Method for generating feature-based localization map for GNSS and / or feature-based localization

Abstract

To provide a method for generating a feature-based localization map for GNSS and / or feature-based localization.SOLUTION: A method for generating a feature-based localization map includes at least following steps i.e., the steps of: i) generating the feature information for a feature-based localization map by using at least one GNSS information; ii) generating the GNSS-related meta information in which a GNSS situation, which is a basis for generating the feature information, can be reversely inferred; and iii) associating the GNSS-related meta information generated in step ii) with the feature information generated in the step i).SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a method for generating a feature-based localization map for GNSS- and / or feature-based localization, and a method for GNSS- and / or feature-based localization. Furthermore, a computer program, a machine-readable storage medium, and a localization device for a vehicle are described. The present invention is particularly usable in GNSS-based localization systems for autonomous or semi-autonomous driving. [Background technology]

[0002] Prior art Feature-based localization maps, or so-called mapping techniques, are widely used to generate environmental maps from sensor observations of individual vehicles or large fleets. They are used, for example, in the fields of robotics, logistics, automotive, aviation, aerospace, and consumer goods. Odometry, environmental sensor measurements, and GNSS measurements are often used to create feature-based localization maps.

[0003] Global Navigation Satellite Systems (GNSS) refer to a constellation of satellites that provide signals from space that transmit position and time data to a GNSS receiver. The receiver then uses these data to determine its current location. GNSS systems are widespread in many application areas, such as the automotive industry, surveying, avionics, or agriculture, and provide a low-cost, continuous, and global solution for positioning. To meet the high requirements for accuracy, availability, and completeness in automated and assisted driving applications, multi-GNSS (e.g., GPS, Galileo, Glonass, or Beidou) i.e., multi-frequency signals are often processed and GNSS correction services are considered to correct signal errors in space and to calculate accurate and secure PVAT (position, velocity, attitude, time) solutions.

[0004] Feature-based localization systems can determine the current location of a surround-sensing-enabled (sensors are, for example, video, radar, or LiDAR) mobile device by decoding current location coordinates from observable markers. In such systems, markers (e.g., landmarks such as traffic signs, poles, and / or lane markers) are placed at specific locations in an area. By measuring the viewing angle from the device to the marker, the device can estimate its own current location coordinates relative to the marker. Summary of the Invention [Problem to be solved by the invention]

[0005] Fusion of vehicle pose estimates from multiple different vehicle localization modalities is considered to be prior art. However, prior art techniques do not adequately consider the independence of the localization paths used. In particular, the use of GNSS for feature map generation can result in a loss of independence between feature-based and GNSS-based localization. [Means for solving the problem]

[0006] Disclosure of the Invention Thus, according to claim 1, there is provided a method for generating a feature-based localization map for GNSS and / or feature-based localization, comprising: At least the following steps: i) generating feature information for a feature-based localization map using at least one GNSS information; ii) generating GNSS-related meta-information that allows inversely inferring the GNSS situation on the basis of which the feature information was generated; iii) associating the GNSS-related meta-information generated in step ii) with the feature information generated in step i); A method is proposed which includes:

[0007] To implement the method, steps i), ii) and iii) can be performed, for example, at least once and / or repeatedly, in the order described. Furthermore, steps i), ii) and iii), in particular steps i) and ii), can be performed at least partially in parallel or simultaneously. The method is used, in particular, to generate feature-based localization maps for GNSS- and / or feature-based localization plausibility checks.

[0008] According to a further aspect, there is provided a method for GNSS and / or feature based localization, comprising: At least the following steps: a) receiving GNSS signals from GNSS satellites; b) retrieving at least one feature information from the feature-based localization map, wherein at least one GNSS-related meta-information associated with the feature information is further retrieved, the meta-information enabling inverse inference of the GNSS situation on the basis of which the feature information was generated; c) performing a GNSS and / or feature based localization using the received GNSS signals and / or at least one feature information and taking into account at least one GNSS related meta information associated with the feature information; A method is also proposed which includes:

[0009] To implement the method, steps a), b), and c) can be performed, for example, at least once and / or repeatedly in the order described. Furthermore, steps a), b), and c), in particular steps a) and b), can be performed at least partially in parallel or simultaneously. The method can be implemented, for example, for localization of a vehicle. The vehicle can be, for example, a motor vehicle, preferably configured for at least partially automated or autonomous driving operation. In principle, localization can be performed based on only one of the techniques (GNSS-based or feature-based). However, this is performed in particular taking into account at least one GNSS-related meta-information associated with each feature information.

