Monitoring and detection of sensor errors in an inertial measurement system - Patents.com

Abstract

The present invention uses the angular velocity (ω A ,ω B ) and specific force (f A ,f B The present invention relates to a method for detecting a failure (F) of two inertial measurement units (4, 5) used in a vehicle (22) for measuring a vehicle speed (V) and a vehicle position (V), the method comprising the steps of: detecting a failure (F) of two inertial measurement units (4, 5) used in a vehicle (22) for measuring a vehicle speed ... In order to detect a fault (F) by comparing the sensor signals of the master inertial measurement unit (4), a fault model parameter is estimated with respect to the master inertial measurement unit (4) and is used as a reference value for correcting the measurement values ​​of the slave inertial measurement unit (5), and in an object detection unit (17), one of the two inertial measurement units (4, 5) whose motion estimation is causing an erroneous prediction of the calculated position of an object around the vehicle (22) is detected as a fault, among the motion estimations calculated by Kalman filters (13, 14) arranged downstream of each of the two inertial measurement units (4, 5).

Description

[Technical field]

[0001] The present invention relates to a method for detecting faults in an inertial measurement unit according to the preamble of claim 1. [Background technology]

[0002] Inertial Measurement Units (IMU) are an arrangement of accelerometers and gyroscopes that are used to measure the specific force (acceleration) and rotational speed of an object in space. Inertial Measurement Units play a key role in estimating the vehicle's motion state, which is required for vehicle control systems, driver assistance systems and autonomous driving to correctly interpret information from e.g. ambient sensors (camera, LiDAR, radar). The term Vehicle Motion Observer (VMO) has been established for an arrangement and method for calculating the vehicle specific motion from the IMU data in combination with additional vehicle sensor systems (Non-Patent Document 5). Such an arrangement is shown in Figure 1. In Figure 1, an application example of an IMU 1 for motion estimation and navigation according to the prior art is shown. The IMU 1 transmits rotational speed signals from the gyroscopes and specific force signals from the accelerometers (Non-Patent Document 9). In the sensor integration unit 2, motion variables such as the position angle α, the speed v and / or the position P are calculated by integration from these values ​​and are corrected with the aid of other sensors or sensor units 3 (odometer, magnetometer, barometric altimeter, steering angle, vehicle level and / or GNSS receiver, etc.). In the following, the result of the described data fusion is called the estimated motion state or motion estimate for short. Where the vehicle motion can be confused with the motion of objects around the vehicle, the terms proper motion state or proper motion estimate should be used for the vehicle.

[0003] An IMU sensor is said to be faulty if the defined (and therefore predetermined by design) relationship between the physical measurements and the output sensor signals is lost in the sensor, so that the specified measurement behavior does not occur. In that case, the corresponding IMU sensor cannot be used. If a failure in the sensor goes undetected and is propagated to the motion calculations, it can lead to significant errors or grossly erroneous values ​​in the useful signals, and therefore a safety risk in safety-related applications.

[0004] Vehicle control systems, driver assistance systems, and autonomous driving require extremely high safety standards, which means that IMU sensor failures must be automatically detected and localized, and the inaccurate sensor must be removed from motion calculations, for example by switching to a redundant sensor (backup, backup solution), within a time period set by the system (FTTI: Fault Tolerant Time Interval).

[0005] The class of automotive IMU systems considered here (automotive-grade IMUs) already has rudimentary monitoring capabilities, especially with regard to electrical faults. This type of monitoring is referred to below as monolateral monitoring, since it is based only on the information of one sensor each. Monolateral methods have low sensitivity and can only detect errors with relatively large amplitudes, i.e. the monitoring threshold is generally too high for driver assistance systems or autonomous driving. Similarly, the diagnostic coverage for minor errors that could represent a safety risk is also low. This diagnostic coverage can be significantly improved by the introduction of redundant sensor elements.

