WO1988005155A1

WO1988005155A1 - Navigation process for vehicles equipped with an electronic compass

Info

Publication number: WO1988005155A1
Application number: PCT/DE1987/000559
Authority: WO
Inventors: Heinz Ewinger; Reinhard HELLDÖRFER; Ulrich Kanzler; Hans Rauch
Original assignee: Robert Bosch Gmbh
Priority date: 1986-12-30
Filing date: 1987-11-28
Publication date: 1988-07-14
Also published as: DE3644682A1; JPH02501856A; AU599727B2; AU8238787A; EP0340219A1

Abstract

A process for indicating the direction of travel of a vehicle by determining the locus curve of a magnetic field on a magnetometer fixed to the vehicle. The magnetometer is provided with two probes perpendicular to each other and is connected to an evaluation circuit that cyclically processes as electric measurement values the components of the instantly effective magnetic field measured by the probes. The measurement values (P) are processed to determine the orientation of the Earth's magnetic field, the north or the direction of travel of the vehicle, by processing a plurality of measurement values (P) taken during a first circular journey of the vehicle in order to determine the parameters of the locus curve. By averaging several measurement values (P1, P2) cyclically read one after the other among the plurality of measurement values (P) determined during the circular journey of the vehicle, mean values (m1) are deduced and temporarily stored in a data cache (12a) of fixed capacity in order to be later used for calculating the parameters of the locus curve.

Description

Navigation method for vehicles with electronic compass

State of the art

The invention is based on a method for determining the locus of a magnetic field on a magnetometer which is permanently arranged in a vehicle, according to the preamble of the main claim. Such a method is known from DE-OS 35 09 548.

When measuring the earth field vector in a vehicle for navigation purposes, a complete rotation of the vehicle does not result in a circular locus curve around the coordinate origin of the measuring system due to the interference fields present in the vehicle, but rather an elliptical locus curve which is independent of the direction due to a hard magnetic component of the interference fields in the direction Interference field vector is shifted from the coordinate origin of the measuring system and which also causes the elliptical shape and a certain rotation of the ellipse in the coordinate system due to the soft magnetic portion of the interference fields in the vehicle. In the known calibration method according to DE-OS 35 09 548, relatively few measuring points are used to calculate the parameters for the elliptical locus of the magnetic field on the magnetometer during a circular travel or complete rotation of the vehicle. Thereby however, it follows that the detection of the interference fields in the vehicle by this calibration method is only relatively coarse and therefore often too imprecise.

With the present solution, the aim is to determine the locus of the magnetic field effective on the magnetometer or a magnetic field meter with increased accuracy with a circle or rotation of the vehicle when calibrating an electronic compass, in order to determine the direction on the display of the electronic compass during the subsequent navigation to improve.

Advantages of the invention

The method according to the invention with the characterizing features of the main claim has the advantage that, by averaging a plurality of measured values recorded in succession, measured value tolerances when determining the elliptical. Locus curve are taken into account and compensated, so that the mean values obtained in this way enable the elliptical locus curve and its parameters to be calculated much more precisely. Another advantage is that with this method, a considerably more accurate directional display of the electronic compass can be realized after the calibration run.

Advantageous further developments and improvements of the features specified in the main claim are possible through the measures listed in the subclaims. To further increase the accuracy of the display, it is advantageous to first determine a first center point of the locus curve and its displacement from the coordinate origin from the mean values found of the elliptical locus curve and then to break down the locus curve into a certain number of equally large angle segments. The mean values stored in the data buffer are then assigned to these angle segments and then all the mean values of each angle segment can be calculated a so-called real value of the locus is formed and temporarily stored. Now a further center point can be determined from the real values of the locus curve obtained in this way and with this, the locus curve can be broken down again into angle segments of the same size. New real values of the locus can now be formed from the mean values assigned to them in the data buffer and temporarily stored. The determination of a further center point and new real values of the locus curve is advantageously repeated until the center point has remained the same in two successive calculation runs.

drawing

An embodiment of the invention is shown in the drawing and explained in more detail in the following description. FIG. 1 shows a navigation system for a motor vehicle in a block diagram, FIG. 2 shows the schematic representation of a data buffer for forming the mean values, FIG. 3 shows the division of the elliptical locus into angle segments and FIG. 4 shows the elliptical locus for calculating its parameters.

Description of the embodiment

FIG. 1 shows a dead reckoning navigation system for motor vehicles, with which the driver can find the desired destination in a foreign environment by specifying the direction of travel and the air line distance (destination finding system). It consists of an input and output unit 10, a microprocessor 11 with a data memory 12 and a displacement sensor 13 and a travel direction sensor 14. The microprocessor 11 is usually included in the input and output unit 10 together with the data memory 12. Pulse generators of a tachometer generator or corresponding transmitters can be used as displacement transmitters 13 the vehicle wheels are used, which may already be in the vehicle. A magnetometer or a magnetic field meter should be used as the direction indicator. The microprocessor with its periphery and the data memory 12 is connected both to the displacement sensor 13, the magnetometer 14 and also to the control buttons and a liquid crystal display 15 of the unit 10.

