular in a vehicle. The vehicle is in particular a motor vehicle. However, the vehicle can in principle also be another land, rail, water, air, or space vehicle, for example a drone or an air taxi. In particular, a vehicle is also provided comprising at least one device according to one of the described embodiments.

[0023] Parts of the device, in particular the data processing device, can be implemented individually or collectively as a combination of hardware and software, for example, as program code executed on a microcontroller or microprocessor. However, it can also be provided that parts are implemented individually or collectively as an application-specific integrated circuit (ASIC) and / or a field-programmable gate array (FPGA).

[0024] In one embodiment, it is provided that driving situations are differentiated at least between straight-ahead driving and / or cornering and / or uphill driving. This allows cases to be differentiated in which the rate of error growth differs. It can also be provided that further distinctions are made during cornering, for example, based on a radius of curvature. Furthermore, during uphill driving, a distinction can be made between positive or negative gradients and / or a change in the gradient.

[0025] In one embodiment, it is provided that the current driving situation is determined based on sensor data from at least the acceleration sensors and the yaw rate sensors. This allows the current driving situation to be determined in a particularly simple manner. For this purpose, the sensor data from at least the acceleration sensors and the yaw rate sensors are evaluated continuously by means of the data processing device. For example, patterns in the sensor data can be recognized, for example with the aid of machine learning or artificial intelligence methods. Alternatively or additionally, the sensor data, for example the acceleration values, can be compared with threshold values ​​in order to identify a current driving situation. In order to distinguish cornering from straight-ahead driving, lateral acceleration values, for example, can be compared with one or more predetermined threshold values.If the threshold is exceeded, it is assumed that the vehicle is cornering. If, however, the threshold is not exceeded, it is assumed that the vehicle is traveling straight ahead. With several graduated thresholds, the radius of curvature of a corner can be determined, taking into account the vehicle's yaw rate. The greater the detected lateral acceleration, the shorter the time intervals between corrections to the estimated vehicle pose can be, for example. The type of curve (right-hand curve or left-hand curve) can be identified from the sign of the lateral acceleration.

[0026] In a further embodiment, it is provided that a single degree of freedom of the acceleration sensors and / or the yaw rate sensors is monitored and evaluated to determine the current driving situation. This makes monitoring and evaluation particularly simple and requires only minimal computing power and memory. The degree of freedom can, for example, coincide with or include a lateral acceleration of the vehicle.

[0027] In an alternative, further development, it is provided that a plurality of degrees of freedom of the acceleration sensors and / or the yaw rate sensors are monitored and evaluated to determine the current driving situation. This allows different driving situations to be better differentiated from one another. In particular, patterns in the sensor data from the majority of degrees of freedom can be recognized more effectively, in particular with greater resolution.

[0028] In one embodiment, it is provided that the vehicle pose estimated based on the sensor data of at least the acceleration sensors and the yaw rate sensors is compared with a further vehicle pose estimated based on wheel revolutions and steering angles, wherein the estimated vehicle pose is corrected, in particular additionally, if a difference between the estimated vehicle pose and the estimated further vehicle pose exceeds a predetermined limit value. This creates a further, in particular additional, possibility of determining the time intervals for the corrections directly based on the sensor data. In particular, this embodiment can be implemented with little technical effort, since the aforementioned sensors are usually present in vehicles anyway.

[0029] In one embodiment, the reference vehicle pose is determined based on a weighted mean value, wherein the mean value is composed of at least the vehicle pose, which is estimated based on the sensor data from at least the acceleration sensors and the yaw rate sensors, and another vehicle pose, which is estimated based on wheel revolutions and steering angles. This can reduce the effort required to determine the reference vehicle pose, since the aforementioned sensors are usually already present in vehicles. A weighting factor can be determined and specified, for example, based on empirical test series and / or simulations.

[0030] In one embodiment, the reference vehicle pose is determined using acquired camera data from at least one surroundings camera of the vehicle and / or acquired radar data from at least one radar sensor of the vehicle and / or acquired lidar data from at least one lidar sensor of the vehicle. This allows the reference vehicle pose to be determined particularly reliably and precisely. Known methods for localization in surroundings are used to determine the reference vehicle pose. In particular, features in the surroundings can be recognized and used to localize the vehicle in a (surroundings) map.