[0010] The above method and the localization device described further below each advantageously contribute to improving, and in particular ensuring, the independence of GNSS-based and feature-based localization paths. Particularly advantageously, this can be achieved here without excluding the use of GNSS data in mapping. The method can be implemented, for example, for validity checks of GNSS- and / or feature-based localization.

[0011] The feature information may correspond to, for example, a segment or section from a feature-based localization map, which is typically a digital map that may store the feature information and GNSS-related meta-information associated with the feature information.

[0012] The GNSS related meta information is particularly suitable for at least contributing to or describing the GNSS situation on the basis of which the characteristic information is generated. The at least one GNSS related meta information comprises the following (GNSS) information (GNSS metadata): satellite type and constellation, satellite geometry (e.g., DOP information), Positioning type (code only, float ambiguity resolution, integer ambiguity resolution), GNSS-only (and / or sensor-fused) position offset information, Percentage of discarded measurements of pseudorange and carrier phase, RMS of the measurement residuals, · Residual variability, Signal-to-noise ratio information, Multifunction displays, and / or Cycle slips (discontinuities in the carrier phase measurement) may include one or more of:

[0013] According to an advantageous embodiment, it is proposed that the feature-based localization map is generated by the method described above.

[0014] According to a further advantageous embodiment, it is proposed to perform an assessment of the reliability of the localization result. In this context, the reliability of the feature-based localization map can also be determined. The reliability of the localization result and / or the localization map can be determined in particular depending on at least one GNSS-related meta-information associated with the feature information. The reliability can be provided in the form of an integrity measure, for example a so-called "integrity level".

[0015] According to a further advantageous embodiment, it is proposed to determine a measure for the independence of GNSS-based position determination from feature-based position determination taking into account at least one GNSS-related meta-information associated with the feature information, which measure for the independence of GNSS-based position determination from feature-based position determination can advantageously be (jointly) factored into or form the basis for an assessment of reliability.

[0016] In this context, it is further preferred to also take into account at least one GNSS-related meta-information associated with the GNSS signal to determine a measure of the independence of the GNSS-based position determination from the feature-based position determination. For example, the at least one GNSS-related meta-information associated with the feature information can be compared with the at least one GNSS-related meta-information associated with the GNSS signal. This can also be expressed as, in particular, comparing the GNSS-related meta-information stored for the feature information with the current GNSS-related meta-information (belonging to the currently received GNSS signal).

[0017] According to a further aspect, a computer program for carrying out the method presented herein is proposed, in other words a computer program product comprising instructions that, when executed by a computer, cause the computer to carry out the method described herein.

[0018] According to a further aspect, a machine-readable storage medium is proposed on which a computer program as proposed herein is saved or stored. Typically, a machine-readable storage medium is a computer-readable data carrier.

[0019] According to a further aspect, there is provided a location determination device for a vehicle, comprising: a GNSS-based positioning module for determining first location information based on GNSS measurements; a feature-based localization module for determining second location information based on at least one feature information from the feature-based localization map; an independence evaluation module for determining independence information based on at least one GNSS-related meta-information associated with the feature information, the independence evaluation module enabling inverse inference of a GNSS situation on the basis of which the feature information was generated; a fusion module for determining a location result using the first location information and the second location information and taking into account the independence information; It is proposed a location determination device comprising:

[0020] The vehicle may be, for example, a vehicle such as a car. Each module may be physically or functionally implemented in the location device, for example by a corresponding software architecture.

[0021] According to an advantageous embodiment, a localization device is configured to perform the method described herein. The localization device may, for example, include a computer and / or a controller capable of executing instructions for performing the method. To this end, the computer or controller may, for example, execute the computer program described above. For example, the computer or controller may access the storage medium described above in order to be able to execute the computer program.

