POSITION ESTIMATION IN DIFFERENT COORDINATE SYSTEMS

Description

[0001] The present invention relates to a method, a computer program with instructions and a device for position estimation, in particular for simultaneous position estimation in at least two coordinate systems.

[0002] In a land vehicle, position estimates are required for a variety of applications, particularly in the area of ​​driver assistance systems and highly automated driving. The position estimates must sometimes have different qualities for different applications.

[0003] For example, algorithms that independently control driving behavior require smooth trajectories and motion information. For such algorithms, short-term stability is important, whereas long-term drift is tolerable. Absolute position information is not necessary. This is therefore a relative local coordinate system.

[0004] Algorithms that independently create a map of the environment or use data from custom-built local maps require precise knowledge of the position in that environment. Long-term drift must be compensated for in this case. This is therefore an absolute local coordinate system.

[0005] Algorithms that require the vehicle's position on the globe, for example, for map comparison, require precisely this information. Here, too, long-term drift is compensated for. The coordinate system is therefore absolutely global.

[0006] For example, US 2018 / 0031387 A1 describes a state estimation system that uses long-range visual stereo odometry and integrates GPS, barometric, and inertial measurements. The system consists of two main components: an EKF (Extended Kalman Filter) and a visual odometry. For the visual odometry, the system uses the EKF information for robust camera position tracking. The visual odometry outputs serve as measurements for the EKF state update.

[0007] DE 10 2017 200 234 A1 describes a method for referencing a local trajectory in a global coordinate system, wherein the local trajectory comprises ordered local positions in a local coordinate system.

[0008] DE 10 2018 207 857 A1 describes a method for transforming a position of a vehicle between a first coordinate system and a second coordinate system.

[0009] DE 10 2015 217 497 A1 describes a method for synchronizing a first local environment map with a second local environment map in an environment of a motor vehicle.

[0010] The position calculation or estimation for the various applications is therefore performed in different coordinate systems, each of which, viewed individually, has properties that meet different requirements. However, the simultaneous, complete calculation of positions in multiple coordinate systems is runtime-intensive, requiring a powerful and therefore expensive computer. The requirements for available computing power can be reduced by introducing simplifications. For example, correct estimation of covariances is omitted or the condition of smoothness of the relative trajectory is relaxed.

[0011] It is an object of the present invention to provide solutions for improved position estimation in at least two coordinate systems.

[0012] This object is achieved by a method for position estimation having the features of claim 1, by a computer program with instructions having the features of claim 5, and by a device for position estimation having the features of claim 6. Preferred embodiments of the invention are the subject of the dependent claims.

[0013] According to a first aspect of the invention, a method for position estimation in at least two coordinate systems comprises the steps: Receiving sensor data for a position estimation in a relative local first coordinate system; performing a position estimation in the first coordinate system using the sensor data; providing position information resulting from the position estimation from the first coordinate system for a position estimation in an absolute global second coordinate system; fusing data from a global navigation satellite system and the provided position information into a position estimate in the second coordinate system; determining correlations between states for two points in time, wherein a state clone of a state measured in the second coordinate system is generated in the first coordinate system for determining the correlations; and outputting position estimates (LP, GP) in the first coordinate system and in the second coordinate system to at least one external application.

[0014] According to a further aspect of the invention, a computer program comprises instructions which, when executed by a computer, cause the computer to carry out the steps of a method according to claim 1.

[0015] The term "computer" should be understood broadly. In particular, it also includes control units, controllers, embedded systems, and other processor-based data processing devices.

[0016] The computer program may, for example, be made available for electronic retrieval or stored on a computer-readable storage medium.

[0017] According to a further aspect of the invention, a device for position estimation in at least two coordinate systems comprises: a receiving unit for receiving sensor data for a position estimation in a relative local first coordinate system; a first position estimator for performing a position estimation in the first coordinate system using the sensor data and for providing position information resulting from the position estimation from the first coordinate system for a position estimation in an absolute global second coordinate system; a second position estimator for fusing data from a global navigation satellite system and the provided position information into a position estimate in the second coordinate system; wherein correlations between states are determined for two points in time, wherein for determining the correlations a state clone of a state measured in the second coordinate system is generated in the first coordinate system; and an output (27) for outputting position estimates (LP, GP) in the first coordinate system and in the second coordinate system to at least one external application.