[0031] In one embodiment, the current driving situation is determined based on a road map. This makes it possible, for example, to take past and / or future road courses into account. Based on the current vehicle pose, the vehicle is located on the road map and the current driving situation, for example a road course (straight ahead or cornering, etc.), is determined. Furthermore, other features of a current environment can be taken into account, such as negative or positive gradients and / or a change in the gradient, a road surface, etc. In particular, negative or positive gradients and / or changes in the gradient can be used to determine the current driving situation. In an exemplary scenario, the vehicle drives on a road with a slight positive or negative gradient (gradient ≈ 0), with pronounced gradients occurring at specific points on the road.These marked locations can then be used to determine the current driving situation, whereby the current driving situation (e.g. uphill, downhill, etc.) is stored in the road map.

[0032] In one embodiment, the reference vehicle pose is determined based on a road map, taking the current driving situation into account. This allows the reference vehicle pose to be determined, in particular, without additional sensor technology.

[0033] In particular, it is provided that the reference vehicle pose is determined based on a road course in the road map that corresponds to the current driving situation. For this purpose, the sensor data from at least the acceleration sensors and the yaw rate sensors are evaluated and compared with a road course in the road map. At specific positions or situations, a vehicle pose can then be easily estimated. For example, the end of a curve can be determined by a vanishing lateral acceleration. If this is the case, the vehicle position is set accordingly to the end of the curve in the road map. The vehicle orientation is set according to a direction of travel stored in the road map. The vehicle position and the vehicle orientation determined in this way then result in a vehicle pose that is used as the reference vehicle pose.In this embodiment, too, a negative or positive gradient can be used to determine the position, for example, if locations with a pronounced gradient are stored in the road map. The respective gradient can be determined by comparing it with the recorded sensor data from at least the acceleration sensors (especially those for vertical acceleration), and by comparing it with the road map, the position and orientation (vehicle pose) of the vehicle in the road map can be determined.

[0034] Accordingly, other situations can be found in which determining a reference vehicle pose in this way with the help of the road map is possible. Examples include S-curves, where the point in time of transition between the curve shapes can be determined by a change in the sign of the lateral acceleration, and thus the vehicle pose can be deduced from the road map. For example, in the case of a change in the sign of the lateral acceleration, the vehicle pose can be set such that a vehicle position in the (right) lane lies at the transition point between the two curves of the S-curve, and a vehicle orientation is set according to the direction of travel in the lane.Turning maneuvers at intersections can also be used as an example, since, similar to curves, the end of the turn can be determined very precisely and a vehicle pose can be estimated based on the road map. The respective intersection can be identified, in particular, by taking into account the intersection angle of the intersecting streets.

[0035] In one embodiment, the reference vehicle pose is determined based on a signal from a global navigation satellite system. This allows for particularly simple provision of the reference vehicle pose. In particular, most modern vehicles already have a navigation system that evaluates the signals from a global navigation satellite system (e.g., GPS), thus saving costs for additional equipment.

[0036] It may also be provided to estimate the reference vehicle pose by taking into account several methods, for example by means of sensor data fusion.

[0037] Further features of the device design will become apparent from the description of embodiments of the method. The advantages of the device are the same as those of the embodiments of the method.

[0038] The invention will be explained in more detail below using preferred embodiments with reference to the figures. Fig. 1 shows a schematic representation of an embodiment of the device for determining a vehicle pose of a vehicle; Fig. 2 shows a schematic flow diagram of an embodiment of the method for determining a vehicle pose of a vehicle.

[0039] In Fig. 1A schematic representation of an embodiment of the device 1 for determining a vehicle pose 20 of a vehicle 50 is shown. The device 1 comprises acceleration sensors 2, yaw rate sensors 3, and a data processing device 4. The data processing device 4 comprises a computing device 4-1, for example, a microprocessor, and a memory device 4-2. The acceleration sensors 2 and the yaw rate sensors 3 are, for example, part of an inertial navigation system. The method for determining a vehicle pose 20 of a vehicle 50 is explained below using the device 1.

[0040] The data processing device 4 is configured to estimate a vehicle pose 20 by means of integration based on sensor data 10 from the acceleration sensors 2 and the yaw rate sensors 3 and a known initial pose. Furthermore, the data processing device 4 obtains, in particular receives, at least one reference vehicle pose 21. The data processing device 4 corrects the estimated vehicle pose 20 using the at least one reference vehicle pose 21. In doing so, the data processing device 4 determines time intervals between the corrections, taking into account a current driving situation 11 of the vehicle 50.

[0041] The current driving situation 11 is determined, for example, based on the sensor data 10 of the acceleration sensors 2 and the yaw rate sensors 3 by means of the data processing device 4. Based on the determined current driving situation 11, the data processing device 4 can, for example, be configured to retrieve and use time intervals between the corrections in a lookup table (not shown) stored in the memory 4-2.

