WO2015189204A1

WO2015189204A1 - Method and system for the improved detection and/or compensation of error values

Info

Publication number: WO2015189204A1
Application number: PCT/EP2015/062831
Authority: WO
Inventors: Nico Steinhardt
Original assignee: Continental Teves Ag & Co. Ohg
Priority date: 2014-06-11
Filing date: 2015-06-09
Publication date: 2015-12-17
Also published as: DE102014211180A1

Abstract

The invention relates to a method for the improved detection and/or compensation of error values, wherein measured values are captured by a sensor system (101, 103, 104, 201, 203, 204), wherein the measured values describe physical quantities, wherein the measured values are burdened with error values, which describe deviations of the measured values from the described physical quantities, wherein the error values are detected and/or compensated by means of a comparison and wherein measured values that exceed a limit value are not used to detect and/or compensate error values of other measured values. The invention further relates to a corresponding system and to a use of the system.

Description

Method and system for improved detection and / or compensation of error values

The invention relates to a method for improved detection and / or compensation of error values according to the preamble of claim 1, a system for improved detection and / or compensation of error values according to the preamble of claim 14 and a use of the system.

All measurement data are subject to errors and in many cases there is no consistent availability of the measurement data. In addition to the dependence of the measured data on sensor-inherent properties, the measured data are often also dependent on ambient conditions. Sensor errors or measurement errors can be in quasi-stationary, constant over several measurements shares, such as. a so-called offset, and statistical, random from measurement to measurement shares, such. Noise, subdivide. While the random parts are in principle not deterministically correctable, quasi-stationary errors can generally be given

Correct observability. Uncorrectable significant errors can usually be avoided, given the detectability.

In the prior art sensor fusion methods are already known in this context, which are usually also suitable for correcting or filtering measurement data from different sensors or sensor systems. In the automotive sector in particular, special requirements have to be taken into account, since a multiplicity of different sensors detect a common environment situation or a motor vehicle state by means of different measurement principles and describe this environment situation or this motor vehicle state by means of a multiplicity of different measurement data. For a sensor fusion applicable in the automotive sector, the greatest possible robustness against incidental disturbances and identification and compensation of systematic errors is required. Likewise, temporal influences on the measured data must be corrected and temporary failures or the unavailability of sensors bridged. DE 10 2012 216 215 A1 describes a sensor system which comprises a plurality of sensor elements and a signal processing device. The signal processing device is designed so that it evaluates the sensor signals of the sensor elements at least partially together. Furthermore, the signal processing device is designed such that the

Measurement data of physical quantities each time information is assigned, which includes information about the time of each measurement directly or indirectly, wherein the signal processing device takes into account this time information at least in the generation of a fusion data set in a fusion ^¬ filter. For the generation of the fusion data set, measurement data are used which either have matching time information or, if no measurement data with matching time information is available, a corresponding measured value with the required time information is created by means of interpolation. Furthermore, the fusion filter assumes that error values of the measured data change only negligibly over a defined period of time. DE 10 2012 219 478 A1 discloses a sensor system for the independent evaluation of the integrity of its data. The sensor system is preferably used in motor vehicles and comprises a plurality of sensor elements which are designed such that they cover at least partially different primary measurement parameters, or at least partially utilize different Messprin ^¬ ciples. The sensor system further comprises a signal processing device which evaluates the sensor signals at least partially jointly and simultaneously assessing the Informa tion ^¬ quality of the sensor signals. The Signalverar- beitungseinrichtung also provides information about the consistency of at least a date of a physical ^¬'s size, wherein the data of the physical quantity on the basis of the sensor signals is calculated of sensor elements, which either directly detect the physical quantity or from whose sensor signals the physical quantity can be calculated. The information about the consistency of the date is now calculated on the basis of directly or indirectly redundant sensor information.

From DE 10 2010 063 984 AI a sensor system comprising several sensor elements is known. The sensor elements are formed so that they cover at least partially different primary measured variables and at least partially under use ^¬ schiedliche measuring principles. From the primary measured variable of the sensor elements, at least in part further measured quantities are derived. Furthermore, the sensor system comprises a signal processing device, an interface device and a plurality of functional devices. The sensor elements and all functional devices are connected to the signal ^¬ processing device. The primary measured variables thus provide redundant information that can be compared with one another in the signal processing device or can support one another. From the comparison of observables calculated in different ways, conclusions can be drawn about the reliability and accuracy of the observables. The signal processing device qualifies the accuracy of the observables and makes the observables, together with an accuracy specification, available to various functional devices via an interface ^device .

The well-known in the art, generic methods and sensor systems, however, are disadvantageous fraught than they algorithms slowly increasing failure of certain sensor systems, such as global navigation satellite systems, or only under comparatively great effort by computational effort Al ^¬ such as the so-called. Multiple Model Adaptive Estimation (MMAE) algorithm, as such. However, the non-recognition of these slowly increasing errors inevitably leads to corresponding deviations or disturbances in the functions which use the sensor data. It is therefore an object of the invention to propose an improved method for the detection or compensation of error values. This object is achieved by the method for improved detection and / or compensation of error values according to claim 1.