[0018] In the inventive solution, the simultaneous position estimation in different coordinate systems is structured in such a way that, for at least one coordinate system, the results of the position estimation from another coordinate system are further processed, i.e., previous results are used. In this way, unnecessary redundant calculations can be eliminated. The inventive solution is thus more runtime-efficient. Since the estimation of the relative trajectory occurs only once in the system, the computing time is reduced. At the same time, the conditions of all coordinate systems can be met. A further advantage is that the position estimations have a modular structure, thus reducing complexity. Compared to a monolithic estimation for each coordinate system, the complexity per model is low in the inventive solution.

[0019] According to the invention, the first coordinate system is a relative local coordinate system, whereas the second coordinate system is an absolute global coordinate system. The basis of the inventive solution is the estimation of a relative local position, i.e., the first coordinate system is a relative local coordinate system. This is advantageous because a large number of measurements are usually available for estimating the relative local position. The results of the relative local position estimates can then be used for an absolute global position estimate.

[0020] According to one aspect of the invention, the position information in the first coordinate system is provided as a relative movement between two points in time or as a continuous position with increasing variance. To make the relative position in the first coordinate system available for a position estimation in the second coordinate system, it can be output, for example, as a relative movement between two points in time. Alternatively, it can be output as a continuous position with a continuously increasing variance.

[0021] According to one aspect of the invention, correlations between states are determined for two points in time. For the correct calculation of relative movements between two points in time, it is advantageous if, in addition to knowledge of the uncertainty of the states themselves, knowledge of correlations between the two states is also available. Knowledge of correlations can increase the accuracy of position estimates.

[0022] According to one aspect of the invention, to determine correlations, a state clone of a state measured in the second coordinate system is generated in the first coordinate system. One possibility for calculating the correlations is the use of the so-called "stochastic cloning" method [1]. A further advantage of this method is that knowledge of state correlations can be used to merge relative measurements, such as those that occur in a scan matching method.

[0023] According to one aspect of the invention, a covariance matrix of the state is carried along to the state clone. This ensures a correct estimation of the full covariance matrix despite distributed estimation algorithms.

[0024] According to one aspect of the invention, position estimates in two absolute coordinate systems are used for plausibility checks. The modularity of the inventive solution makes it possible to integrate multiple absolute coordinate systems. If at least two different absolute coordinate systems with the associated measurements are available, it is possible to perform a plausibility check of the position estimates.

[0025] According to one aspect of the invention, position estimates in the first coordinate system and in the second coordinate system are output to at least one external application. The position estimates are thus available for the various applications with the required quality.

[0026] Preferably, a method or device according to the invention is used in a means of transportation, e.g., in an autonomous, semi-autonomous, or motor vehicle equipped with at least one assistance system. For example, a solution according to the invention can be used in this context for systems for controlling driving behavior or for global positioning using map matching.

[0027] Further features of the present invention will become apparent from the following description and the appended claims taken in conjunction with the figures. Figure overview

[0028] Fig. 1 schematically shows a method for position estimation in at least two coordinate systems; Fig. 2 schematically shows a first embodiment of a device for position estimation in at least two coordinate systems; Fig. 3 schematically shows a second embodiment of a device for position estimation in at least two coordinate systems; Fig. 4 schematically shows a means of transport in which a solution according to the invention is implemented; Fig. 5 schematically shows a conventional approach for position estimation in a local and a global coordinate system; Fig. 6 schematically shows an inventive approach for position estimation in a local and a global coordinate system; and Fig. 7 schematically shows the sequence of an inventive method using stochastic cloning. Character description

[0029] To better understand the principles of the present invention, embodiments of the invention are explained in more detail below with reference to the figures. Like reference numerals are used in the figures for like or equivalent elements and are not necessarily described again for each figure. It is understood that the invention is not limited to the illustrated embodiments and that the described features can also be combined or modified without departing from the scope of the invention as defined in the appended claims.

[0030] Fig. 1 shows a schematic view of a method for position estimation in at least two coordinate systems. In a first step S1, sensor data for a position estimation in a first coordinate system is received. Using the sensor data, a position estimation is then carried out in the first coordinate system S2. Position information from the first coordinate system resulting from the position estimation is subsequently provided S3 for a position estimation in a second coordinate system, e.g. as a relative movement between two points in time or as a continuous position with increasing variance. The provided position information is then merged S4 into a position estimate in the second coordinate system. The position estimates are finally output to at least one external application S5. The position estimate in the first coordinate system can in particular be output as a continuous position with increasing variance.The first coordinate system is a relative local coordinate system, whereas the second coordinate system is an absolute global coordinate system. To correctly calculate relative movements between two points in time, correlations between states are preferably determined for two points in time. For this purpose, a state clone of a state measured in the second coordinate system can be generated in the first coordinate system. Preferably, a covariance matrix of the state is included for the state clone.

