Method for determining position data with a GNSS system

Abstract

The invention relates to a method for determining navigation data using a GNSS system, wherein a measurement noise of a slip angle (108) of a vehicle (100) is modeled in the method and the method comprises at least the following steps: a) detecting a wheel speed (104), a yaw rate and a steering angle (102) of the vehicle (100); b) if a current speed and / or a current yaw rate of the vehicle (100) falls below a predetermined threshold value, modeling the measurement noise of the slip angle (108) on the basis of the detected wheel speed (104); c) if the current speed and / or a current yaw rate of the vehicle reaches or exceeds the predetermined threshold value, modeling the measurement noise of the slip angle (108) only on the basis of a slip angle (108) determined from the detected yaw rate and the detected steering angle (102); and d) determining the navigation data taking into account the measurement noise modeled in step b) or step c).

Description

[0001] R. 414958

[0002] - 1 -

[0003] Description

[0004] title

[0005] Methods for determining position data using a GNSS system

[0006] State of the art

[0007] The invention relates to a novel method for determining position data using a GNSS system.

[0008] The Kalman filter (also Kalman-Bucy filter, Stratonovich-Kalman-Bucy filter or Kalman-Bucy-Stratonovich filter) is a mathematical filter model for the iterative estimation of system states based on faulty input data, which are in particular sensor data.

[0009] The Kalman filter is used to estimate system quantities that cannot be measured directly, while optimally reducing errors in the observations. Based on the input data, the Kalman filter typically maintains an internal mathematical model as a constraint, which is incorporated into the parameter estimation and takes into account dynamic relationships between the system quantities. This mathematical model, for example, represents equations of motion that describe the relationship between changing positions and velocities.

[0010] Simulating accelerations. Incorrect input data relating to positions, velocities, and / or accelerations, fed into the Kalman filter, can be correlated within the filter. This allows for the creation of precise estimates based on (incorrect) input data for positions and velocities.

[0011] Kalman filters are particularly useful for the iterative estimation of system states based on observations that are typically subject to error. In this context, Kalman filters have proven especially advantageous for applications where sensor information from different sensors needs to be combined (or fused) with model information. Furthermore, R. 414958

[0012] - 2 -

[0013] Kalman filters are frequently used in embedded systems because their calculations are advantageous, accurate, and robust. Furthermore, microcontrollers can perform Kalman filter calculations with advantageous efficiency.

[0014] Kalman filters are primarily used to fuse data from various sensors that determine vehicle positions, thereby generating highly precise navigation data (especially data concerning position and speed). Navigation data here refers specifically to position data, as well as data concerning speed and acceleration. Sensors whose data can be processed or fused using Kalman filters include, for example, GNSS sensors for determining navigation data (or localization) via GNSS satellites, inertial sensors, and, for instance, wheel sensors and steering angle sensors used in motor vehicles to monitor the vehicle's movement via the chassis. Such sensors thus generate a kind of precursor data for navigation data, which is used as input data in a Kalman filter to determine highly precise navigation data.By jointly considering this described data in a Kalman filter using the system models, parameters and / or system states stored in the Kalman filter, it becomes possible to generate highly precise navigation data from the described input data.

[0015] Kalman filters are used in particular to obtain highly precise navigation data for highly automated driving functions of motor vehicles and especially for autonomous driving functions.

[0016] In a Kalman filter, a continuous correction of internal parameters and system states is typically performed. The term "corrected parameters" is used here to refer to all corrections taking place within the Kalman filter. The quality of the internal parameters and / or system states in the Kalman filter is crucial for improving navigation data through the use of the Kalman filter.

[0017] However, regarding the precision and confidence of navigation data obtained with Kalman filters, there is a fundamental R. 414958

[0018] - 3 -

[0019] There is a need for improvement because, in principle, very high precision and high confidence in the navigation data used are desirable for applications of highly automated and autonomous driving.

[0020] GNSS systems, in particular, use measurements of speed and steering angle to decompose the speed into lateral and longitudinal components. At high speeds (e.g., highway driving), noise from the steering angle is therefore crucial for the accuracy of lateral speed observation.