[0006] Non-Patent Document 4 summarizes the current state of the art for simple redundant sensor arrangements. A solution to this is shown in Fig. 2. In this case, two redundant inertial measurement units 4 (IMU A) and 5 (IMU B) are first calibrated to each other with respect to their systematic errors (bias, sensitivity, misalignment, etc.), and then the corresponding sensor values ​​thus calibrated are compared. If the difference in the measurements exceeds a threshold, a fault is considered in the respective sensor pair. However, it is not yet possible to assign the fault to one of the two sensors, since an additional reference value would be required for this. Non-Patent Document 4 attempts to solve this problem by monitoring the innovations of the respective Kalman filters for estimating the proper motion. By fusing the motion state predicted with the aid of the IMU in the two Kalman filters with the remaining vehicle sensor system (wheel rotation speed, steering angle), making the vehicle sensor system a third independent source of information, the IMU error can be localized by evaluating the innovations (weighted difference between prediction and measurement). A drawback of this procedure is that the monitoring procedure described in [4] is not equally selective in all degrees of freedom of motion of the IMU, since the measurements derived from the vehicle sensor system are mainly limited to translational speed. For example, the sensitivity to faults causing deviations in the rotational degrees of freedom is low. Furthermore, the detection of faults with small amplitudes is limited by the resolution of the wheel speed sensor system. In particular, in the case of fusion of estimated eigenmotion states with ambient information, as occurs in object tracking, there are high requirements, in part, due to the detection of IMU faults by smaller detection thresholds and shorter error detection times, which the described method probably cannot meet.

[0007] In Non-Patent Document 7 and Non-Patent Document 6, a general claim is made that it is possible to detect faults in the inherent motion estimation during the fusion of ambient sensors, but the sources do not include specific examples of this. Furthermore, the claim is generally focused on faults that result in physically unrealistic information about the vehicle's motion state. Nothing is said about faults where the motion state is physically possible but incorrect. Furthermore, the solutions proposed here do not include redundancy in the inertial measurement unit and in the calculation of the vehicle's own motion. However, redundancy in the inertial sensors is necessary to shorten the error detection time and to increase the sensitivity of error diagnosis due to low error thresholds. Summary of the Invention [Problem to be solved by the invention]

[0008] The object of the invention is to propose a new method for detecting faults in an inertial measurement unit. [Means for solving the problem]

[0009] This problem is solved according to the invention by a method having the features of claim 1.

[0010] Advantageous embodiments of the invention are the subject matter of the dependent claims.

[0011] In a method according to the present invention for detecting failures in two inertial measurement units used in a vehicle to measure angular velocity and specific force, two inertial measurement units each equipped with a number of sensors including an acceleration sensor and a gyro sensor are used, for example using a multi-hypothesis approach.

[0012] According to the present invention, a first inertial measurement unit is used as a master inertial measurement unit, and a second inertial measurement unit, which may have a lower performance than the first inertial measurement unit, is used as a slave inertial measurement unit. At this time, the measurement value of the master inertial measurement unit is used as a reference value for correcting the measurement value of the slave inertial measurement unit by estimating a fault model parameter with respect to the master inertial measurement unit in order to detect a fault by comparing two corresponding sensor signals from each of the two inertial measurement units. At this time, in the object tracking unit, the inertial measurement unit of the motion estimation, which causes an incorrect prediction of the calculated position of the object around the vehicle, among the motion estimations estimated by the Kalman filters arranged downstream of each of the two inertial measurement units, is detected as a fault.

[0013] In one embodiment, an error state from a monitoring unit supplied with sensor values ​​from an inertial measurement unit is transmitted to the object tracking unit, and in the event of a fault, multiple hypotheses regarding the motion state can be checked, while in the event of normal operation, one motion state is used.

[0014] In one embodiment, the vehicle's inherent motion is described by the vehicle pose and transmitted to the object tracking unit.

[0015] In one embodiment, in the object tracking unit, a new object position of an object or new object positions of multiple objects are predicted in vehicle fixed coordinates from the calculated hypothesis regarding the vehicle pose and a previously calculated object position of an object or multiple object positions of multiple objects.

[0016] In one embodiment, object positions of one or more objects measured by the vehicle's surrounding sensors are compared to predicted object positions from inherent motion hypotheses for the vehicle identified using a Kalman filter, and inherent motion hypotheses for which the deviation between the predicted object positions and the measured object positions exceeds a threshold are detected as incorrect.

[0017] In one embodiment, if a failure occurs in one of the inertial measurement units, the failure is detected by comparison with the redundant sensors of the other inertial measurement unit and a switchover occurs to the other inertial measurement unit if the proper motion hypothesis calculated with the inaccurate sensor values ​​is rejected as inaccurate.