[0031] Fig. 2 shows a simplified schematic representation of a first embodiment of a device 20 for position estimation in at least two coordinate systems, which in this example are a relative local coordinate system and an absolute global coordinate system. The device 20 has an input 21, via which a receiving unit 22 receives sensor data S and data GNSS of a global navigation satellite system for position estimations. The sensor data S enable a position estimation in a first coordinate system. A first position estimator 23 is configured to perform a position estimation in the first coordinate system using the sensor data S. The first position estimator 23 is also configured to provide a second position estimator 24 with position information BI, LP from the first coordinate system resulting from the position estimation for a position estimation in the second coordinate system, e.g.as a relative movement BI between two points in time or as a continuous position LP with increasing variance. The second position estimator 24 is further configured to merge the provided position information BI, LP into a position estimate GP in the second coordinate system using the GNSS data of the global navigation satellite system. The position estimates LP, GP are output to an external application via an output 27 of the device 20 for further use. To correctly calculate relative movements between two points in time, correlations between states for two points in time are determined. For this purpose, a state clone of a state measured in the second coordinate system is generated in the first coordinate system.

[0032] Preferably, a covariance matrix of the state is carried along to the state clone.

[0033] The receiving unit 22, the first position estimator 23, and the second position estimator 24 can be controlled by a control unit 25. Settings of the receiving unit 22, the two position estimators 23, 24, or the control unit 25 can be changed via a user interface 28. The data generated in the device 20 can be stored in a memory 26 of the device 20 if necessary, for example, for later evaluation or for use by the components of the device 20. The receiving unit 22, the two position estimators 23, 24, and the control unit 25 can be implemented as dedicated hardware, for example, as integrated circuits. Of course, they can also be partially or completely combined or implemented as software running on a suitable processor, for example, a GPU or a CPU.The input 21 and the output 27 can be implemented as separate interfaces or as a combined bidirectional interface.

[0034] Fig. 3 shows a simplified schematic representation of a second embodiment of a device 30 for position estimation in at least two coordinate systems. The device 30 has a processor 32 and a memory 31. For example, the device 30 is a control unit or a controller. Instructions are stored in the memory 31 which, when executed by the processor 32, cause the device 30 to carry out the steps according to one of the described methods. The instructions stored in the memory 31 thus embody a program executable by the processor 32 which implements the method according to the invention. The device 30 has an input 33 for receiving information, in particular measurement data from one or more sensors. Data generated by the processor 32 are provided via an output 34. Furthermore, they can be stored in the memory 31.The input 33 and the output 34 can be combined to form a bidirectional interface.

[0035] The processor 32 may include one or more processor units, such as microprocessors, digital signal processors, or combinations thereof.

[0036] The memories 26, 31 of the described devices can have both volatile and non-volatile memory areas and can comprise a wide variety of storage devices and storage media, for example hard disks, optical storage media or semiconductor memories.

[0037] Fig. 4 shows a schematic representation of a means of transport 40 in which a solution according to the invention is implemented. In this example, the means of transport 40 is a motor vehicle, in particular an autonomous vehicle, a semi-autonomous vehicle or one equipped with assistance systems 41. The motor vehicle has a sensor system 42 with which measurement data can be recorded. The sensor system 42 can in particular comprise sensors for detecting the environment, e.g. ultrasonic sensors, laser scanners, radar sensors, lidar sensors or cameras, as well as sensors for determining relative movements, e.g. sensors for odometry, gyroscopes or acceleration sensors. The motor vehicle also has a device 20 for estimating position in at least two coordinate systems, which device can be used, for example, by the assistance system 41. The device 20 can of course also be integrated into the assistance system 41.In this example, other components of the motor vehicle include a navigation system 43 and a data transmission unit 44. A connection to a backend can be established using the data transmission unit 44. A memory 45 is provided for storing data. Data exchange between the various components of the motor vehicle takes place via a network 46.

[0038] In the following, further details of the solution according to the invention will be explained with reference to Fig. 5 bis Fig. 7 The position determination is shown for two coordinate systems, but the described solution is scalable to any number of coordinate systems.

[0039] Fig. 5 schematically shows a conventional approach for position estimation in a local and a global coordinate system. The sensor data S and the GNSS data of a global navigation satellite system are provided to a local position estimator 50 and a global position estimator 51. Both position estimators 50, 51 perform a complete calculation or estimation of the position LP, GP in their respective coordinate systems. The resulting position estimates LP, GP are then provided for further use.

