Fusion self-motion calculation for a vehicle

A device combining kinematic and environmental sensors with data fusion techniques addresses inaccuracies in vehicle position determination, enhancing precision through weighted averaging and Kalman filters.

DE102015202230B4Active Publication Date: 2024-11-28AUMOVIO AUTONOMOUS MOBILITY GERMANY GMBH
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
DE102015202230
Authority / Receiving Office
DE · DE
Patent Type
Patents
Current Assignee / Owner
Filing Date
2015-02-09
Publication Date
2024-11-28
Estimated Expiration
2035-02-09

AI Technical Summary

Technical Problem

Existing vehicle driver assistance systems face challenges in accurately determining the relative change in vehicle position between two time points due to inconsistencies and inaccuracies in sensor data, particularly from wheel speed and wheel angle sensors during slipping and cornering.

Method used

A device utilizing a first sensor unit for kinematic data and a second sensor unit for environmental detection, combined with data fusion techniques like weighted averaging and Kalman filters, to enhance the accuracy of vehicle position determination by merging self-motion data with environmental data.

Benefits of technology

The solution provides a significantly more reliable and precise detection of the vehicle's own movement by integrating multiple sensors and data fusion algorithms, reducing errors from sensor inconsistencies and improving positional accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

Device (8) for determining the self-propulsion of a vehicle (9), comprising the device (8): a first sensor unit (21) for determining kinematic data of the vehicle (9) at a first time point in time and a second point in time and a second sensor unit (22) which is designed to detect the environment of the vehicle (9) and to determine a relative movement of the vehicle (9) to the environment, further comprising: a computing unit (30) which is executed to determine a change in position of the vehicle (9) between the first and the second time point based on the kinematic data of the first sensor unit (21) and the relative motion determined by the second sensor unit (22), wherein the kinematic data of the first sensor unit (21) and the relative motion determined by the second sensor unit (22) are calculated together by means of a weighted averaging to determine the change in position of the vehicle (9), wherein the kinematic data of the first sensor unit (21) and the relative motion determined by the second sensor unit (22) are fed to a Kalman filter to determine the change in position of the vehicle (9), and wherein the computing unit (30) is designed to use the self-motion data together with data from a radar system for environmental mapping.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] The invention relates to a device for determining the self-propulsion of a vehicle. Background of the invention

[0002] Some driver assistance systems in vehicles require a very precise determination of the vehicle's own motion. For example, for environmental mapping systems, the vehicle's movement may be necessary as additional information. In such systems, which emit electromagnetic waves in discrete time steps and receive their reflections, the relative change in position during the period between two time steps can be of interest.

[0003] DE 10 2010 048 896 A1 and US 2012 / 0169532 A1 concern radar systems in vehicles for environmental detection using the vehicle's own movement, where the movement is determined, for example, from acceleration, yaw rate, steering angle, and wheel speed sensors.

[0004] DE 10 2010 063 133 A1 relates to a method for determining the self-movement of a vehicle using a vehicle camera and based on the recorded motion flow.

[0005] DE 10 2005 000 651 A1 relates to a method for determining the self-motion of a vehicle using an imaging device and based on optical flows.

[0006] DE 10 2010 015 723 A1 relates to a method for determining the self-motion of a vehicle using a receiving unit, in particular a radar unit and based on the determination of an angular function.

[0007] DE 102 51 949 A1 discloses a method for vehicle dynamics control in a motor vehicle, wherein a sensor records a measured value, wherein an actuator for vehicle dynamics control is controlled depending on the measured value, wherein an image sensor system generates image information from the vehicle environment for vehicle dynamics control, wherein at least two image sensors are provided which record essentially the same scene.

[0008] DE 10 2012 219 834 A1 describes a lane tracking system for tracking the position of a vehicle in a lane, comprising a camera designed to provide a video signal representing a field of view, and a video processor designed to receive the video signal from the camera and generate latent video-based position data indicating the position of the vehicle in the lane.