The invention relates to a method for improved recognition and / or compensation of error values, wherein measured values are detected by a sensor system, wherein the measured values describe physical quantities, wherein the measured values are subject to Def ^¬ lerwerten describing deviations of the measured values of the above physical quantities , wherein the error values detected by means of a comparison and / or com ^¬ compensated, and wherein a limit-border measured values are not used in order to detect errors values of other measured values and / or compensate. The invention therefore offers the advantage that - in particular in a sensor data fusion system or. are detected in accordance with the prior art only with difficulty or not at all detectable fault types on the one hand and on the other hand ^¬ not used to detect errors values of other measured values or to compensate for - in a sensor data fusion. On the one hand, these types of errors are so-called "creeping faults", whereby two different types of creeping faults must be distinguished: 1. creeping faults which can be calculated and compensated for by a system model, eg offset errors, scale factor errors (in particular of an inertial navigation system) ) and tires ^¬ semi-diameter error. These errors (if they are observable) almost no influence on a fusion record, but rather are merely an indication that the corresponding

Measured values are not suitable for detecting or correcting other measured values. 2. creeping faults that can not be compensated and thus directly have a comparatively considerable negative influence on a fusion data set, eg ionospheric influences in so-called Pseudorange measurements of a satellite navigation system.

Simultaneous errors are almost unavoidable when using, for example, a satellite navigation system and an odometry navigation system in a motor vehicle because there are always several simultaneous errors due to multipath propagation of satellite navigation signals or odometry disturbances due to uneven The probability that several such errors occur simultaneously and that their effects are consistent with each other is comparatively low, so that this case is negligible in real use. since the plausibility check rejects such faulty measurements, these are no longer used for the known per se integrity assessment. Only the case of false detections by too few verifiable redundant Mes ^¬ sun This is not covered by the plausibility check, which, however, does not contradict the assumption of a single error per measure epoch on which integrity is based.

The limit can z. B. the specifications resp. a data sheet of the sensor system are taken.

The measured values describe the physical quantities by representing values of the physical quantities.

Faulty measured values which exceed the error ^limit are preferably discarded.

By the measured values which are associated with such recognized error values or with error values exceeding the limit value are not used for the detection or correction of other measured values, is advantageously prevented that error detection and Fehl ^¬ corrections done on substantially error-free readings.

According to a preferred embodiment of the invention, it is provided that the comparison is a comparison with measured values of at least one further sensor system, wherein the measured values and the measured values of the at least one further sensor system directly or indirectly describe identical physical quantities and wherein error values detected by the comparison are determined by means of increments be continuously compensated. Thus, the invention enables the use of so-called AIME (Autonomous Integrity Monitoring by Extrapolation) and RAIM (Receiver Autonomous Integrity Monitoring) algorithms for error detection and error compensation.

A weakness of RAIM and AIME known in the art is the non-detectability of creeping faults. In current practice, these algorithms are therefore used only to a limited extent, instead, with respect to processing time and storage requirements consuming procedures such as the

MMAE algorithm used. A disadvantage of these methods, however, is that only modeled errors are reliably detected. A typical case of creeping errors is, for example, the time-varying disturbance in a pseudorange measurement caused by changes in the ionosphere. This takes place sufficiently slowly so that the estimated position of the fusion filter Corridor ^¬ is yaws according wrong, the amount of deviation from one measurement epoch to the next, however, not sufficient to detect an error with a so-called. SNAPS hot process. In the case of the fusion filter, creeping errors are expected for the sensors used in the following cases:

• Inertial navigation system: Due to defects or external influences, such as the ambient temperature, caused drift of offset or scale factor error.

• Satellite navigation system: pseudo-orange measurements due to ionospheric influence and multipath reception, whereas delta-range measurements are not affected by the temporal differentiation of the measured values. • Odometrienavigationssystem: Slow variable error of the measured velocity by changes of the rolling radius, for example by temperature creep ^¬ sponding pressure loss or alteration of the of the tires.

The sizes of an inertial navigation system and potentially those affected by creeping errors

Odometrial navigation systems are usually already implemented as an error model in the fusion filter, and in each measurement epoch, the raw measurements are preferably corrected by the known, continuously further estimated errors. Thus, the slow growth of these errors does not lead to significant errors in the fused data as long as the fusion filter corrects them sufficiently quickly by redundant measurements. Problematic as much about the dynamics filter out ^¬ continuous interference of the measurement data in question are against a defined detection probability and thresh recognizable both preventable by a plausibility check and by the. In so-called NIS test. Furthermore, by checking the summed absolute values of the error corrections with defined maximum values, a detection of measured values outside their specified error range can be realized.

Creeping faults of individual pseudo-orange measurements, which are not modeled as faults in the fusion filter, lead to contradictions in the one described by the above-mentioned independent of the fusion filter.

Plausibilisierung carried out geometric comparison of pseudoranges with each other, and are therefore recognizable with a defined detection threshold.