[0042] It can be provided that the reference vehicle pose 21 is determined using acquired camera data 12 from at least one surroundings camera 51 of the vehicle 50 and / or acquired radar data 13 from at least one radar sensor 52 of the vehicle 50 and / or acquired lidar data 14 from at least one lidar sensor 53 of the vehicle 50. The reference vehicle pose 21 can be determined, for example, with the aid of an environment detection system 54 configured accordingly for this purpose, using known localization methods. In this case, the environment detection system 54 can also be designed as part of the device 1. The determination of the reference vehicle pose 21 is carried out (only) when a correction of the estimated vehicle pose 20 is to be carried out. This makes it possible to keep the computing power required for evaluating the sensor data 12, 13, 14 low.

[0043] It can be provided that in order to determine the current driving situation 11, a single degree of freedom of the acceleration sensors 2 and / or the yaw rate sensors 3 is monitored and evaluated.

[0044] It can be provided that in order to determine the current driving situation 11, a plurality of degrees of freedom of the acceleration sensors 2 and / or the yaw rate sensors 3 are monitored and evaluated.

[0045] It can be provided that at least straight-ahead driving and / or cornering and / or uphill driving are distinguished as driving situations 11. This is done, for example, by using the data processing device 4 to compare a lateral acceleration mapped in the sensor data 10 with one (or more) predetermined threshold values, the exceedance of which indicates cornering. Furthermore, a rotation rate mapped in the sensor data 10 can be taken into account and evaluated. To detect uphill driving, in particular, an acceleration in the vertical direction of the vehicle 50 is also evaluated.

[0046] It can be provided that the vehicle pose 20 estimated on the basis of the sensor data 10 of at least the acceleration sensors 2 and the yaw rate sensors 3 is compared with a further vehicle pose 22 estimated on the basis of wheel revolutions 15 and steering angles 16, wherein the estimated vehicle pose 21 is corrected, in particular additionally, if a difference between the estimated vehicle pose 20 and the estimated further vehicle pose 22 exceeds a predetermined limit value.

[0047] It can be provided that the reference vehicle pose 21 is determined based on a weighted mean value, wherein the mean value is composed in a weighted manner of at least the vehicle pose 21, which is estimated based on the sensor data 10 of at least the acceleration sensors 2 and the yaw rate sensors 3, and a further vehicle pose 22, which is estimated based on wheel revolutions 15 and steering angles 16.

[0048] It can be provided that the current driving situation 11 is determined based on a road map 30. The road map 30 is stored, for example, in the memory 4-2. In particular, it can be provided that the reference vehicle pose 21 is determined based on the road map 30, taking into account the current driving situation 11.

[0049] In a further development, it can be provided that the reference vehicle pose 21 is determined based on a road course in the road map 30 corresponding to the current driving situation 11.

[0050] It can be provided that the reference vehicle pose 21 is determined based on a signal from a global navigation satellite system (not shown), for example a global positioning system (GPS).

[0051] The Fig. 2 shows a schematic flow diagram of an embodiment of the method for determining a vehicle pose of a vehicle.

[0052] In a measure 100, the vehicle pose is estimated by integration based on sensor data from at least acceleration sensors and yaw rate sensors and a known initial pose. The initial pose is, in particular, the last estimated vehicle pose. The value for the estimated vehicle pose is provided.

[0053] In a measure 101, a current driving situation is determined. This occurs, for example, based on sensor data from at least the acceleration sensors and the yaw rate sensors. It can be provided that a single degree of freedom of the acceleration sensors and / or the yaw rate sensors is monitored and evaluated to determine the current driving situation. Alternatively, it can be provided that a plurality of degrees of freedom of the acceleration sensors and / or the yaw rate sensors are monitored and evaluated to determine the current driving situation. In particular, it is provided that a distinction is made between at least straight-ahead driving and cornering as driving situations.

[0054] Alternatively or additionally, it may be provided that the current driving situation is determined based on a road map.

[0055] In a measure 102, time intervals between correction times are determined taking into account the specific current driving situation of the vehicle.

[0056] In step 103, a check is made to determine whether the time interval for a specified time interval has already elapsed. If this is not the case, the program jumps to step 100, with the initial pose now corresponding to the vehicle pose estimated in step 100. If this is the case, the program continues with step 104.

[0057] In step 104, a reference vehicle pose is obtained, in particular received. It can be provided that the reference vehicle pose is determined based on a weighted mean value, wherein the mean value is composed of a weighted average value composed of at least the vehicle pose, which is estimated based on the sensor data of at least the acceleration sensors and the yaw rate sensors, and another vehicle pose, which is estimated based on wheel revolutions and steering angles.