[0009] DE 10 2011 055 795 A1 relates to a method for detecting an impending vehicle rollover, comprising the following steps: Step a) Recording a sequence of images of the vehicle's surroundings, in particular of the road ahead, with a camera (1a; 1b). Step b) Determining and tracking point features in the image sequence. Step c) Eliminating point features located on objects moving relative to the stationary vehicle surroundings. Step d) Calculating at least one parameter of the rotational camera movement (roll, pitch, and / or yaw rate) taking into account the remaining point features.

[0010] Furthermore, DE 10 2012 100 164 A1 discloses a vehicle lateral control system comprising a lane marker module configured to determine a direction of travel and deviation of a vehicle in response to images received from a second sensor device, a lane information fusion module configured to generate vehicle and lane information in response to data received from heterogeneous vehicle sensors, and a lane control configured to generate a collision-free vehicle path in response to the vehicle and lane information from the lane information fusion module and an object map. Summary of the invention

[0011] The object of the invention can be considered to be to improve the accuracy of determining the relative change in position of a vehicle between two points in time.

[0012] This problem is solved by the subject matter of the independent claim. Further developments of the invention are described in the dependent claims and in the following description.

[0013] According to one aspect of the invention, a device for determining the self-motion of a vehicle is provided, the device comprising a first sensor unit for determining kinematic data of the vehicle at a first time and a second time, and a second sensor unit configured to detect the vehicle's environment and determine a relative motion of the vehicle to the environment, further comprising a computing unit configured to determine a change in the vehicle's position between the first and second time based on the kinematic data of the first sensor unit and the relative motion determined by the second sensor unit.

[0014] In other words, a variety of sensors are used to determine the relative change in position between two time steps. These sensors consist of a first sensor unit for determining kinematic data and a second sensor unit for observing the environment.

[0015] Measurements can also be taken at further time steps. Typically, measurements are performed at a large number of time points, so the first and second time points described are representative of a large number of time steps.

[0016] Kinematic data is information about motion, especially information about the motion of a vehicle. Information about motion can relate to the position of the vehicle itself, but also to changes in position, especially changes in position over time, i.e., speed, or the acceleration of the vehicle.

[0017] The second sensor unit observes and records the vehicle's immediate surroundings as they move, in order to infer the vehicle's own motion relative to its environment. This data on the vehicle's own motion is then fused with that from the other sensors, i.e., the first sensor unit.

[0018] Not all sensors consistently deliver accurate readings. One potential source of error is wheel slippage, which can cause wheel speed sensors to produce unusable values ​​even when measuring correctly. Lateral wheel slippage can also lead to inaccurate readings from wheel angle sensors. Furthermore, the difference in rotational speed during cornering can distort the results. Therefore, a verification algorithm can be employed during sensor data fusion to check the measured motion data for consistency, plausibility, and statistical parameters such as the standard deviation to identify measurement noise.

[0019] The data fusion itself can be performed, for example, using weighted averaging. A Kalman filter can also be used.

[0020] The computing unit is specifically designed to determine a change in the vehicle's position based on the kinematic data of the first sensor unit and on the data of the second sensor unit.

[0021] Beyond a certain number of measurement data points from different time points, the measurement data can be dynamically filtered. For example, complementary filtering can be used depending on the signal-to-noise ratio of individual sensors. In complementary filtering, the signals from one sensor are filtered by a low-pass filter, and the signals from another sensor, which provides essentially the same information as the first sensor, are filtered by a high-pass filter. The filtered signals are then combined.

[0022] Two sensors that provide essentially the same information include, for example, a Doppler radar and a wheel speed sensor, when both are used for speed measurement.

[0023] For example, only the frequencies passed through a low-pass filter can be used from the typically very noisy acceleration measurements. The signals from a Doppler radar, for instance, are suitable for utilizing their high frequencies.

[0024] The terms "high frequencies" and "low frequencies" can be freely defined. Typically, a bandwidth is determined for the low-pass filter, and the bandwidth of the high-pass filter is adjusted accordingly so that both together provide a complete spectrum in the frequency range of interest.

[0025] An advantageous effect of the invention may be based on the fact that the data fusion of the data from the sensors of the first sensor unit with those of the second, environment-monitoring sensor unit enables a significantly more reliable and accurate detection of the vehicle's own movement.