The physical variables described indirectly are preferably calculated from the measured values which describe them and from known physical or mathematical relationships. According to a further preferred embodiment of the invention, it is provided that only measured values with an identical time stamp are used for the comparison of the measured values. This has the advantage that - at least in error-free Measurement - match all subject to the comparison values because they write the same physical size at the same time be ^¬. This simplifies the detection of erroneous values, since the cause of a mismatch in this case must necessarily be an erroneous value.

Particularly preferably, it is provided that the values subjected to the comparison are generated by means of interpolation, if there are no values recorded with an identical time stamp. Since the measured data are usually acquired at different times due to different signal output delays and generally due to lack of synchronization between the sensor systems, or are output by the sensor systems at different times, the required values can be calculated by means of the interpolation. Preferably, measurement data or values of the sensor system with the lowest signal output delay are generated by means of interpolation, ie, that these measurement data or values are generated as a function of the acquisition times of the measurement data or values of the other sensor systems. For the generation of a value by means of interpolation, it is expedient to use the two values of the sensor system with the lowest signal output delay closest in time and enclosing them to be generated. The value generated by means of interpolation is then subjected to the comparison as described. The sensor system with the lowest signal output delay is most preferably the basic sensor system. Furthermore, it is particularly preferred that changes in the values for the interpolation be assumed to be proportional to the time. So therefore are used for linear INTERPO ^¬ lation. This results in the advantage that the interpolation is relatively simple and correspondingly executable with only a small amount of computation.

According to a further preferred embodiment of the invention, it is provided that the comparison is a comparison of Measured values of one and the same sensor system with each other taking into account a sensor-specific stochastic model. This results in the advantage that existing redundancies can be optimally utilized optimally, taking into account the sensor-specific stochastic model. The sensor-specific stochastic model is used to compare the measured values from the past with the current measured values. If, for example, an unexpectedly large deviation is detected here or a limit value is exceeded, the corresponding measured value can be recognized as defective or can not be used to detect and correct error values of other measured values.

According to a further preferred embodiment of the invention, it is provided that for the comparison taking into account the sensor-specific stochastic model only measurement values are used, have different time stamps on ^¬. Thus, the current measured values can be comparatively easily compared against a limit, which was formed from the measured values of the past. Such a procedure is also known as a so-called snapshot method. The snapshot method is essentially the well-known NIS test, but without the consideration of covariances. Thus, the snapshot method is comparatively less computationally expensive and favors real-time detection of error values.

According to a further preferred embodiment of the invention, it is provided that it is assumed that after carrying out the comparison of the measured values of one and the same

Sensor system each detection cycle more than a single measured value exceeds the limit. This be ^¬ günstigt the use of the aforementioned AIME- and

RAIM algorithms even further.

In accordance with a further preferred embodiment of the invention, it is provided that the sensor system and the at least one further sensor system are an inertial navigation system global satellite navigation system and / or an odometry navigation system. Thus, the present invention is particularly suitable for navigation purposes and for navigation systems, preferably in motor vehicles. The sensor systems thus determine the position, in particular the position of a motor vehicle, from the output data. The global navigation satellite system may be, for example, a so-called GPS navigation system. The

The odometry navigation system initially determines the speed, for example, via the known rolling circumference of the motor vehicle tires and thus makes it possible to determine the position while taking into account the steering angle in the context of dead reckoning. It is particularly expedient that the Satellitenna ^¬ vigationssystem least two satellite signal receiver comprises is. This improves the quality of the recorded

Satellite signals and thus the reliability and accuracy of the satellite navigation system.

In addition, it is particularly preferable that the inertial navigation system is the sensor base system. The inertial navigation system as a sensor base system has the advantage that it has the comparatively highest availability, since it has a comparatively high output rate of the acquired input data and, moreover, operates largely independently of external interference influences.

According to a particularly preferred embodiment of the invention, it is provided that the limit value of the inertial navigation system is a limit value for zero error and / or

Scale factor error is.

According to another particularly preferred embodiment of the invention, it is provided that the limit of the Satelli ^¬ tennavigationssystems is a limit value for frequency error of an oscillation frequency of a receiver clock oscillator.

According to a further particularly preferred embodiment of the invention, it is provided that the limit value of Odometer navigation system is a limit for radius error of Abrollradius of vehicle wheels.

According to a further preferred embodiment of the invention, it is provided that the detected by comparing error values by means of an error-state-space filter, in particular ^¬ sondere are detected by means of an error-state-space Kalman filter and / or compensated. The error-state space filter represents a fusion filter for the fusion of the output data or navigation information, in particular for the fusion of normally distributed output data or navigation information. At the same time, the error state space filter preferably estimates or determines the error values of at least the sensor base system. By means of the at least one correction system, the error values and possibly also unknown variables of the inertial navigation system can then be estimated or determined. A specifics ^¬ derheit of the error-state-space filter, then, is that error values are incrementally estimated or determined instead of the sensor signals or the input data only and are subsequently corrected. The error values namely have a signi ficantly lower ^¬ temporal dynamics than the output data itself, whereby a significant decoupling of the dynamics of the error-state-space filter on the characteristics of Sensorba ^¬ sissystems or the at least one correction system

is reached.