[0058] It can also be provided that the reference vehicle pose is determined by means of recorded camera data from at least one surrounding camera of the vehicle and / or recorded radar data from at least one radar sensor of the vehicle and / or recorded lidar data from at least one lidar sensor of the vehicle.

[0059] Furthermore, it can be provided that the reference vehicle pose is determined based on a road map, taking into account the current driving situation. In particular, it can be provided that the reference vehicle pose is determined based on a road course in the road map that corresponds to the current driving situation.

[0060] It may also be provided that the reference vehicle pose is determined based on a signal from a global navigation satellite system.

[0061] In a step 105, the estimated vehicle pose is corrected using at least one reference vehicle pose. The corrected estimated vehicle pose is provided.

[0062] Then, step 100 is continued, with the initial pose now corresponding to the corrected estimated vehicle pose.

[0063] Further embodiments of the method have already been described above with reference to the Fig. 1 described. List of reference symbols

[0064] 1Device 2Acceleration sensor(s) 3Yaw rate sensor(s) 4Computing device 4-1Computing device 4-2Storage device 10Sensor data 11Current driving situation 12Camera data 13Radar data 14Lidar data 15Wheel revolutions 16Steering angle 20Estimated vehicle pose 21Reference vehicle pose 22Further vehicle pose 30Road map 50Vehicle 51Environmental camera 52Radar sensor 53Lidar sensor 54Environmental detection system 100-105Measures of the method

Claims

1. Method for determining a vehicle pose (20) of a vehicle (50), wherein the vehicle pose (20) is estimated by means of integration based on sensor data (10) from at least acceleration sensors (2) and angular rate sensors (3), and a known initial pose, wherein the estimated vehicle pose (20) is corrected by means of at least one reference vehicle pose (21), wherein time intervals between the corrections are established taking into account a current driving situation (11) of the vehicle (50).

2. Method according to claim 1, characterized in that the current driving situation (11) is determined based on sensor data (10) from at least the acceleration sensors (2) and the angular rate sensors (3).

3. Method according to claim 2, characterized in that, to determine the current driving situation (11), an individual degree of freedom of the acceleration sensors (2) and / or the angular rate sensors (3) is monitored and evaluated.

4. Method according to claim 2, characterized in that, to determine the current driving situation (11), a plurality of degrees of freedom of the acceleration sensors (2) and / or the angular rate sensors (3) are monitored and evaluated.

5. Method according to any of the preceding claims, characterized in that at least driving straight-ahead and / or cornering and / or uphill driving can be distinguished as driving situations (11).

6. Method according to any of the preceding claims, characterized in that the vehicle pose (11) estimated on the basis of the sensor data (11) from at least the acceleration sensors (2) and the angular rate sensors (3) is compared with an additional vehicle pose (22) estimated on the basis of wheel revolutions (15) and steering angles (16), the estimated vehicle pose (20) being corrected if a difference between the estimated vehicle pose (20) and the estimated additional vehicle pose (22) exceeds a predetermined limit value.

7. Method according to any of the preceding claims, characterized in that the reference vehicle pose (21) is determined based on a weighted mean value, the mean value being composed, in a weighted manner, of at least the vehicle pose (20) which is estimated based on the sensor data (10) from at least the acceleration sensors (2) and the angular rate sensors (3), and an additional vehicle pose (22) which is estimated based on wheel revolutions (15) and steering angles (16).

8. Method according to any of the preceding claims, characterized in that the reference vehicle pose (21) is determined by means of recorded camera data (12) from at least one surrounding camera (51) of the vehicle (50) and / or recorded radar data (13) from at least one radar sensor (52) of the vehicle (50) and / or recorded lidar data (14) from at least one lidar sensor (53) of the vehicle (50).

9. Method according to any of the preceding claims, characterized in that the current driving situation (11) is determined based on a road map (30).

10. Device (1) for determining a vehicle pose (20) of a vehicle (50), comprising: acceleration sensors (2), angular rate sensors (3), and a data processing device (4), wherein the data processing device (4) is designed to estimate a vehicle pose (20) by means of integration based on sensor data (10) from at least the acceleration sensors (2) and the angular rate sensors (3), and a known initial pose, to obtain at least one reference vehicle pose (21) and to correct the estimated vehicle pose (20) by means of the at least one reference vehicle pose (21), and to establish time intervals between the corrections, taking into account a current driving situation (11) of the vehicle (50).