[0026] According to another embodiment, the second sensor unit is an optical sensor unit. An optical sensor unit receives, in particular, the electromagnetic waves reflected in the vicinity of the vehicle, for example, visible light.

[0027] According to another embodiment, the second sensor unit is a Doppler radar. A Doppler radar determines a speed from the frequency shift of the waves when these waves are emitted by a transmitter moving relative to the reflector.

[0028] According to another embodiment, the kinematic data from the first sensor unit and the relative motion determined by the second sensor unit are combined using a weighted averaging process to determine the vehicle's change in position. For example, the kinematic data from the first sensor unit can be weighted differently in the averaging process with the relative motion calculated by the second sensor unit, thus accounting for the accuracy or inaccuracy of a sensor unit. The weighting can vary over time, for example, if the accuracy of a sensor unit fluctuates. Depending on the data type of the kinematic data from the first sensor unit, a relative motion must first be calculated from this data.If, for example, the first sensor unit provides the speed and steering angle, the relative motion calculated from this data is fused with the relative motion calculated by the second sensor unit. Therefore, the processing unit can use either the raw data from the first sensor unit or the relative motion determined by the first sensor unit to calculate the relative change in the vehicle's position between a first and a second point in time.

[0029] According to another embodiment, the kinematic data of the first sensor unit and the relative motion determined by the second sensor unit are fed to a Kalman filter to determine the change in position of the vehicle.

[0030] According to one aspect of the invention, the determination of self-motion using inertial sensors, speed and steering angle sensors is supplemented by environmental sensing.

[0031] According to one embodiment of the invention, the first sensor unit comprises at least one of the following: wheel angle sensor, wheel speed sensor, direction sensor unit, satellite-based positioning system, translational acceleration sensor, rotational acceleration sensor, yaw rate measuring device, or attitude gyroscope sensor. The first sensor unit differs from the second sensor unit in that the first does not detect the vehicle's surroundings.

[0032] According to a further embodiment of the invention, the second optical sensor unit is a camera unit. This can, for example, be a special camera that captures images of the environment in as deterministic, equidistant time steps as possible, so that the time of the recording can be precisely synchronized with the processing unit.

[0033] The operation of a camera unit is as follows: First, an image of the surroundings is captured. Second, another image of the surroundings is captured. Using image processing algorithms and the known time interval between the captures, the relative change in position is determined. This relative change in position can be expressed, for example, as translation and rotation in a local Cartesian coordinate system. For two-dimensional movement, these transformations can be expressed, for example, by a two-dimensional displacement vector and a rotation angle.

[0034] According to a further embodiment of the invention, the second optical sensor unit comprises two cameras. These can, for example, be stereo cameras. Multiple cameras have the advantage of being able to view the environment from several angles and thus capture the environment in its spatial structure.

[0035] According to a further embodiment of the invention, the processing unit is configured to use the vehicle's own motion data together with data from a radar system for mapping the vehicle's surroundings. In other words, the vehicle's own motion detection, combined with data fusion from inertial sensors, wheel speed sensors, steering angle sensors, or other sensors related to the vehicle itself, provides an interface with the sensors monitoring the surroundings. This interface allows the acquired data on the vehicle's own motion to be used together with a radar system for mapping the surroundings. Environmental mapping specifically describes the creation of a model of the environment, providing data on the positional relationships of objects to each other and to the vehicle.

[0036] According to a further embodiment of the invention, a device as described is provided which is configured to perform a method comprising the steps of: capturing vehicle motion information in a first sensor unit, capturing environmental information and determining a relative movement of the vehicle to the environment in a second sensor unit, fusing the data from the first and the second sensor unit and determining a change in position that the vehicle has approximately experienced.

[0037] Furthermore, a vehicle is provided, having the device as described above. Brief description of the characters Fig. Figure 1 schematically shows a method for determining the self-propulsion of a vehicle. Fig. Figure 2 schematically shows a device for determining the self-propulsion of a vehicle. Fig. Figure 3 schematically shows an example of a possible data fusion. Detailed description of implementation examples

[0038] Fig. Figure 1 shows a procedure comprising the following steps: - Capture S1 of vehicle movement information in a first sensor unit, - Capture S2 environmental information and determine the relative movement of the vehicle to the environment in a second sensor unit, - Fusing S3 of the data from the first and second sensor units, - Determine S4 of a change in position that the vehicle has approximately experienced.