[0021] Consequently, the vehicle's lateral speed cannot be estimated with sufficient accuracy. This inaccuracy also affects the sideslip angle. The sideslip angle, which is the ratio of lateral to longitudinal speed, is particularly relevant for applications such as autonomous lane changes or safety-critical scenarios, and its low accuracy represents a weakness in common GNSS / INS systems.

[0022] Disclosure of the invention

[0023] Based on this, a particularly advantageous method for determining navigation data with a GNSS system will be described, in which an improved modeling of the swimming angle is carried out.

[0024] The invention relates to a method for determining navigation data in a vehicle with a GNSS system, wherein the method models measurement noise of a vehicle's sideslip angle and the method comprises at least the following steps: a) acquiring a wheel speed, yaw rate, and steering angle of the vehicle; b) when a current vehicle speed and / or current vehicle yaw rate falls below a predetermined threshold, modeling the measurement noise of the sideslip angle based on the acquired wheel speed; c) when the current vehicle speed and / or current vehicle yaw rate reaches or exceeds the predetermined threshold, R. 414958

[0025] - 4 -

[0026] Modeling the measurement noise of the sideslip angle solely based on a sideslip angle determined from the recorded yaw rate and steering angle; and d) determining the navigation data taking into account the measurement noise modeled in step b) or step c).

[0027] For the purposes of this application, the term "navigation data" can be understood to mean geographical positions and / or speeds and / or accelerations of the vehicle that can be determined using the method according to the invention.

[0028] The measured wheel speed is preferably the wheel speed of the vehicle's front wheels. Furthermore, and preferably, the measured wheel speed of both front wheels is averaged.

[0029] The fundamental approach of the method described here is to improve the determination of navigation data by enhancing the accuracy of a swim angle measurement. This improvement in swim angle accuracy is based on an improved modeling of the swim angle measurement noise. This improved modeling of the measurement noise is achieved in particular through the modeling performed in steps b) and c), which is specifically dependent on the predetermined threshold value. This hybrid system makes it possible to optimize the measurement noise for different scenarios and thus incorporate it into the determination of navigation data.

[0030] A key difference between modeling the measurement noise of the sideslip angle according to step b) and step c) is that in step c), the wheel speed is not used for the modeling. This is based on the idea that the additional inclusion of the wheel speed at high speeds can lead to non-negligible errors in the modeling. For example, when the vehicle is traveling at 100 km / h with a steering angle of 0 degrees, a steering angle error of 0.5 degrees can lead to a deviation in the determination of the lateral speed of 0.9 km / h. Thus, a combined use of R. 414958

[0031] - 5 - of steering angle and speed, especially at high speeds, is disadvantageous and is avoided by the method according to the invention.

[0032] According to one embodiment, the wheel speed and / or the yaw rate and / or the steering angle of the vehicle are detected in step a) by means of sensors, in particular by means of a sensor unit. As already mentioned, the detection of the aforementioned quantities preferably takes place at the front axle of the vehicle. Since this axle is usually the steered axle of the vehicle, a precise determination of the aforementioned quantities, in particular the steering angle, can thus be made. The sensors or the sensor unit can be, or in particular can be, arranged in the vehicle.

[0033] Furthermore, it can be advantageous if, in step b), the recorded wheel speed is divided into longitudinal and lateral components using the steering angle. This division allows for a more precise inclusion of the wheel speed and / or the steering angle in the modeling of the measurement noise of the sideslip angle.

[0034] According to another embodiment, in the case of modeling the measurement noise according to step b), an extended Kalman filter is used to determine the navigation data in step d). Since the modeling in step b) is a non-linear model without ratio calculation, an extended Kalman filter has proven to be particularly suitable.

[0035] In one embodiment, the detected steering angle is corrected by a slip angle in step c). This improves the modeling of the measurement noise of the slip angle.

[0036] Furthermore, it can be advantageous to use an unscented Kalman filter for determining the navigation data in step d) when modeling the measurement noise according to step c). Since the modeling according to step c), unlike the previously explained modeling according to step b), involves a highly nonlinear observation equation due to a transformation and ratio of the velocities, the unscented Kalman filter has proven suitable. R. 414958

[0037] - 6 -

[0038] In particular, a hybrid Kalman filter is used for the execution of the method. For the purposes of this application, the term "hybrid Kalman filter" can be understood to mean a Kalman filter that includes both an extended and an unscented Kalman filter.

