EP2643661A1

EP2643661A1 - Method and device for determining motion parameters

Info

Publication number: EP2643661A1
Application number: EP11788097.1A
Authority: EP
Inventors: Guido Arnold
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 2010-11-26
Filing date: 2011-11-11
Publication date: 2013-10-02
Also published as: WO2012069324A1; DE102010062016A1; CN103299160A

Abstract

The invention relates to a method for determining motion parameters of moving objects, in particular for rail vehicles, and to a corresponding device. In order to be able to dispense with expensive gyroscopes, according to the invention the acceleration of the object along at least one of the three spatial axes (X, Y, Z) is measured by means of MEMS (micro-electromechanical systems) acceleration sensors and/or the rotational speed of the object about at least one of the three spatial axes (X, Y, Z) is measured by means of MEMS gyro sensors. For each digital or digitized rotational speed measured value (GyroY') in relation to a spatial axis (X) rotated through 90°, trailing and/or leading mean values (MwGXn and/or MwGXv) based on the associated standard deviation (StabwGXn and/or StabGXv) compared with a threshold value (SgX) are used for offset correction of the rotational speed measured value (GyroY').

Description

description

Method and device for determining motion parameters of moving objects

The invention relates to a method for determining movement parameters of moving objects, in particular for rail vehicles, as well as a device relating thereto.

Movement parameters such as acceleration and inclination of the object relative to at least one of the three spatial axes are used, for example, for locating or navigating the object. The use of gyro compasses is known, but they are extremely expensive and expensive, so that they can not be used in many areas. Significantly cheaper are MEMS (micro-electro-mechanical system) sensors, which are used for example as acceleration sensors for airbag controls. The problem, however, is their use as gyro sensors for determining the inclination of the object about an axis, since the inclination can only be determined cumulatively by the discrete measurement of rotation angles designated as rotation rates. The measured rotation rates are added continuously and must therefore be determined extremely accurately. However, functional reasons have MEMS considerable Off ^¬ set- and drift errors.

The invention has for its object to provide a method and apparatus for determining motion parameters, which allow high accuracy with little effort.

According to the method, the object is achieved in that along at least one of the three spatial axes by means of MEMS (Micro-Electro-Mechanical System) acceleration sensors acceleration and / or the rotational rates of the object are measured by at least one of the three spatial axes by means of MEMS gyro sensors, with trailing and / or leading average values for each digital or digitized rotational measured value with respect to a spatial axis rotated by 90 ° as a function of the associated, standard deviation used for offset correction of the rotation rate measured value.

The object is also achieved with a device for carrying out the method, wherein at least one MEMS acceleration sensor for determining the acceleration along at least one of the three spatial axes and / or at least one MEMS gyro sensor is provided for determining the rotation rates about at least one of the three spatial axes are /, whose digital or digitized measured values are supplied via a storage medium to a control device for calculating the movement parameters, wherein means for offset correction of the rotation rate measured by means of a rotated by 90 ° spatial axis trailing and leading average values and to ^¬ associated with a threshold to be compared Standard deviations are provided.

Only by this offset correction simple MEMS sensors for tilt calculation, for example, for rail vehicles, can be used.

While an object to be viewed is in motion befin ^¬ det are represented by the MEMS acceleration sensors, the acceleration, which is exposed to the object detected along one, two or all three spatial axes. Preferably, all the rotations of the object are likewise detected by one, two or all three spatial axes by the MEMS gyro sensors. For the measurement of accelerations and Rotary rates can be used both individual MEMS sensors and multi-axis MEMS sensors. Also MEMS Senso ^¬ ren that monitor all six degrees of freedom can be used. The simultaneity of data collection is indispensable to some extent. If the MEMS sensors have analog output signals, these are converted into digital output signals by means of A / D converters. The digital or digitized measured values of the accelerations and the rotation rates are stored, whereby it is always possible not to have to monitor all the axes.

According to claim 2 it is provided that trailing and vorlau ^¬ Fende standard deviations are compared with a threshold value for digital or digitized acceleration measurement values, wherein the acceleration measuring values are used to Neigungsbe ^¬ bill, if at least one of the two standard deviations below the threshold value, and ge ^¬ optionally one of a maximum permissible inclination corre ^¬ sponding maximum acceleration of the accelerometer falls below. In this way, the cumulative rotation rate measurement, in which a propagation of offset and drift influences according to claim 1, although minimized but can not be excluded, quasi bypassed. It used only acceleration measurements that are not cumulative ermit ^¬ telt. However, this preferable algorithm assumes that in the main movement direction no acceleration, but a constant speed movement takes place. As soon as an acceleration is detected, the rotation rate measured value with offset correction according to claim 1 must be immediately available again. Therefore, the simultaneous determination of the rotation rates and their storage is crucial. Are expected curves in the path of movement of the observed object, it is provided according to claim 3, that are used to identify a curve rotation rate measurement values to a vertical spatial axis and for calculating a kuvenbedingten bank of the object yaw rate measurements to a main direction of movement ^¬ processing of the object. In this way, influences of the bank in curves, for example due to centrifugal force to the outside or in the transverse direction beveled roadway inwards, can be determined and used to correct the inclination calculation.