According to a further preferred embodiment of the invention, it is provided that the measured values represent navigation information. This enables the use of the invention for navigation purposes, in particular for future applications in motor vehicles, which require a highly accurate localization capability.

According to a further preferred embodiment of the invention, it is provided that the measured values are fused to a fusion data set. A common merger data set is usually more reliable and more precise than the individual measured values and, in particular, it allows using a fault Estimate a comparatively reliable assessment of the accuracy or reliability of the fused measurements or physical quantities. A fusion filter, which generates the fusion data set, thereby corrects those error values or error values which correspond to the stochastic model of the filter within given tolerances. Measured values whose error values are outside the model, however, are preferably rejected.

The invention further relates to a system for improved detection and / or compensation of fault values, comprising a sensor base system, at least one further sensor system and a fusion filter, wherein the sensor base system and the at least one further sensor system are configured to acquire measured values, the measured values being physical variables wherein the measured values are subject to error values which describe deviations of the measured values from the described physical quantities, wherein the system is designed to detect and / or compensate for error values by means of a comparison and wherein the fusion filter is designed to exceed a limit value not to approach ^¬ pull measurements to detect errors values of other measurements and / or compensate. The system according to the invention thus comprises all devices necessary for carrying out the method according to the invention.

It is preferably provided that the system is designed to carry out the method according to the invention.

Moreover, the invention relates to a use of the system according to the invention in a motor vehicle.

Further preferred embodiments will become apparent from the subclaims and the following description of an embodiment with reference to figures.

Show it Fig. 1 by way of example a possible embodiment of a system according to the invention, which is designed for position ^¬ determination, in a motor vehicle and

2 shows by way of example a further possible embodiment of a system according to the invention, which is likewise designed for position determination, in a motor vehicle. Fig. 1 shows a schematic representation of an embodiment of the system according to the invention, which is provided for arrangement and use in a motor vehicle (not shown). The illustrated system is designed according to the example for improved detection or compensation of error values and is suitable for determining the position of the motor vehicle. In this case, all elements or components or sensor systems comprised by the system are illustrated as functional blocks and their interaction with each other is shown. The navigation system includes inertial navigation system 101 configured to sense at least the accelerations along first, second, and third axes and at least the yaw rates about the first, second, and third axes. The first axis corresponds, for example according to the longitudinal axis of the motor vehicle, the second axis corresponds to the transverse axis of the motor vehicle and the third axis corresponds to the vertical axis of the motor vehicle. These three axes form a Cartesian coordinate system, the so-called motor vehicle coordinate system.

Inertial Navigation System 101 forms, for example according to the so-called sensor-based system, whose measured values using the method described in the following further sensor systems, so-called. Korrektursys ^¬ systems can be corrected. The correction systems are Odometrienavigationssystem 103 and Satellitennavigationssys ^¬ tems 104th

The system according to the invention also has a so-called. Strapdown algorithm unit 102, in which a so-called. Strapdown algorithm is performed by means of which the measured values of inertial navigation system 101, among other things, are converted into position data. These are the readings of inertial navigation system 101, which naturally describe Accelerat ^¬ fixing certificates, integrated twice over time. By means of a simple integration over time, the orientation and the speed of the motor vehicle are further determined. In addition, strapdown algorithm unit 102 compensates for a Coriolis force acting on inertial navigation system 101.

The output data from strapdown algorithm unit 102 to ^¬ therefore consider the following physical quantities:

the speed, the acceleration and the rate of rotation of the motor vehicle, according to the example of said three

Axes of the motor vehicle coordinate system and, for example, additionally in each case based on a world coordinate system, which is suitable for describing the orientation or of dynamic variables of the motor vehicle in the world. By way of example, the named world coordinate system is a

GPS coordinate system. In addition, the output data from the strapdown algorithm unit 102. ^¬ clearly includes the position with respect to the vehicle coordinate system and the orientation of the world coordinate system. In addition, the output data from the strapdown algorithm unit 102 have the Va ^¬ rianzen to as an information on the quality of the above navigation information. These variances are such as not be ^¬ expects strap-down algorithm unit 102, but only used by this and forwarded. The values calculated by the strapdown algorithm unit 102 og Navi ^¬ gationsinformationen be output via output module 112 and provided to other vehicle systems.

The system according to the invention also comprises

Odometry navigation system 103 in the form of wheel speed sensors for each wheel of the motor vehicle. By way of example, it is a four-wheeled motor vehicle with four wheel speed sensors, each of the speed of their associated wheel and its Detect direction of rotation. Furthermore includes

Odometrienavigationssystem 103 a steering angle sensor element that detects the steering angle of the motor vehicle. In addition, the exemplary system shown

Satellite navigation system 104, which is designed so that it determines the distance between each associated satellite and the vehicle and the speed between the associated satellite and the vehicle.

The system also includes fusion filter 105. Fusion filter 105 provides a fusion data set 106 as the odometry navigation system 103, satellite navigation system 104, and inertial navigation system 101 collectively evaluate the measurements. Fusion data set 106 has the acquired measured values of the different sensor systems, wherein fusion data record 106 includes, for example, additional error values and variances associated with the error values which describe the data quality.