[0018] In one embodiment, the monitoring results of two inertial measurement units are incorporated into the fault detection of the object tracking unit, where a fault is detected only if the deviation of the predicted object position for one of the hypotheses exceeds a threshold value but not for the other hypothesis, and at the same time an error is detected by the sensor comparison of the IMU fault detection unit.

[0019] In one embodiment, fault detection includes only one object detected as stationary or multiple objects detected as stationary. Faults are detected for a hypothesis where the predicted object position changes significantly while the other hypothesis is that the object position remains constant with respect to location.

[0020] The present invention appropriately exploits the higher sensitivity of the object tracking method with respect to errors in inherent motion estimation for error detection, thereby achieving detection even for very small amplitude IMU faults.

[0021] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. [Brief description of the drawings]

[0022] [Figure 1] FIG. 1 is a schematic diagram of a typical use case of an Inertial Measurement Unit (IMU) for estimating vehicle motion variables. [Diagram 2] FIG. 1 is a schematic diagram of IMU channel monitoring with simple redundancy. [Diagram 3] FIG. 1 is a schematic diagram of a solution according to the invention including ego-motion (proper motion) solution evaluation and error detection within the scope of object tracking. [Figure 4]1A and 1B are schematic diagrams for explaining the basic concept of the present invention in a two-dimensional space. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] In all the drawings, the same reference numerals are used to designate corresponding parts.

[0024] A method for detecting a fault F in an inertial measurement unit with simple redundancy uses two inertial measurement units 4 and 5, each equipped with multiple accelerometer and gyroscope sensors, as shown in FIG.

[0025] In the figure, the two inertial measurement units 4 and 5 are also denoted as IMU A and IMU B. Each of the inertial measurement units produces a vector with three angular rate signals:

number

number

number

number

[0026] In FIG. 2, a first inertial measurement unit 4 is used as a master inertial measurement unit, and a second inertial measurement unit 5, which may have lower performance than the first inertial measurement unit 4, is used as a slave or backup inertial measurement unit 12. In this case, the monitoring unit 12 estimates systematic fault parameters (fault model parameters, error model parameters) in the model of the slave inertial measurement unit 5 with respect to the master inertial measurement unit 4, and is used to detect a signal deviation between the corrected signals. A signal deviation that is too large may be caused by a potential fault F in either one of the two inertial measurement units 4 or 5. However, it is not possible to detect which of the two inertial measurement units 4 or 5 is affected by the malfunction (fault) F only from the detection in the monitoring unit 12. However, it is possible to detect which specific sensor pair is affected.

[0027] To calculate the vehicle's motion state, the angular velocity (vector ω A , vector ω B ) and the specific force signal (vector f A , vector f B ) are provided as inputs to one Kalman filter 13, 14 (Vehicle Motion Observer, VMO for short) where they are fused with measurements of other vehicle sensors (wheel speeds and steering angle, etc.). Each of the two Kalman filters 13, 14 then uses the same measurements. The Kalman filters 13, 14 determine the respective motion states BZ A, BZ B for the respective inertial measurement units 4, 5. As can be seen from FIG. 1, the aforementioned data fusion of the signals of the inertial measurement unit 1 with the remaining vehicle sensor systems is a standard approach for determining the vehicle motion variables.

[0028] The solution proposed in Non-Patent Document 4 as prior art is intended to monitor the innovations of two Kalman filters 13 and 14 in a monitoring unit 15, as shown in FIG. 2, and to use them together with the information sent from the monitoring unit 12 for error detection (fault detection) IF in a fault detection unit 16.

[0029] In the following description, the following mathematical notations are used:

[0030] t0: time u , u′ : Vehicle attitude estimated from IMU and vehicle sensor system Δq eb n : Vehicle position angle (vehicle attitude, vehicle direction) increment Δp eb n : Vehicle position increment C(Δq eb n ): Rotation matrix calculated from position angle increments H( u ): Homogeneous transformation from Earth-fixed coordinates to vehicle-fixed coordinates x :Object position z : Object position measurement using ambient sensor systems K: Kalman filter gain ν Innovation S: measurement error covariance matrix P - : State error covariance matrix before correction by measurement (a priori) R: Covariance matrix of measurements κ: threshold Ψ: Vehicle yaw angle relative to the Earth-fixed coordinate system ΔΨ: Yaw angle increment ω z : Yaw rate δω z : Yaw rate error v x : Vehicle longitudinal speed v y : Vehicle lateral speed Δt: the time interval between two measurements l r : Distance from the rear axle of the vehicle to the vehicle-fixed coordinate system