[0039] To switch between the two filters and thus between modeling the measurement noise of the swimming angle according to step b) or step c), the predetermined threshold is stored in an observation switch, which switches between step b) and c) depending on whether the threshold is undershot, reached or exceeded.

[0040] Thus, depending on the driving situation, especially depending on the speed and / or yaw rate of the vehicle, the modeling of the measurement noise of the sideslip angle can be determined by the observation switch in order to determine it in the best possible way for the respective driving situation and thus generally for the determination of the navigation data.

[0041] According to one embodiment, the predetermined threshold for the current yaw rate of the vehicle has a value of less than 10%, in particular less than 5% and specifically less than 3%.

[0042] Furthermore, according to an alternative or supplementary embodiment, the predetermined threshold for the current speed of the vehicle has a value of less than 50m / s, in particular less than 30m / s and specifically less than 22m / s.

[0043] The aforementioned values ​​have proven particularly suitable for thresholds. These thresholds are stored, for example, in an editable format on a data storage device and can be used from there for the process.

[0044] Furthermore, a computer program product, comprising instructions that, when executed by a computer, cause the computer to execute the described procedure, is to be described. R. 414958

[0045] - 7 -

[0046] The computer program product is preferably installed on the described control unit.

[0047] The following will be described in more detail: a computer-readable storage medium comprising instructions which, when executed by a computer, cause it to execute the described procedure or the steps of the described procedure.

[0048] The invention and its technical context are explained in more detail below with reference to the figures. The figures show preferred embodiments, to which the invention is not limited. It should be noted in particular that the figures, and especially the size relationships shown in the figures, are only schematic. They show:

[0049] Fig. 1 : a schematically represented block diagram of the method according to the invention;

[0050] Fig. 2: a diagram comparing the longitudinal and lateral velocity errors measured by different methods; and

[0051] Fig. 3: a diagram showing a comparison between the measurement noise of the float angle according to the prior art and the method according to the invention.

[0052] In the figures, identical or equivalent components are always marked with the same reference symbols.

[0053] In Fig. 1, the process according to the invention is schematically illustrated in its sequence by means of a block diagram.

[0054] In the first step a), a wheel speed 104a, 104b, a yaw rate, and a steering angle 102 of a vehicle 100 are determined. The vehicle 100 is only schematically represented as a rectangle in Fig. 1, which is intended to illustrate that the method is preferably used in such a vehicle 100. The aforementioned parameters are preferably recorded at the front axle of the vehicle 100. The wheel speed 104a, 104b is recorded for both the left and the right front wheel. R. 414958

[0055] - 8 -

[0056] Thus, a left wheel speed of 104a and a right wheel speed of 104b are obtained, from which an average is formed in an intermediate step to obtain a (mean) wheel speed of 104.

[0057] The next two steps b) and c) are each carried out as alternatives to each other in an observation switch 106. Here, the quantities recorded in step a) are compared with a predefined threshold value for the wheel speed 104 and / or a yaw rate.

[0058] If the detected value of the wheel speed 104 and / or the yaw rate is below the specified threshold, step b) is performed, which (viewed in the image plane) is represented in the right path of the observation switch 106.

[0059] In step b), the measurement noise of the sideslip angle 108 is modeled based on the recorded wheel speed 104. This step is therefore also referred to as "front velocity observation".

[0060] If the detected value of the wheel speed 104 and / or the yaw rate reaches or even exceeds the predetermined threshold, step c) of the method is carried out according to the invention. Step c) is shown (viewed in the plane of the image) in the left path of the observation switch 106.

[0061] According to step c), the modeling of the measurement noise of the side slip angle is based solely on a side slip angle 108 determined from the recorded yaw rate and steering angle. This step is therefore also referred to as "side slip angle observation". In other words, the wheel speed 104 is not used for the modeling in step c).