The invention will be explained in more detail with reference to figurative representations. Show it:

FIG. 1 shows the essential components of a device according to the invention,

Figure 2 shows a first variant for the determination of

Motion parameters,

3 shows a second variant for the determination of the BEWE ^¬ supply parameter,

FIG. 4 shows an algorithm for offset correction in the inclination calculation;

FIG. 5 shows an algorithm for inclination calculation by means of

Acceleration measurements and

FIG. 6 shows an algorithm for tilt correction during cornering.

FIG. 1 shows a MEMS sensor system 1 with individual sensors for determining acceleration values AccX, AccY, AccZ and for Rotational rate values GyroX, GyroY and GyroZ, in each case for the three spatial axes X, Y, Z. The analog values AccX, AccY, AccZ, GyroX, GyroY and GyroZ are digitized via A / D converter 2 and stand as discrete measured values AccX ', AccY ', AccZ', GYROX ', GyroY' and GyroZ 'on the output side for further processing proces ^¬ available. These measured values AccX ', AccY', AccZ ', GyroX', GyroY 'and GyroZ' are first fed to any storage medium 3 and then evaluated in an evaluation device 4 for calculating the desired motion parameters 5 using different algorithms. These evaluation algorithms are illustrated in FIGS. 4, 5 and 6.

In principle, the second algorithm shown in FIG. 5 is to be preferred, as FIG. 2 shows. In this second algorithm, only acceleration values Acc are needed to determine the desired motion parameters 5. The acceleration values Acc may be provided as opposed to the rotation rate with ^¬ means of the MEMS sensor 1 in a non-cumulative method. The more error-prone rotation rate measurement values Gyro required for the algorithms one and three shown in FIGS. 4 and 6 are only used for the calculation of the motion parameters 5 if the algorithm two is not applicable.

However, it is also possible to reverse the method sequence according to FIG. 2 and to have the Acc algorithm two run parallel to the gyro algorithm one and / or three, as FIG. 3 shows, and only then the decision on the one for the calculation of the motion parameters 5 algorithm to use.

The algorithm one will be explained below with reference to the flowchart shown in FIG. This algorithm one is required for offset and drift compensation in the rate-of-turn based on MEMS gyro sensors. The algorithm is explained using the example of the inclination in the direction of the main movement direction of the moving object. The use of the algorithm is of course not limited to this case, but can be used in any spatial axis. To the interference offset - offset - to minimize and zero drift in the tilt calculation ^¬ means of rotation rates, ended the current betrach- time t upstream and a trailing consideration of the standard deviation StabwGXv and StabwGXn the rate of rotation about the horizontal, to the main motion direction by 90 ° turned axis used over a fixed period. Falls below at least one of the two standard deviations StabwGXv and StabwGXn a fixed threshold value SGX, it signals a rectilinear motion, which is accelerated or may be the same ^¬ shaped. In the case of such a movement, the rotation rate is zero. For the specified period, a leading and a trailing average MwGXv and MwGXn are calculated. These averages MwGXv MwGXn and provide the offset before and after the currently considered point in time t. Falls below only one of the two standard deviations StabwGXv and StabwGXn SGX the threshold, only the dazuge ^¬ hearing mean MwGXv or MwGXn is used. If both standard deviations StabwGXn and StabwGXv exceed the

Threshold SgX, the mean value is again formed from the two mean values MwGXv and MwGXn. The thus ultimately decisive mean - trailing, or Running forward with ^¬ average value of both - can now be used to compensate the current offset. The compensation can be carried out in the various stages of the inclination calculation, namely with respect to the rotation rate, the calculated angle or the calculated inclination. In this way, there is a running compensation of the current offset. Ultimately, it also compensates for disturbances of the drift.