[0031] Unlike the conventional technology shown in FIG. 2, according to the present invention, as shown in FIG. 3, the calculation result of the vehicle motion state (correct vehicle posture) u and incorrect vehicle posture u′ The motion states (proper motion) and information (error state) from the monitoring unit 12 are fed (transmitted) to the object tracking unit 17. Here, the motion states are the positions of objects around the vehicle. x - Then, the predicted position x - In the object position correction unit 19, the object measurements from the surrounding sensor system are z and corrected accordingly (the corrected positions of the objects around the vehicle x + If there is a fault in the inertial measurement units 4 and 5, the predicted position x - and the measured or corrected position x + This, in combination with independent IMU error detection in the monitoring unit 12, is then used to detect IMU faults F, identify their location (assign the corresponding IMU) and, if necessary, isolate the IMU errors by switching to a redundant solution, i.e. the respective error-free inertial measurement unit 4, 5. If neither of the inertial measurement units 4, 5 has a fault F, the predicted position x - and the measured or corrected position x + There will be good agreement between

[0032] The detection problem is solved by propagating two redundant proper motion estimates to the object tracking unit 17, where these estimates can be used as hypotheses for the vehicle's proper motion.

[0033] The object tracking unit 17 typically receives data from various sensor sources (cameras, radar, lidar, ultrasound, etc.) about objects around the vehicle. Typically, these data are fused and tracked over time. The object tracking unit 17 thereby extracts object information, tracks the motion of these objects over time and predicts their future motion. In this way, dynamic objects such as vehicles and pedestrians can be tracked. Detection of static objects such as small obstacles on the road surface is another example of a task of the object tracking unit 17.

[0034] When fusing objects around the vehicle, the vehicle's proper motion needs to be corrected, which allows the calculation of the correct position and velocity of these objects. For this purpose, accurate and reliable motion estimates are used as input from the VMO 13 or 14 to the object tracking unit 17, where the proper motion states are used to predict the positions of the surrounding objects during the observation by the surrounding sensor system. Inaccurate proper motion states lead to incorrect prediction of the positions of these objects. For example, this can cause the position of stationary objects to change from one observation time to the next. All stationary objects being tracked are considered to be affected by this at the same time. This leads to the following basic idea of ​​the invention: if a sudden motion is predicted for objects around the vehicle that were previously classified as stationary objects (such as walls), there is a high possibility that there is an anomaly in the motion calculation.

[0035] The solution according to the invention consists in calculating two inherent motion estimates as hypotheses for the two redundant inertial measurement units 4 and 5 in their respective Kalman filters 13, 14 or VMOs. At the same time, the sensor values ​​of the inertial measurement units 4, 5 are fed to the monitoring unit 12. If the two inertial measurement units 4, 5 are working properly, they send similar sensor signals, the difference of which is below the threshold of the monitoring unit 12, and the inherent motion states calculated therefrom are also the same and correspond to the objects around the vehicle observed by the object tracking unit 17. On the one hand, if a fault F occurs in one of the IMU sensors 4, 5, this result is detected by the monitoring unit 12. On the other hand, the hypotheses of the vehicle inherent motion states differ, and an inaccurate solution leads to a discrepancy between the predicted and measured positions of the objects. On the other hand, the error-free motion states calculated from the error-free IMUs correspond to the observed objects. By combining the two schemes, both the inaccurate IMUs 4, 5 and the sensor with the fault F are unambiguously located, and the system can switch to the backup solution.

[0036] In one embodiment, the motion states BZ A, BZ B calculated from the two redundant inertial measurement units 4, 5 may be provided to an object tracking unit 17 for further processing.

[0037] In one embodiment, the error state from the IMU monitoring unit 12, ascertained by comparison of corresponding IMU sensors of the two inertial measurement units 4, 5, can be provided to the object tracking unit 17, so that the two hypotheses about the motion state only need to be checked if a fault F occurs.