The measured values of inertial navigation system 101 are stored during operation of the motor vehicle in a dedicated electronic data memory 113 of fusion filter 105 for a predetermined period of time. Inertial navigation system 101 represents the so-called sensor base system, while odometry navigation system 103 and satellite navigation system 104 represent the so-called correction systems whose measured values are used to correct the measured values of the sensor base system. This ensures that measured values that were at least apparently acquired at an identical point in time can always be subjected to comparison.

For example, fusion data set 106 provided by fusion filter 105 includes that which has been checked for plausibility

Output data of the correction systems certain quantitative errors of the sensor system. Strapdown algorithm unit 102 is now corrected by Fu ^¬ sion data record 106, the measured values of the sensor base system.

Fusion data set 106 is calculated by fusion filter 105 from the measurements of odometry navigation system 103, satellite navigation system 104, and inertial navigation system 101.

By way of example, fusion filter 105 is embodied as an Errator-State-Space Kalman filter, that is to say as a Kalman filter, which in particular carries out a linearization of the measured values and in which the quantitative error values of the measured values are calculated or estimated and which se ^¬ quentiell operates and thereby corrects the available in each functi ^¬ onsschritt the sequence of output data.

Fusion filter 105 is designed to always be asynchronous to the most recent ones of inertial navigation system 101,

Odometry navigation system 103 and satellite navigation system 104 detected. By way of example, the measured values are guided via the motor vehicle model unit 107 and the ^{outage model unit} 109.

Motor vehicle model unit 107 is configured to calculate from the measured values of odometry navigation system 103 at least the speed along a first axis, the speed along a second axis and the rate of rotation about a third axis and provide these fusion filters 105.

The exemplary system also includes tire parameter estimation unit 110 configured to include at least the radius, for example, the dynamic one

Radius, calculated all wheels and additionally calculated the slip stiffness and the slip stiffness of all wheels and this motor vehicle model unit 107 provides as additional input variables. Tire parameter estimation unit 110 is further configured to use a substantially linear tire model to calculate tire sizes. The example according to inputs from Reifenparameterschät- wetting unit 110 are the wheel speeds and the steering angle describing input data, at least partially, from the output values of ^¬ strapdown algorithm unit 102 as well as the particular fusion filter 105 variances.

The exemplary system also includes

GPS error detection and plausibility unit 111, which is configured such that it receives as input data, for example, the output data from satellite navigation system 104 as well as at least partial output data from

Strapdown algorithm unit 102 receives and considered in their Be ^¬ calculations.

GPS error detection and validation unit 111 checks the output data against a stochastic model adapted to satellite navigation system 104. If the measured values correspond to the model within the framework of a tolerance that takes account of the noise, they are checked for plausibility. GPS error detection and plausibility check unit 111 is additionally connected to data-level fusion filter 105 and transmits the plausibility measured values to fusion filter 105. GPS error detection and plausibility check unit 111 is exemplified to provide a method for selecting a satellite, and the like. by means of the following method steps:

Measuring position data of the motor vehicle relative to the satellite based on the sensor signals of satellite navigation system 104,

- determining redundant to the determined based on the sensor signals from the satellite navigation system 104 Positionsda ^¬ th reference position data of the motor vehicle, - selecting the satellite when a comparison of the position data and the reference position data of a predetermined condition is satisfied,

- whereby the position data and the comparison reference position data, a difference is formed between the position data and the reference position data,

wherein the predetermined condition is a maximum permissible deviation of the position data from the reference position data,

- the maximum allowable deviation from a standard deviation depends ^¬, which is calculated based on a sum of a reference variance for the reference position data and a measurement variance for the position data, and

- Where the maximum allowable deviation corresponds to a multiple of the standard deviation such that a probability that the position data fall in a dependent of the Standardab ^¬ scattering interval falls below a predetermined threshold.

The exemplary system also has

Standstill detection unit 108, which is designed so that it can detect a stoppage of the motor vehicle and in case of a detected stoppage of the motor vehicle at least fusion filter 105 information from a

Stability model provides. The information from a standstill model describes that the rotation rates around all three axes have the value zero and the velocities along all three axes have the value zero.

In this case, the standstill detection unit 108 is designed in such a way that it uses the measured values of the wheel speed sensors of the odometry navigation system 103 as well as the measured values of the inertial navigation system 101 as input data. The example contemporary system used, for example in accordance with a first set of measurement values which relate to a Kraftfahrzeugkoordi ^¬ natensystem and in addition a second set of measurement values relating to a world coordinate system, the world coordinate system is used to describe the orientation and dynamic variables of the motor vehicle. By means of alignment model unit 109 is determined ^¬ an alignment angle between the motor vehicle coordinate system and the world coordinate system. The particular orientation of the model unit 109, Reg ^¬ tung angle between the motor vehicle coordinate system and the world coordinate system is determined on the basis of the following physical quantities:

- the vectorial speed with respect to the

World coordinate system,

- the vectorial speed with respect to the

Motor vehicle coordinate system,

- the steering angle and

- the respective quantitative error of the said

Size descriptive output data.