[0040] Fig. 6 shows a schematic of an inventive approach for position estimation in a local and a global coordinate system. The basis of the illustrated system is the estimation of a relative local position LP by the local position estimator 50 based on the sensor data S. This relative local position LP is output to external applications via an interface for further use. Position information BI, LP resulting from the position estimation is also made available to the global position estimator 51. This can, for example, take the form of a relative movement BI between two points in time or as a continuous position LP with constantly increasing variance. The global position estimator 51 uses this data as well as the GNSS data of the global navigation satellite system to estimate the global position GP.This global position GP is also output to external applications via an interface for further use.

[0041] To correctly calculate relative movements between two points in time, in addition to the uncertainty of the states themselves, knowledge of correlations between the two states is also relevant. Calculating these correlations is made possible, for example, by stochastic cloning.

[0042] Fig. 7shows a schematic of the flow of a method according to the invention using stochastic cloning. The flow is shown in the form of a simplified UML sequence diagram (UML: Unified Modeling Language). The starting point of the method is input data, which in this case consists of sensor data S and GNSS data from a global navigation satellite system. The sensor data S forms the basis for an estimate of the relative local position, and the GNSS data from the navigation satellite system forms the basis for an estimate of the absolute global position. In addition to the measurements (RelativeMeasurement()) in the local coordinate system, the generation of a state clone (CreateStateClone()) is triggered at the time of a measurement (AbsoluteMeasurement()) in the absolute coordinate system in the estimation model of the relative trajectory.Until the next measurement in the absolute coordinate system, stochastic cloning ensures that the full covariance matrix, including the correlation, is correctly maintained from the current state to the state clone. The maintained covariance matrix and the relative motion of the state clone can then be used for the estimation model of the absolute trajectory (GetMotionAndCovariance()). At the time of the next measurement in the absolute coordinate system, the creation of a new state clone or the update of the existing state clone is triggered in the estimation model of the relative trajectory (CreateOrUpdateStateClone()). References

[0043] [1] SI Roumeliotis et al.: "Stochastic cloning: a generalized framework for processing relative state measurements", Proceedings of the 2002 IEEE International Conference on Robotics and Automation (ICRA), pp. 1788-1795.

Claims

1. Method for estimating a position in at least two coordinate systems, having the steps of: - receiving (S1) sensor data (S) for a position estimation in a relative local first coordinate system; and - performing a position estimation (S2) in the first coordinate system using the sensor data (S); - providing (S3) position information (BI, LP) resulting from the position estimation (S2) from the first coordinate system for a position estimation in an absolute global second coordinate system; - fusing (S4) data from a global navigation satellite system and the position information (BI, LP) provided to form a position estimation in the second coordinate system; - determining correlations between states for two points in time, wherein a state clone of a state measured in the second coordinate system is generated in the first coordinate system for determining the correlations; and - outputting position estimations (LP, GP) in the first coordinate system and in the second coordinate system to at least one external application.

2. Method according to Claim 1, wherein the position information in the first coordinate system is provided (S3) as a relative movement (BI) between two points in time or as a continuous position (LP) with increasing variance.

3. Method according to Claim 1 or 2, wherein a covariance matrix of the state is carried to the state clone.

4. Method according to one of the preceding claims, wherein position estimations (GP) in two absolute coordinate systems are used for a plausibility check.

5. Computer program having instructions which, when executed by a computer, cause the computer to carry out the steps of a method according to one of Claims 1 to 4 for estimating a position in at least two coordinate systems.

6. Device (20) for estimating a position in at least two coordinate systems, having: - a receiving unit (22) for receiving (S1) sensor data (S) for a position estimation in a relative local first coordinate system; - a first position estimator (23) for performing a position estimation (S2) in the first coordinate system using the sensor data (S) and for providing (S3) position information (BI, LP) resulting from the position estimation (S2) from the first coordinate system for a position estimation in an absolute global second coordinate system; - a second position estimator (24) for fusing (S4) data from a global navigation satellite system and the position information (BI, LP) provided to form a position estimation in the second coordinate system; wherein correlations between states are determined for two points in time, wherein a state clone of a state measured in the second coordinate system is generated in the first coordinate system for determining the correlations; and - an output (27) for outputting position estimations (LP, GP) in the first coordinate system and in the second coordinate system to at least one external application.