[0038] Here, in one embodiment, the proper motion of the vehicle 22 is calculated as the vehicle attitude u It is described by:

number

[0039] Here, Δq ebn x n Axis and y n is the position increment (angular increment) of the vehicle 22 relative to a local Earth-fixed (tangential) coordinate system 20 having axes Δp eb n is the position increment of the corresponding vehicle 22 in the Earth-fixed coordinate system 20. Also, C(Δq eb n ) represents the rotation matrix that can be calculated from the position increments. The overall transformation (rotation and displacement) of the object from Earth-fixed coordinates to vehicle-fixed coordinates H( u ) can be written as follows by a homogeneous matrix (homogeneous transformation from Earth-fixed coordinates to vehicle-fixed coordinates) well known from practical use (Non-Patent Document 8, Non-Patent Document 10):

number

[0040] In one embodiment, a prediction of the positions of objects around the vehicle can be calculated from two conditions: the vehicle's inherent motion and previous object positions.

[0041] Δp o n represents the object's position in Earth-fixed coordinates, and hence its position in vehicle-fixed coordinates, Δp o b is given by the following formula:

number

[0042] In one embodiment, the object position of the object in Earth fixed coordinates x denotes the state of the object for object tracking.

number

[0043] Object location measurement using ambient sensor systems z are the measurements sent by the surrounding sensors (object position measurements by the surrounding sensor system).

number

[0044] These measurements need to be fused in a Kalman filter 13, 14 to localize the object. For simplicity, we assume that the object has already been classified as a static object. In this case, x must be constant, and the fusion task in the Kalman filter13,14 is to z Estimation of object state using x^ (the estimated position of the object) z k represents a sequence of measurements of an object at discrete time instants k=0,1,2,..., u k is the relevant vehicle attitude, x^ k - is the uncorrected estimate of the object position (a priori), and then the improved (a posteriori) position estimate x^ k + The correction to is performed according to the following formula:

number

[0045] The difference between the measurement and the predicted measurement determined from the vehicle specific motion and a priori position estimation is then the innovation ν k (Non-Patent Document 2, Non-Patent Document 3, Non-Patent Document 1) and is expressed by the following formula:

number

number

[0046] Here, P k - is the covariance matrix of the a priori state estimation error, and R k is the covariance matrix of the measurement errors. The latter includes basic statistical assumptions about the measurement errors and the accuracy (error distribution) of the proper motion estimates. The innovation vector ν k Each component ν ik satisfies the following formula:

number

[0047] That is, the difference between the measured and predicted measurements is the diagonal component of the covariance, s iik (For κ=3, 99.73% of the innovations in the statistical mean must be in the range described by equation (9); for κ=4, it would be 99.994%). If this is not the case, there is either an error in the measurement (outlier) or an error in the prediction (e.g. due to incorrectly calculated proper motion), where u k A and u k B are the two hypotheses of proper motion based on inertial measurement units A and B, 4 and 5, respectively; x^ k - Let,be the a priori estimated position of objects around the vehicle that are classified as static.,Faults in IMU,A,4,and the resulting u k A Due to the margin of error in the calculation, the following applies to the allocated innovation element:

number

[0048] In addition, the hypothesis u k B The following applies to the innovation elements allocated to:

number

[0049] In this case, hypothesis A is clearly false and it is detected that there is an error in proper motion A. In this way, the error detected by the IMU monitoring of the monitoring unit 12 can be unambiguously assigned to one of the two IMUs 4, 5.

[0050] This procedure is explained using a simple example in two-dimensional space with reference to Figures 4(A) and (B). Figure 4(A) shows a two-dimensional space with an Earth-fixed coordinate system 20 and x b Axis and y b The relationship between the two coordinate systems is shown in Fig. 2(a) and the vehicle-fixed coordinate system 21 with its axes. The yaw angle Ψ then represents the rotation between the two coordinate systems. The corresponding rotation matrix is:

number

[0051] In the example shown in FIG. 4B, the vehicle 22 moves at a constant speed v x , the vehicle 22 is moving straight ahead toward an object position 25 located directly in front of the vehicle 22, which is represented by the vector x It is described by:

number

[0052] At time t0, which corresponds to the discrete time k=0, the vehicle 22 performs a yaw angle increment ΔΨ0=0 and a position increment Δp b0 n It has.