[0022] The localization device may, for example, be a component of a motion and position sensor that can be or is located in or on the vehicle, in particular, or may be connected to such sensors for information exchange. In this context, for example, a GNSS sensor and / or a localization device may be a component of a motion and position sensor.

[0023] The details, features and advantageous embodiments discussed in relation to the method for generating may also be implemented in the method for localization and / or computer program and / or storage medium and / or localization device presented in this specification, and vice versa. To this extent, reference is made to the description therein in its entirety in order to express the nature of the features in more detail.

[0024] The solution and its technical environment presented in this specification will be described in more detail below with reference to the drawings. It should be noted that the present invention should not be limited by the illustrated embodiments. In particular, unless otherwise specified, partial aspects of the factual content illustrated in the drawings can be extracted and combined with other components and / or knowledge from other drawings and / or this specification. [Brief explanation of the drawings]

[0025] [Figure 1] 1 is a schematic diagram of an exemplary structure of a location system according to the prior art; [Figure 2] FIG. 2 is a schematic diagram of an exemplary sequence of the method presented herein. [Figure 3] FIG. 2 is a schematic diagram of an exemplary sequence of a further method presented herein. [Figure 4] 1 is a schematic diagram of an exemplary structure of a localization device presented herein. [Figure 5] 1 is a schematic diagram of a vehicle having a location device as described herein. DETAILED DESCRIPTION OF THE INVENTION

[0026] 1 shows a schematic diagram of an exemplary structure of a prior art localization device 1. The localization device 1 typically includes a GNSS-based localization module 2, a feature-based localization module 3, and a fusion module 4.

[0027] This allows for fusion of vehicle localization estimates from different vehicle localization modalities. In this context, exemplarily here, first position information 5 from the GNSS-based localization module 2 and second position information 6 from the feature-based localization module 3 are directly fed to the fusion module 4, which determines a localization result 7 from these position information. However, in this case, the independence of the localization paths used is not properly taken into account.

[0028] Current and future automated (AD) and assisted (DA) vehicle systems rely to a large extent on additional information from digital maps. Thus, increasingly complex driving tasks can be automatically completed based on a reduced set of sensors. Map information can be used to localize (position and orientation) the vehicle relative to the map being used.

[0029] For safety-related functions, combined localization solutions based on independent localization paths are usually considered. A typical combination of localization paths is GNSS-based localization and feature-based localization. These two localization paths should be as independent of each other as possible. However, the use of GNSS data, especially for map construction, can compromise this independence.

[0030] In relation to the prior art, it could be observed that the use of GNSS data in map construction, for example, can lead to a loss of independence between feature-based and GNSS-based localization, which can lead to critical system failures and should be avoided.

[0031] 2 illustrates an exemplary sequence of the method presented herein, which is used to generate a feature-based localization map for GNSS and / or feature-based localization. The order of steps i), ii), and iii) indicated by blocks 110, 120, and 130 is exemplary and can be completed in the illustrated order, e.g., at least once, to implement the method.

[0032] In block 110, according to step i), it is implemented to generate feature information for a feature-based localization map using at least one GNSS information. In block 120, according to step ii), it is implemented to generate GNSS-related meta information that allows inversely inferring the GNSS situation on the basis of which the feature information was generated. In block 130, according to step iii), it is implemented to associate the GNSS-related meta information generated in step ii) with the feature information generated in step i).

[0033] 3 shows a schematic diagram of an exemplary sequence of a further method presented herein, the method being used for GNSS and / or feature-based localization. The order of steps a), b), and c) indicated by blocks 210, 220, and 230 is exemplary and may be completed in the illustrated order, e.g., at least once, to implement the method.

[0034] In block 210, according to step a), receiving GNSS signals from GNSS satellites is performed. In block 220, according to step b), retrieving at least one feature information from the feature-based localization map is performed, wherein at least one GNSS-related meta-information associated with the feature information is also retrieved, which allows inversely inferring the GNSS situation on the basis of which the feature information was generated. In block 230, according to step c), performing GNSS- and / or feature-based localization is performed using the received GNSS signals and / or the at least one feature information and taking into account the at least one GNSS-related meta-information associated with the feature information.