Alignment model unit 109 uses all of the output data from strapdown algorithm unit 102.

Alignment model unit 109 is, for example according formed so that they have an in ^¬ formation calculated in addition to the orientation angle on the quality of the orientation angle in the form of a variance and provides fusion filter 105th

Fusion filter 105 uses the orientation angle and the variance of the orientation angle in its calculations, the results of which it provides via fusion data set 106

Strapdown algorithm unit 102 forwards.

Fusion filter 105 thus detects the measured values of inertial navigation system 101, the sensor base system, as well as of

Odometry navigation system 103 and satellite navigation system 104, the correction systems.

Since it is not possible to detect sensor errors that signify a departure from the specification with the increments usually applied, the method according to the invention is used. This means that measured values for which a limit value is known are used to detect errors in the sensor systems. For this purpose, the absolute size, in which the corrections output by fusion filter 105 are summed, must be checked against a limit value. Especially creeping error, the dynamics of which is below the dynamics of fusion filter 105 and the result are thus mutually offset by Fu ^¬ sion filter 105, which are hereby detectable. In particular, errors of the inertial navigation system 101 are thus detectable by comparison with known absolute errors specified in the data sheet with the maximum values for zero point and scale factor errors permissible according to the specification of the inertial navigation system 101.

Errors from satellite navigation system 104 are detectable by checking the summed receiver clock error. In particular, the drift of the receiver clock error makes it possible to check the frequency error of the receiver clock oscillator for compliance with the specification.

By means of the summation corrections of the wheel-roll radii, by comparison with the nominal values both defects and defects of the tires and sensor systems as well as special situations such as, for example, the mounting of an emergency wheel can be detected.

FIG. 2 shows by way of example a further possible embodiment of a system according to the invention, which method is designed for improved recognition or compensation of error values in a motor vehicle (not shown). The system includes, for example, inertial navigation system 201, satellite navigation system 204, and odometry navigation system 203 as different sensor systems. Trägheitsnavigati ^¬ onssystem 201, GPS 204, and

Odometry navigation system 203 outputs measured values which directly and indirectly describe physical variables, namely a position, a velocity, an acceleration, an orientation, a yaw rate or a yaw acceleration, to fusion filter 205. The output of the measured values takes place via a vehicle data bus, for example via a so-called CAN bus. For example, satellite navigation system 204 outputs its raw data readings. As a central element in a position determination of

Motor vehicle inertial navigation system 201, which is a so-called MEMS IMU

(Micro-Electro-Mechanical System Inertial Measurement Unit) is in combination with strapdown algorithm unit 207 used, since this is assumed to be error-free, that is, it is assumed that the measured values of Trägheitsnavigati ^¬ onssystem 201 always their stochastic model correspond to the fact that they have only noise influences and thus free of external or random errors or. Disturbances are. The noise and remaining un-modeled errors of inertia ^¬ navigation system 201, such as non-linearity, are thereby (so-called. Gaussian white noise) over the measuring range as a zero-mean, stationary, and normally distributed adopted.

Inertial navigation system 201 comprises three yaw rate sensors which mutually register orthogonally and three acceleration sensors which detect each other orthogonally in each case.

Satellite navigation system 204 includes a GPS receiver, which initially carries out over the satellite signal propagation time Entfer ^¬ voltage measurements to the receivable GPS satellites and also from the change of the satellite signal transit time as well as additionally from the change in the number of wavelengths of the satellite signals, a distance traveled by the vehicle path ^¬ distance determined. Odometrienavigationssystem 203 includes depending ^¬ weils a wheel speed sensor at each wheel of the motor vehicle, and a Lenwinkelsensor. The wheel speed sensors each determine the Raddrehgschwindigkeit their associated wheel and the steering angle sensor determines the chosen steering angle.

Inertial navigation system 201 outputs its measured values to preprocessing unit 206 of inertial navigation system 201. Pre-processing unit 206 now corrects the measured values or the physical variables described therein by means of corrections, which pre-processing unit 206 receives from fusion filter 205. The thus corrected measured values or in it described physical quantities are continued on strapdown algorithm unit 207.

Strapdown algorithm unit 207 now makes a position determination on the basis of the corrected measured values or the physical quantities of preprocessing unit 206. This Po ^¬ sitionsbestimmung is a so-called. Dead reckoning based on inertial navigation system 201. For this, the output from preprocessing unit 206 corrected measured values or the physical quantities are continuously integrated over time and added up. Strapdown algorithm unit 207 further compensates for a Coriolis force acting on inertial navigation system 201, which may affect the measurements of inertial navigation system 201. For position determination, strapdown algorithm unit 207 performs a two-fold integration over time of input data acquired by inertial navigation system 201 describing accelerations. This allows an updating of a previously known position as well as an updating of a previously known orientation of the motor vehicle. For determining a speed of the motor vehicle, strapdown algorithm unit 207 performs a simple integration of the input data acquired by inertial navigation system 201 over time. Further corrected

Strapdown algorithm unit 207 and the particular position by means of corresponding correction values of fusion filter 205. Fusion filter 205 performs in this example, the correction only indirectly via strapdown algorithm unit 207 from. The measured values determined by strapdown algorithm unit 207 or the physical variables, ie the position, the speed, the acceleration, the orientation, the rotation rate and the rotational acceleration of the motor vehicle, are now fed to output module 212 and to fusion filter 205. The so-called strapdown algorithm unit 207 performs.