The navigation system is operated with five control buttons. Numerical values can be changed in the direction larger and smaller on the LCD 15 using a rocker button 16. With the actuation of an acknowledgment key 17, the current numerical value displayed on the LCD 15 is stored, which appears in the lower area of the LCD 15 on the display 18. With a function selection key 19, the navigation system is switched within an offered menu, in accordance with a label 25 on the left edge area of the unit 10, the information displayed on the LCD 15 being indicated by an arrow 20 on the LCD 15 at the level of the label 25 . Another key switch 21 is used to switch the dead reckoning system on and off. Another display 22 in the upper LCD area is used to identify targets. A rosette 23 of the LCD 15 with sixteen different invisible arrow segments is used for directional information, the controlled directional arrow 24 representing either the north direction, the direction of travel or the direction of the destination to be approached.

The navigation system is designed in such a way that it can be installed in any motor vehicle. To calibrate the navigation system, it is first necessary to take a circular drive with the vehicle to determine the size, shape and position of the elliptical locus. During the subsequent navigation trips, the measured values transmitted from the magnetometer 14 to the microprocessor 11 can then each be assigned to a point on the local curve on the basis of the determined and stored locus curve, and the north direction or the direction of travel of the vehicle can be calculated and displayed therefrom. FIG. 2 shows a data buffer 12a with a fixed memory length, which forms part of the data memory 12 from FIG. 1. In the data buffer 12a, during the calibration of the navigation system, averages are formed from the large number of measured values which are output by the magnetometer 14 during the circular movement and recorded by the evaluation circuit by averaging a plurality of measured values which are read in cyclically and one after the other and are temporarily stored in the data buffer 12a . Since the data buffer 12a has a fixed length, only a certain number of mean values can be temporarily stored in it. The number of measured values of the magnetometer recorded during the circular travel of the vehicle is not predetermined, since the measured values occur in a fixed cycle time of approximately 100 ms and thus the total number of measured values recorded from the driving speed or rotational speed of the vehicle during the circular travel or the full one Vehicle rotation depends. The display of the measured values used for averaging is therefore determined by the evaluation circuit from the ratio of the total number of all measured values recorded during the circular drive and the number of storage locations in the data buffer 12a. For this purpose, mean values formed several times in succession or previously are read into the data buffer 12a, these are then averaged in pairs and the new mean values are then temporarily stored in the data buffer 12a instead of the previous values.

The procedure shown in the schematic illustration in FIG. 2 is advantageous. There, averaging must be carried out in several passes whenever the data buffer 12a is filled beforehand during the calibration run. At the beginning of the calibration run, a first part of the measured values P1 to P50 is read into the data buffer 12a and buffered until it is fully loaded. This can be seen from the left column of the schematic illustration in FIG. 2. The first averaging follows, in which two successive measurements are made from the stored measured values evaluate z. B. P1, P2 a mean ml is formed. The mean values ml to m25 obtained in this way are now temporarily stored in the data buffer 12a instead of the previous measured values P1 to P50, which are now only half full. Now further successively recorded measured values are first averaged in pairs in front of the data buffer 12a and then these mean values, e.g. P51 / 2 + P52 / 2 entered in the data buffer 12a until it is refilled. This first averaging can be seen in column 2 of the schematic illustration in FIG. 2. Then, by means of a second averaging, two mean values are again formed in the data buffer 12a and these are then stored in the data buffer 12a instead of the previous mean values. For example, they result from m1 / 2 + m2 / 2. The data buffer 12a is again only half full and it is again filled up with mean values from a corresponding number of further measured values P100 / 4 to P104 / 4 etc. If the circular drive has not yet ended, a third mean is again used to form a new mean from two successive mean values. It corresponds to the mean value from the first eight measured values P1 ... P8 or from the first four mean values m1 ... m4. The data buffer 12a, which is then only half full, is then refilled with further mean values from a corresponding number of further measured values. This averaging and filling of the data buffer 12a is continued until, at the end of the first circular travel of the vehicle, all measured values P recorded with the same averaging are temporarily stored in the data buffer 12a as the final average values m. These mean values now form points for the elliptical locus of the magnetic field effective on magnetometer 14. From these mean values in the data buffer 12a, the large and the small semiaxis, the rotation and the displacement of the elliptical locus curve with respect to the origin of the coordinates are determined in the evaluation circuit as parameters. With the help of FIG. 3 and FIG. 4, it is explained how the parameters for the elliptical locus curve to increase the measurement accuracy of the navigation system are checked several times at the end of the circular drive and corrected if necessary. In a first step, according to DE-OS 35 09 548, the maximum and minimum values on the X and Y axes of the coordinate system are acquired from the mean values m stored in the data buffer 12a and a first center point M of the locus O is calculated therefrom. Its shift from the origin of the coordinate forms the hard magnetic, direction-independent interference field fector after this preliminary determination of the