[0035] In the method described herein (FIG. 3), it is particularly advantageous if the feature-based localization map is generated by the method described above (FIG. 2).

[0036] The method can advantageously include an assessment of the reliability of the localization result. In this context, the reliability of the feature-based localization map can also be determined. The reliability of the localization result and / or the localization map can be determined in particular depending on at least one GNSS-related meta-information associated with the feature information. The reliability can be provided, for example, in the form of an integrity measure, such as a so-called "integrity level".

[0037] Furthermore, the method may determine a measure of independence of GNSS-based position determination from feature-based position determination taking into account at least one GNSS-related meta-information associated with the feature information. The measure of independence of GNSS-based position determination from feature-based position determination may also be determined taking into account at least one GNSS-related meta-information associated with the GNSS signals. For example, in this context, the at least one GNSS-related meta-information associated with the feature information and the at least one GNSS-related meta-information associated with the GNSS signals may be compared with each other.

[0038] 4 shows a schematic diagram of an exemplary structure of the localization device 1 presented herein, which is suitable for or provided and configured for use in a vehicle 10.

[0039] The positioning device 1 includes a GNSS-based positioning module 2 for determining first position information 5 based on GNSS measurements. The positioning device 1 further includes a feature-based positioning module 3 for determining second position information 6 based on at least one feature information from a feature-based positioning map. The positioning device 1 further includes an independence evaluation module 8 for determining independence information 9 based on at least one GNSS-related meta-information associated with the feature information, which enables inverse inference of a GNSS situation that was the basis for generating the feature information. The positioning device 1 further includes a fusion module 4 for determining a positioning result 7 using the first position information 5 and the second position information 6 and taking the independence information 9 into consideration.

[0040] A particular advantage of the localization device 1 resides in the independence assessment module 8. The independence assessment module 8 assesses the independence of the localization outputs based in particular on GNSS-related meta-information.

[0041] The localization device 1 can be configured to implement, for example, the above-described method.

[0042] The above method and localization device 1 can each contribute to ensuring the independence of (redundant) GNSS-based and feature-based localization paths in a location system that is as secure as possible. A particular advantage of the method and localization device 1 described herein resides in particular in the generation, storage and / or use of GNSS-related meta-information in feature-based localization maps.

[0043] Exemplary GNSS-related meta-information that can be used herein is: satellite type and constellation, satellite geometry (e.g., DOP information), Positioning type (code only, float ambiguity resolution, integer ambiguity resolution), GNSS-only (and / or sensor-fused) position offset information, Percentage of discarded measurements of pseudorange and carrier phase, RMS of the measurement residuals, · Residual variability, Signal-to-noise ratio information, Multifunction display, Cycle slip is.

[0044] A particular advantage resides in the use of GNSS-related meta-information during operation of the location device 1, especially in the independence assessment module 8.

[0045] The feature-based localization module 3 can load feature map segments (examples of feature information) that match the current map-relative vehicle position and vehicle orientation (attitude). The (feature-based) vehicle pose (second location information 6) can then be advantageously estimated based on the current sensor measurements and the loaded feature map segments. The corresponding sensor measurements can be performed by the vehicle's sensors, such as GNSS sensors, inertial sensors, and / or environmental sensors (cameras, radar, LiDAR, ultrasonic, etc.).

[0046] The vehicle position estimate thus determined (second position information 6) can then be transmitted to an independence assessment module 8 together with at least one piece of GNSS-related meta-information associated with the characteristic information.

[0047] Furthermore, the GNSS-based localization module 2 may send GNSS-related meta-information and the current (GNSS-based) vehicle position estimate (first position information 5) to the independence assessment module 8.

[0048] The independence assessment module 8 is able to assess the meta-information and advantageously assess the degree of independence between the GNSS-based positioning input and the feature-based positioning input, the result of which can advantageously be represented by an independence flag.

[0049] If the independence flag indicates a sufficiently high degree of independence, both vehicle position estimates can be fused in the fusion module 4. If a dependency is identified by the independence evaluation module 8, the fusion module 4 can only provide a positioning result that is (correspondingly) lower or of correspondingly reduced integrity. Correspondingly reduced integrity may result in degradation of the DA / AD functionality using the positioning result.