[0062] In the final step d) of the procedure, the navigation data is determined taking into account the measurement noise modeled in step b) or step c). In the case of step b), the measurement noise of the swimming angle 108 is modeled, and consequently, the navigation data is determined using an extended Kalman filter 110. R. 414958

[0063] - 9 -

[0064] In step c), the measurement noise of the swimming angle 108 is modeled, and consequently the navigation data is determined, using an unscented Kalman filter 112. This is because the modeling of the measurement noise according to step c) involves a highly nonlinear observation equation. Unscented Kalman filters 112 have proven to be particularly suitable and, in particular, more accurate for solving such equations.

[0065] The Extended Kalman Filter 110 and the Unscented Kalman Filter 112 are preferably part of a hybrid Kalman Filter 114, which is symbolized by a rectangle in Fig. 1.

[0066] Fig. 2 shows a comparison of both the longitudinal velocity error and the lateral velocity error at different speeds.

[0067] In this case, the two errors were recorded solely based on wheel speed. This is illustrated graphically by curve I.

[0068] Furthermore, this diagram also shows the error curves where the modeling was based solely on the sideslip angle determined from the recorded yaw rate and steering angle. This is illustrated graphically by curve II.

[0069] The third curve, III, is shown as modeling using the observation switch 106 and thus according to the method of the invention. As can be seen from Fig. 2, curve III exhibits a profile that shows the lowest errors over the entire time band.

[0070] Figure 3 shows a diagram that compares the measurement noise of the float angle according to the prior art and the method according to the invention.

[0071] Curve I represents the course of the measurement noise according to the prior art. Curve II represents the course of the measurement noise according to the method according to the invention. It is also clearly visible here, R. 414958

[0072] - 10 - that curve II has a lower amplitude, which means less measurement noise.

Claims

R. 414958 - 11 - Patent claims 1. A method for determining navigation data in a vehicle (100) with a GNSS system, wherein the method models measurement noise of a sideslip angle (108) of the vehicle (100) and the method comprises at least the following steps: a) Acquiring a wheel speed (104), a yaw rate, and a steering angle (102) of the vehicle (100); b) If a current speed of the vehicle (100) and / or a current yaw rate of the vehicle (100) falls below a predetermined threshold, modeling the measurement noise of the sideslip angle (108) based on the acquired wheel speed (104); c) If the current speed and / or a current yaw rate of the vehicle (100) reaches or exceeds the predetermined threshold, modeling the measurement noise of the sideslip angle (108) solely based on a sideslip angle (108) determined from the acquired yaw rate and the acquired steering angle (102);and d) Determining the navigation data taking into account the measurement noise modeled in step b) or step c).; 2. Method according to claim 1, wherein the wheel speed (104) and / or the yaw rate and / or the steering angle (102) of the vehicle (100) are detected in step a) by means of sensors, in particular by means of a sensor unit.

3. Method according to claim 1 or 2, wherein in step b) the detected wheel speed (104) is divided into a longitudinal and lateral component by means of the steering angle (102).

4. Method according to claim 3, wherein in the case of modeling the measurement noise according to step b) an Extended Kalman Filter (110) is used for determining the navigation data in step d).

5. Method according to claim 1 or 2, wherein in step c) the detected steering angle (102) is corrected by a slip angle. R. 414958 - 12 - 6. Method according to claim 5, in the case of modeling the measurement noise according to step c) for determining the navigation data in step d) an unscented Kalman filter (11 ) is used.

7. Method according to one of the preceding claims, wherein the predetermined threshold value is stored in an observation switch (106) which switches between step b) and c) depending on whether the threshold value is undershot, reached or exceeded for modeling the measurement noise of the float angle (108).

8. Method according to one of the preceding claims, wherein the predetermined threshold for the current yaw rate of the vehicle (100) has a value of less than 10° / s, in particular less than 5% and especially less than 3% 9. Method according to one of the preceding claims, wherein the predetermined threshold for the current speed of the vehicle (100) has a value of less than 50 m / s, in particular less than 30 m / s and especially less than 22 m / s 10. Control unit comprising a processor adapted / configured to perform the method according to any one of claims 1 to 9.

11. Computer program product comprising instructions which, when the computer program product is executed by a computer, cause the computer to execute the method according to any one of claims 1 to 9.

12. Computer-readable storage medium comprising instructions which, when executed by a computer, cause it to perform the steps of the method according to any one of claims 1 to 9.

Images