In the example illustrated in Figure 5 algorithm to two inclination calculating only acceleration measurements ACCX 'be ^¬ forces. For this purpose, the standard deviations StabwAXv and StabwAXn over a certain period of time are compared with the current acceleration measured value AccX 'and compared with a threshold value SaX. If this threshold SaX below of at least one of the two standard deviations StabwAXv and StabwAXn, it means that the object is in a state in which the Ge ^¬ speed is constant along the axis considered. Only in these phases with constant speed is recourse made to the inclination calculation by means of acceleration measurement. This is driving ^¬ is a nichtkumulatives Ver, so that errors can not propagate. The simultaneity of the measurements of acceleration and yaw rate is therefore very important for this algorithm. If the phase is completed at a constant speed and acceleration takes place, must be able to resort to the Drehra ^¬ tenmessung to calculate the slope immediately. In addition, the case can be considered, in which maximum inclinations must not be exceeded. According to these maximum inclinations, there are also permissible limit values for the acceleration. The slope would be larger in the case of a constant linear acceleration than the maximum slope would allow. Thus, there are two thresholds to consider, namely the standard deviation threshold SaX and the maximum acceleration threshold MaxAX. Consequently, the current acceleration value and min. least one of the two standard deviations StabwAXv StabwAXn or smaller than the respective threshold SaX be rela ^¬ hung MaxAX example, to use the acceleration measurement to the slope calculation.

A third algorithm, illustrated in FIG. 6, is used to compensate for the influences of bank angles. If the object moves around a curve, this is in many cases associated with a bank, ie with an inclination about the axis of the main movement direction. This bank can either be directed to the outside of the curve, for example, as a result of the centrifugal force, or be directed to the inside of the curve, for example in the case of over-elevation curves for maximizing the passage-through speed. Both cases lead to measurable rates of rotation about the horizontal, perpendicular to the main motion ^¬ spatial axis. The effect of the rotation in the curve is directly dependent on the bank. Since the rotation rates around this space axis, which is perpendicular to the main movement direction, serve to calculate the inclination in the main direction of movement, incorrect inclination values occur when cornering in algorithm one, which must be compensated. The Quernei ^¬ supply can of course only in the case of a curve leading to compensation values or the ever-calculated compensation values must be taken into account only in the case of a curve. The transverse acceleration can not be used in this case, since the acceleration caused by the Quernei ^¬ supply, are superimposed by the lateral acceleration, which lead to cornering.

To identify a curve, the rotation GyroZ 'around the vertical Z axis must be considered. To calculate the bank angle, the yaw rate measurements around the main heading are used. Since these in turn require a cumulative calculation, the algorithm must also be one with corresponding axis reference are used. The calculation so ^¬ designated tilt used to compensate for the coming from the bank of error in the tilt calculation. The three algorithms of Figures 4, 5 and 6 allow accurate and reliable computation of motion parameters. The use of the inexpensive compared to gyroscopes with rotating gyroscopic MEMS technology is made possible only by the algorithms described.

Claims

claims

1. A method for determining movement parameters of moving objects, in particular for rail vehicles,

d a d u r c h e c e n c i n e s that

along at least one of the three spatial axes (X, Y, Z) by means of MEMS (Micro-Electro-Mechanical Sys ^¬ system) acceleration sensors, the acceleration and / or at least one of the three spatial axes (X, Y, Z) by means of MEMS gyro sensors, the Rotational and / or leading average values (MwGXn and / or MwGXv) depending on the associated, with. For each digital or digitized rotation rate measured value (GyroY ') with respect to a rotated by 90 ° spatial axis (X) standard deviation (StabwGXn and / or StabwGXv) compared to a threshold value (SgX) can be used for offset correction of the gyro measurement (GyroY ').

2. The method according to claim 1,

d a d u r c h e c e n c i n e s that

for digital or digitized Accax 'AccX', trailing and leading standard deviations (StabwAXv and StabwAXn) are compared to a Threshold (SaX), using the acceleration measurements (AccX ') for inclination calculation if at least one of the two standard deviations (StabwAXv and / or StabwAXn) falls below the threshold value (SAX), and optionally a corresponding maximum acceleration be ^¬ (MaxAX) (from the acceleration measurement value ACCX ') falls below a maximum permissible tilt.

3. The method according to any one of the preceding claims,

characterized in that to identify a plot of yaw rate readings (GyroZ ') about a vertical space axis (Z) and to compute a curve-related roll of the spin rate measured value (GyroX') object about a main motion direction of the object.

4. Apparatus for carrying out the method according to one of the preceding claims,

d a d u r c h e c e n c i n e s that

at least one MEMS (Micro-Electro-Mechanical

System) acceleration sensor for determining the acceleration along at least one of the three spatial axes (X, Y, Z) and / or at least one MEMS gyro sensor is provided for determining the rotation rate about at least one of the three spatial axes (X, Y, Z) her / his digital or digitalized ^¬ catalyzed measured values (ACCX ', accy', AccZ ', GYROX', GyroY '

GyroZ ') are supplied via a storage medium (3) to an evaluation device (4) for calculating the movement parameters, wherein means for offset correction of the rotation rate measured value

(GyroY ') with reference to a rotated by 90 ° spatial axis (X) trailing and leading average values (MwGXn and MwGXv) and associated with a threshold (SgX) to ver ^¬ equivalent standard deviations (StabwGXn and StabwGXv) are provided.