[0053]

number

[0054] The motion of the vehicle 22 is described by the following three equations:

number

[0055] rotation speed ω z is also called yaw rate in this context and is measured by the inertial measurement units 4, 5. x is the body-fixed longitudinal velocity of the vehicle 22, which can be calculated by the VMO 13, 14 by fusion of wheel speed sensor and IMU data. In this example, a common simplification of the calculation is that the lateral velocity of the vehicle 22 is calculated from the following equation:

number

number

[0056] As a result, we obtain the following for the measured position of the object in vehicle coordinates:

number

[0057] Here, at time t0, there is a constant systematic error δω z appears on the IMU yaw rate sensor, the actual yaw rate ω z=0 instead of inaccurate yaw rate ω z =0+δω z (twenty one) is measured by the sensor. Therefore, the integration of equation (15) leads to an incorrect yaw angle at time t1 of ΔΨ1=δω z Δt (22)

[0058] This error propagates into the position calculation, resulting in an incorrect attitude 24 of the vehicle 22. To be more specific, the error in the yaw angle is small, and the trigonometric functions have a small-angle approximation.

number

number

[0059] Rotation matrix obtained from small angle approximation

number

number

[0060] As a result, we obtain the following equation for the measurement prediction from the vehicle eigenmotion:

number

[0061] This result can best be illustrated by a numerical example. v x =30m / s (28) l r=1.5m (29)

number

[0062] For the case of yaw rate without fault F, we obtain the following accurate measurement predictions:

number

[0063] In the case of yaw rate with fault F, we have:

number

[0064] That is, due to the erroneous proper motion, the position of the object 25 in the vehicle coordinates is predicted to deviate more than 0.4 m to the right in the y-component, while the measured object position 26 is always exactly in front of the vehicle 22 even in the case of fault F. The associated standard deviation is e.g.

number

[0065] In one embodiment, two hypotheses for the predicted object position can be compared to the object position 25, 26 measured by the ambient sensor system. If in one of the hypotheses the measured object position 25, 26 exceeds a threshold relative to the predicted object position 25, 26 but in the other hypothesis it does not, then an error in the corresponding inertial measurement unit 4, 5 can be inferred.

[0066] In one embodiment, the results of monitoring the two IMU sensors 4, 5 can be incorporated into the error detection of the object tracking unit 17. In this case, a fault F is detected only if the deviation of the predicted object position 25, 26 exceeds a threshold value for one of the hypotheses but not for the other hypothesis, and at the same time an error is detected by the sensor comparison of the IMU fault detection unit 16.

[0067] In one embodiment, the fault detection includes an object detected as stationary or multiple objects detected as stationary, and a fault F can be detected for a hypothesis in which the predicted object position 25, 26 changes significantly, while for the other hypothesis the object position remains constant.

[0068] The present invention allows fault-tolerant identification of the motion variables of the vehicle 22 in the event of a failure F in an IMU with simple redundancy (fail-to-fail operation), without imposing special requirements on the geometry and sensor quality. Rather, the present invention is also applicable to low-cost heterogeneous IMU hardware that is already partially present in the vehicle 22.

[0069] Taking into account the increased safety requirements during autonomous driving, the method according to the present invention makes it possible to protect the signal of the master inertial measurement unit 4 against sensor errors as a measure to ensure the functional safety of the estimation of the vehicle's inherent motion (continued operation in the event of a failure).

[0070] The key point here is the use of dissimilar (not similar) Inertial Measurement Units (IMUs) 4, 5 mounted at different points on the vehicle 22 in particular.

[0071] The solution according to the invention achieves fault tolerance in the calculation of vehicle motion variables, using systems and methods for object detection and fusion already present in driver assistance systems and autonomous vehicles. [Prior art documents] [Non-patent literature]