Strapdown algorithm is computationally only slightly complex and can therefore be realized as a real-time capable sensor base system. He introduces a procedure for integration The inertial navigation system 201 input data on velocity, orientation and position and does not include filtering, resulting in an approximately constant latency and group delay.

The term sensor base system describes the sensor system whose measured values or. physical values are corrected by means of the measured values or the physical quantities of the other sensor systems, the so-called correction systems. By way of example, as discussed above, the correction systems are odometry navigation system 203 and satellite navigation system 204.

Inertial navigation system 201, preprocessing unit 206 of inertial navigation system 201 and

Strapdown algorithm unit 207 form, for example, to ^{¬ together} the so-called. Sensor base system, to which additionally pro rata and fusion filter 205 is counted. Output module 212 relays the measured values or physical quantities determined and corrected by strapdown algorithm unit 207 to any other systems of the motor vehicle.

The measured values acquired by the satellite navigation system 204 are, for example, in the form of sensor signals via a so-called UART data connection, first forwarded to preprocessing unit 208 of satellite navigation system 204. Pre-processing unit 208 then determines from the output from satnav ^¬ gationssystem 204 measured values, which represent raw GPS data, and also include a description of the orbit of each GPS signal transmitting GPS satellite, a position and a speed of the motor vehicle

GPS coordinate system. In addition, satellite navigation system 204 determines a relative speed of the motor vehicle to the GPS satellites from which GPS signals are received. Furthermore, preprocessing unit 208 corrects a time error contained in the measured values or the physical quantities of a receiver clock of satellite navigation system 204, which caused by a drift of the receiver clock, as well as by means of a correction model, the changes in the signal transit time and the signal path caused by atmospheric influences on the GPS signals transmitted by the GPS satellites. The correction of the time error as well as the atmospheric influences are carried out by means of correction filters obtained by means of Fusionsfilter 205 via the CAN bus.

Satellite navigation system 204 is still approximately feasibility check module 209 associated with which plausibility output by preprocessing unit 208 ^¬ physical quantities, that is, the position and the speed of the motor vehicle, lisiert. The plausibility values or physical variables plausibilized by plausibility module 209 are then output to fusion filter 205.

The system further comprises preprocessing unit 210 of odometry navigation system 203, which receives the measured values acquired by odometry navigation system 203 via the CAN bus. The recorded measured values in this case are the output data of the individual wheel speed sensors as well as the input data of the steering angle sensor. Preprocessing unit 210 now determines the position and orientation of the motor vehicle in the motor vehicle coordinate system from the measured values output by odometry navigation system 203 in accordance with a so-called dead reckoning method. Furthermore, the speed Accelerat ^¬ nigung, the rate of rotation and the rotational acceleration of the motor ^¬ vehicle are determined, also in the vehicle coordinates ^¬ system. In addition preprocessing unit 210 corrects the measured values obtained by Odometrienavigationssystem 203 by means obtained from Fusion filter 205 correction values.

Odometry navigation system 203 is further associated with plausibility module 211, which plausibilizes the measured values output by preprocessing unit 210, that is to say the position, the orientation, the speed, the acceleration, the rotation rate and the rotational acceleration of the motor vehicle. Since the error values of the measured values of In the case of comparatively large wheel slippage, the measured values determined by means of the inertial navigation system 201 and the satellite navigation system 204 are used to make the measured values of the odometry navigation system 203 plausible. First of all, however, the measured values or the physical quantities are also calibrated against a sensor-individual model assigned to them which takes into account measurement uncertainties such as noise influences. If the measured values or the physical quantities correspond to the model within the given limit values or tolerance ranges, a first occurs here

Plausibility checks and the plausibility of measured values or physical quantities are further processed. The plausible measured values or physical quantities are then forwarded to fusion filter 205. If one

Plausibility of these measured values or physical quantities can not be done, the corresponding measured values or physical quantities are discarded and not further processed.

Fusion filter 205 is embodied, for example, as an Eror State Space Kalman filter. By way of example, the main task of fusion filter 205 is to correct or to measure the measured values or the physical quantities of the sensor base system, ie of inertial navigation system 201, by means of the measured values or the physical variables of odometry navigation system 203 and satellite navigation system 204, which represent the correction systems to output corresponding correction values to strapdown algorithm unit 207. For example, since inertial navigation system 201 is assumed to be free from random errors and external disturbances, the measurements of inertial navigation system 201 are exclusively white noise. Since it is in merger filter 205 a so-called. ER- ror state-space Kalman filter, the quantitative error values of the measured values or the physi ^¬-earth values are exclusively determined and corresponding corrections off guided. This simplifies and accelerates the fusion of the measured values or the physical quantities of inertial navigation system 201 made by fusion filter 205,

Odometry navigation system 203 and satellite navigation system 204 to a common fusion data set. Thus, a real-time capable position determination and correction of the position determination is possible.