[0050] For example, the GNSS constellation geometry used for recording GNSS data for mapping and the satellite or GNSS constellation geometry used for GNSS-based positioning may be similar or even poor. Therefore, significant coupling and / or loss of independence between the two positioning paths may occur, and it may be assumed that adverse safety effects cannot be eliminated. In this case, the independence of the two positioning paths cannot be assumed. Therefore, in such cases of DA / AD functionality, vehicle positioning results with a high degree of integrity may not be available.

[0051] In a further example, the GNSS signals are corrupted by multipath propagation, i.e. by reflections on surrounding obstacles, which leads to erroneous signal delays and thus to increased residuals and strong fluctuations in signal strength. In such an exemplary case, the residuals and the current quality indicators, such as CN0, can reflect that the vehicle is in a difficult environment, both in the metadata and in the actually observed data. In this example case, a high level of integrity cannot be guaranteed for the GNSS-based vehicle localization, nor for the map data originally calculated based on poor GNSS signal quality.

[0052] 5 shows a schematic representation of a vehicle 10 having a location device 1 as described herein. The vehicle 10 may be, for example, a motor vehicle, preferably configured for at least partially automated driving operation.

[0053] The above method and the above location device may contribute to achieving one or more of the following advantages: · GNSS-based and property / feature-based location paths can be independent without the need to exclude GNSS data from the property / feature map creation. If independence between mapping (and thus feature-based localization) and GNSS-based localization cannot be achieved, this problem can also be handled in the vehicle. In such cases, the methods and localization devices described herein can improve the independence and provide vehicle localization with the highest possible integrity for AD / DA functions / vehicle systems. The map independence requirement can be significantly reduced, so GNSS data can be used to any extent in mapping. The method and localization device described herein advantageously allow potentially safety-related independence requirements to be transformed into availability requirements. This is possible, among other things, because the safety requirements can be shifted towards the independence assessment module and thus away from the feature map. This can facilitate system design by providing additional design possibilities.

Claims

1. 1. A method for generating a feature-based localization map for GNSS and / or feature-based localization, comprising: At least the following steps: i) generating feature information for said feature-based localization map using at least one GNSS information; ii) generating GNSS-related meta-information that allows inverse inference of the GNSS context on which the characteristic information was generated; iii) associating the GNSS-related meta-information generated in step ii) with the feature information generated in step i); The method includes:

2. 1. A method for GNSS and / or feature based localization, comprising: At least the following steps: a) receiving a GNSS signal from a GNSS satellite; b) retrieving at least one feature information from the feature-based localization map, wherein at least one GNSS-related meta-information associated with said feature information is further retrieved, said meta-information enabling reverse inference of a GNSS situation on the basis of which said feature information was generated; c) performing a GNSS- and / or feature-based position determination using the received GNSS signals and / or the at least one feature information and taking into account at least one GNSS-related meta-information associated with the feature information; The method includes:

3. The feature-based localization map was generated by the method of claim 1. The method of claim 2.

4. An assessment of the reliability of the location result is performed; The method of claim 2.

5. determining a measure of independence of the GNSS-based position location from the feature-based position location taking into account at least one GNSS-related meta-information associated with the feature information; The method of claim 2.

6. determining a measure of independence of the GNSS-based position location from the feature-based position location, also taking into account at least one GNSS-related meta-information associated with the GNSS signals; The method according to claim 5.

7. A computer program for carrying out the method according to claim 1 or 2.

8. A machine-readable storage medium having stored thereon the computer program of claim 7.

9. A location determination device (1) for a vehicle (10), comprising: a GNSS-based position determination module (2) for determining first position information (5) based on GNSS measurements; a feature-based localization module (3) for determining second location information (6) based on at least one feature information from the feature-based localization map; an independence assessment module (8) for identifying independence information (9) based on at least one GNSS-related meta-information associated with said feature information, said independence information enabling a reverse inference of the GNSS context on which said feature information was generated; a fusion module (4) for determining a location determination result (7) using said first location information (5) and said second location information (6) and taking into account said independence information (9); A location determination device (1) comprising:

10. 3. A method for performing a method according to claim 2, A location device (1) according to claim 9.

Images