[0039] Fig. Figure 2 shows the device 8 for determining the self-propulsion of a vehicle 9. This device 8 has the following features: A first sensor unit 21, a second sensor unit 22, wherein the second sensor unit 22 is an environment-monitoring, for example optical, sensor unit configured to detect the environment and, for example, to determine the relative movement of the vehicle with respect to the environment via the optical flux of the environment. Furthermore, this device 8 has a processing unit 30 configured to combine the data from the first sensor unit 21 and the second sensor unit 22 in data fusion and to determine a relative change in position of the vehicle 9 using the fused data.

[0040] Fig.Figure 3 shows an embodiment of a possible data fusion of the steering angle sensor 211 with a wheel speed sensor 212 from the first sensor unit 21 with the data from a second sensor unit 22, for example, a camera unit. The data fusion is performed by a processing unit 30. In the embodiment shown, both the first 21 and the second 22 sensor units provide data on the speed and the direction of travel or change of direction of travel of the vehicle 9. In this example, there is initially redundant information about the relative change in position of the vehicle 9 over a certain time. In this embodiment, the data fusion can be performed, for example, by means of weighted averaging. The data from the wheel speed sensor 212 and the steering angle sensor 211 can also be discarded if another device of a driver assistance system detects wheel slip. Reference symbol list: 8 Device 9 vehicles 21 First sensor unit 22 Second sensor unit 211 Steering angle sensor 212 Wheel speed sensor 30 computing units S1 S2 S3 Merge S4 Determine

Claims

[1] Device (8) for determining the movement of a vehicle (9), the device (8) comprising: a first sensor unit (21) for determining kinematic data of the vehicle (9) at a first time and a second point in time and a second sensor unit (22) which is designed to detect an environment of the vehicle (9) and to determine a relative movement of the vehicle (9) to the environment, further comprising: a computing unit (30) designed to determine a change in position of the vehicle (9) between the first and the second time point based on the kinematic data of the first sensor unit (21) and the relative movement determined by the second sensor unit (22), wherein the kinematic data of the first sensor unit (21) and the relative movement determined by the second sensor unit (22) are offset against one another via a weighted averaging in order to determine the change in position of the vehicle (9), wherein the kinematic data of the first sensor unit (21) and the relative movement determined by the second sensor unit (22) are fed to a Kalman filter in order to determine the change in position of the vehicle (9), and wherein the computing unit (30) is designed to use the self-motion data together with data from a radar system for environmental mapping. [2] The device according to claim 1, wherein the second sensor unit (22) is an optical sensor unit. [3] Apparatus according to claim 1, wherein the second sensor unit (22) is a Doppler radar. [4] Device according to one of the preceding claims, wherein the first sensor unit (21) comprises at least one element from the group consisting of steering angle sensor (211), wheel speed sensor (212), direction sensor unit, satellite-based positioning system, sensor for translatory accelerations, sensor for rotational accelerations, yaw rate measuring device, attitude gyro sensor. [5] Device (8) according to claim 2, wherein the second sensor unit (22) is a camera unit. [6] Device (8) according to claim 5, wherein the camera unit comprises two cameras. [7] Vehicle (9) comprising the device (8) according to one of claims 1 to 6.

Citation Information

Patent Citations

  • Method for determining the own motion of a vehicle

    DE102005000651A1

  • Motor vehicle movement detection method for driver assistance for e.g. navigation function, involves determining trigonometric function by compensation calculation of values of relative movement

    DE102010015723A1

  • Device for detecting surrounding of vehicle i.e. motor car, using radar system, has echo profile generating measured values that are assigned to frequency stages and including surrounding information for synthetic aperture by points

    DE102010048896A1

  • Method and system for determining the self-motion of a vehicle

    DE102010063133A1

  • Procedure for determining an imminent vehicle rollover

    DE102011055795A1