[0072] [Non-Patent Document 1] Berman, Z.: Outliers Rejection in Kalman Filtering-Some New Observations.In: IEEE / ION Position,Location and Navigation Symposium-PLANS 2014,S.1008-1013 9 [Non-Patent Document 2] Grewal,MS;Adrews,AP:Kalman Filtering.New York:John Wiley,2001 9 [Non-Patent Document 3] Groves,PD:Principles of GNSS,Inertial and Multisensor integrated Navigation systems.Artech House,2013 9 [Non-Patent Document 4] Kalkkuhl,J.;Bergmann,M.;Engelhardt,T.;Bleimund,F.:DE 10 2021 004 103 A1:Verfahren und Anordnung zur Uberwachung und Detektion von Sensorfehlern in Inertial-Mess-Systemen 3,17 [Non-Patent Document 5] Klier,W.;Reim,A.;Stapel,D.:Robust Estimation of Vehicle Sideslip Angle-An Approach w / o Vehicle and Tire Models.In:SAE Technical Paper Series (2008),Nr.2008-01-0582 1 [Non-Patent Document 6] Mercep,L.;Pollach,M.:US10553044B2:Self-diagnosis of faults with a secondary system in an autonomous driving system.4

Non-licensed Document 7

Non-licensed literature 9

Non-licensed literature 10

Claims

1. Angular velocity (vector ω Ａ , vector ω Ｂ ) and specific force (vector f Ａ , vector f Ｂ A method for detecting a failure (F) of two inertial measurement units (4, 5) used in a vehicle (22) for measuring a vehicle speed, comprising the steps of: Each of the two inertial measurement units (4, 5) comprises a plurality of sensors including an acceleration sensor and a gyro sensor; A first inertial measurement unit (4) is used as a master inertial measurement unit and a second inertial measurement unit (5), which may be of lower performance than the first inertial measurement unit (4), is used as a slave inertial measurement unit; the measurements of the master inertial measurement unit (4) are used as reference values ​​for correcting the measurements of the slave inertial measurement unit (5) by estimating fault model parameters for the master inertial measurement unit (4) in order to detect faults (F) by comparing two corresponding sensor signals from each of the two inertial measurement units (4, 5); In the object tracking unit (17), a malfunction of the inertial measurement unit of the motion estimation, which causes an erroneous prediction of the calculated position of the object around the vehicle (22), is detected among the motion estimations calculated by the Kalman filters (13, 14) arranged downstream of each of the two inertial measurement units (4, 5). A method comprising:

2. characterised in that an error state is transmitted to the object tracking unit (17) from a monitoring unit (12) to which the sensor values ​​of the inertial measurement units (4, 5) are supplied, and in the event of a fault (F) several hypotheses regarding the movement state (BZ A, BZ B) can be checked, whereas in the event of normal operation one movement state (BZ A, BZ B) is used, The method of claim 1.

3. the proper motion of the vehicle (22) is described by a vehicle pose (u, u') and transmitted to the object tracking unit (17), The method according to claim 1 or 2.

4. in said object tracking unit (17), a new object position (25, 26) of one object or new object positions (25, 26) of several objects are predicted in vehicle fixed coordinates from said calculated hypothesis on the vehicle pose (u, u') and a previously calculated object position (25, 26) of one object or from previously calculated object positions (25, 26) of several objects, The method according to claim 3.

5. characterised in that object positions (25, 26) of one or more objects measured by surrounding sensors of the vehicle (22) are compared with object positions (25, 26) predicted from proper motion hypotheses of the vehicle (22) determined using the Kalman filter (13, 14), and proper motion hypotheses in which the deviation between the predicted object positions (25, 26) and the measured object positions (25, 26) exceeds a threshold value are detected as incorrect.

5. The method according to any one of claims 1 to 4.

6. characterised in that if a fault (F) occurs in one of the inertial measurement units (4, 5), said fault (F) is detected by comparison with a redundant sensor of the other inertial measurement unit (4, 5) and a proper motion hypothesis calculated on the basis of an inaccurate sensor value is rejected as inaccurate, a switchover to the other inertial measurement unit (4, 5) is performed.

6. The method according to any one of claims 1 to 5.

7. the monitoring results of the two inertial measurement units (4, 5) are fed into the fault detection of the object tracking unit (17), characterised in that a fault (F) is detected only if the deviation of the predicted object position (25, 26) for one of the hypotheses exceeds a threshold value but not for the other hypothesis, and at the same time an error is detected by a sensor comparison of an IMU fault detection unit (16), 7. The method according to any one of claims 1 to 6.

8. characterised in that the fault detection involves an object detected as stationary or a number of objects detected as stationary, and a fault (F) is detected for a hypothesis in which the predicted object position (25, 26) changes significantly, while for the other hypothesis the object position remains constant, 8. The method according to any one of claims 1 to 7.

Images