The system shown in FIG. 2 represents a so-called virtual sensor, wherein inertial navigation system 201,

Odometry navigation system 203 and satellite navigation system 204, however, are not components of the virtual sensor. A virtual sensor is a system which always generates the same output data or outputs regardless of the type of sensor systems involved - in this case inertial navigation system 201, odometry navigation system 203 and satellite navigation system 204. On the basis of the output data or outputs is not clear what sensor systems are powered ^¬ connected into the system.

Thus, the system described by way of example receives correction ^¬ measurements of Odometrienavigationssystem 203 and satellite navigation system 204. Fusion filter 205 is, for example expanded according to a two-stage signal plausibility for the detection of error values in the measured values of the correction measurements. This detected the one hand deviations Träg ^¬ integrated navigation system 201 and to the strapdown algorithm, and on the other hand deviations from the individual measured values, eg each other contradictory pseudorange measurement values within a data set with each other. Fusion filter 205 is also an integrity score according to the known one

"Normalized Innovation Squared" test and extended to calculate a protection level according to the well-known RAIM or AIME algorithm as a snapshot method In order to compensate for the disadvantages usually associated therewith, namely the inability to detect creeping errors and impossibilities, to handle several errors occurring at the same time are discarded by means of the plausibility check of all measured values according to the invention, which exceed a predetermined limit, so that it can finally be assumed that there is at most one more faulty measured value per fair epoch. Furthermore, the absolute values of the accumulated error corrections are checked against a defined limit value and, if necessary, the corresponding measured values are rejected when the limit value is exceeded.

Claims

claims

A method for improved detection and / or compensation of error values, wherein measured values are detected by a sensor system (101, 103, 104, 201, 203, 204), the measured values describing physical quantities, wherein the measured values are subject to error values Describe deviations of the measured values from the described physical quantities and wherein the error values are detected and / or compensated by means of a comparison,

characterized in that measured values exceeding a limit value are not used to detect and / or compensate for error values of other measured values.

2. The method according to claim 1,

characterized in that the comparison is a comparison with measured values of at least one further sensor system (103, 104, 203, 204), the measured values and the measured values of the at least one further sensor system (103, 104, 203, 204) being directly or indirectly identical Describe quantities and wherein compensated by the comparison detected error values are continuously compensated by increments.

3. The method according to claim 2,

characterized in that only measured values with an identical time stamp are used for the comparison of the measured values.

4. The method according to claim 1,

characterized in that the comparison is a comparison of measured values of one and the same sensor system (101, 103, 104, 201, 203, 204) with one another taking into account a sensor-specific stochastic model.

5. The method according to claim 4,

characterized in that for the comparison taking into account the sensor-specific stochastic model only measured values are used which have different time stamps.

6. The method according to at least one of claims 4 and 5, characterized in that it is assumed that after execution of the comparison of the measured values of one and the same sensor system with each other per detection cycle a maximum of a single measured value exceeds the limit.

7. The method according to at least one of claims 1 to 6, characterized in that the sensor system (101, 201) and the at least one further sensor system (103, 104, 203, 204) an inertial navigation system (101, 201), a global navigation satellite system ( 104, 204) and / or a

Odometry navigation system (103, 203) are.

8. The method according to claim 7,

characterized in that the limit value of the inertial navigation system (101, 201) is a limit value for zero error and / or scale factor error.

9. The method according to at least one of claims 7 and 8, characterized in that the limit value of the Satellitenna ^¬ vigationssystems (104, 204) is a limit value for frequency error of an oscillation frequency of a receiver clock oscillator.

10. The method according to at least one of claims 7 to 9, characterized in that the limit value of

Odometry navigation system (103, 203) is a limit for Ra ^¬ diusfehler a Abrollradius of vehicle wheels.

11. The method according to at least one of claims 1 to 10, characterized in that the recognized by the comparison

Error values by means of an error-state space filter (105, 205), in particular by means of an error-state-space Kalman filter (105, 205), recognized and / or compensated.

12. A method according to any one of claims 1 and 11, characterized in that the measured values represent Navigationsinfor ^¬ mation.

13. The method according to at least one of claims 1 to 12, characterized in that the measured values are fused to a merger ^¬ record.

14. A system for improved detection and / or compensation of error values, comprising a sensor base system (101, 201), at least one further sensor system (103, 104, 203, 204) and a fusion filter (105, 205), wherein the sensor base system (101, 201) and the at least one further sensor system (103, 104, 203, 204) are designed to record measured values, wherein the measured values describe physical variables, wherein the measured values are subject to error values which describe deviations of the measured values from the described physical quantities and wherein the system is designed to recognize and / or compensate for error values by means of a comparison,

characterized in that the fusion filter (105, 205) is designed not to use measured values exceeding a limit value in order to detect and / or compensate for error values of other measured values.

15. System according to claim 14,

characterized in that the system is ^adapted to ^perform a method according to at least one of claims 1 to 13 from ^¬ .

16. Use of the system according to at least one of claims 14 and 15 in a motor vehicle.