At the center M, the elliptical locus O is now divided into a certain number of, for example, 24 equally large angle segments w with the segment angle μ. The mean values m stored in the data buffer 12a are now assigned to these angle segments w. From a plurality of mean values m for each angle segment w, a so-called real value e of the elliptical locus curve O is then formed by a further averaging of these values and this is temporarily stored in the data memory 12. A new center point Mo is now determined from the real values e of the locus curve O thus obtained for each angle segment w. Then the locus O is broken down again into the angle segments w and new real values e of the locus O are formed from the mean values m assigned to them in the data buffer 12a and temporarily stored. The evaluation circuit then repeats the determination of a further center point and new real values of the locus curve until the center point Mo is no longer shifted in two successive calculation runs. Depending on the position of the angle segments w, individual mean values m in the data memory 12a can be counted in the successive calculation runs first to one angle segment and then to another, adjacent angle segment. It is essential for increasing the accuracy of measurement that the real values e found now represent the locus O at uniform angular intervals. To calculate the large and small semiaxes as well as the rotation α of the elliptical locus O compared to the origin of the coordinate, four points Pa, Pb, Pc and Pd on the locus O, each offset by 90 °, are sought. The first point Pa is chosen arbitrarily by the evaluation circuit. Now the semiaxes a1, a2, b1 and b2 of these points to the center are calculated. A difference value is then determined in pairs from the semiaxes of these points (a1-a2, b1-b2) and this difference value is temporarily stored in the data memory 12. In a further step, a further four points of the locus curve O, each offset by 90 ° to one another, are determined, and here too the difference value of their semiaxes is calculated and compared with the stored difference value. These process steps are now continued by the evaluation circuit and the difference values are compared with one another until a minimal difference value is present. This results at the four points E1, E2, E3 and E4, which are assigned to the large and the small semiaxis of the elliptical locus 0. It is advisable not to simply add the difference values of the large and small semiaxes of four points on the locus to form a total difference value, but rather to form this difference value D of four points on the locus according to the following algorithm:

* b

Weighting factors of <1 are chosen with x1 and x2. The values a1, a2, b1, b2 form the semiaxes of the four points, each offset by 90 ° to the center point Mo of the locus O.

Since the center point M, Mo of the locus curve O is only calculated very imprecisely via the maximum and minimum X and Y measured values, the calibration phase for determining the center point of the ellipse can be shortened by first averaging all of them mean values standing in the data buffer 12a and is determined in the further calculation runs by averaging the real values e found previously. It is also possible to add up all the X and Y measured values recorded during the circular drive, to divide the sum by the number of measured values and thus to form a first X and Y value for the center M of the locus curve O.

After finding the points E1, E2, E3 and E4 with the smallest difference value D, the small and large semiaxis a and b of the ellipse O and their rotation α in the coordinate system are determined in a known manner by the evaluation circuit using the following equations:

φ = 1/4 [φ (El) + ρ (E2) - 90 ° + φ (E3) - 180 ° + φ (E4) - 270 °]

These calculated parameter values of the locus curve O are also stored in the data memory 12. If necessary, they can be corrected in later navigation procedures using known procedures.

Claims

Expectations

1. Method for determining the location curve of a magnetic field on a magnetometer fixedly arranged in a vehicle with at least two probes rotated by 90 ° with their longitudinal axes and with an evaluation circuit which cyclically detects the components of the magnetic field active there as electrical measured values and to determine the direction of the earth's magnetic field, the north direction or the direction of travel of the vehicle, by recording a plurality of measured values of the magnetic field vector effective on the magnetometer during a first circular travel of the vehicle and processing them to determine the parameters of the locus curve of the magnetic field, characterized in that that from the multiplicity of the measured values (Pl ... PX) recorded during the circular travel, by averaging a plurality of measured values (P) read in cyclically one after the other, mean values (m) are formed, which are stored in a data buffer (12a) with a fixed length for calculating the Local area Urven parameters are cached.

2. The method according to claim 1, characterized in that the number of measured values (P) used for a final averaging is determined from the ratio of the total number of all measured values (p) recorded during the circular travel and the number of storage locations in the data buffer (12a).

3. The method according to claim 1 or 2, characterized in that readings (P) or previously formed mean values (m) are read into the data buffer (12a) several times in succession, these are then paired averaged and the new mean values (m) are then temporarily stored in the data buffer (12a) instead of the previous values.

4. The method according to any one of the preceding claims, characterized in that

a) that a first part of the measured values recorded during the circular drive are temporarily stored in the data buffer (12a) with a fixed length until it is fully loaded, b) that an average value (m1) is thus formed from two successive measured values (P1, P2) and these mean values (m) are temporarily stored instead of the measured values (P) in the data buffer (12a), which is then only half full, c) that further measured values (P51, P52) are first averaged over the same number and then these mean values successively in the data buffer (12a) are entered until it is full again, d) that further mean values (m) are then formed again from two successive values (m) in the data buffer (12a) and these are then stored in the data buffer (12a) instead of the previous mean values e) that the data buffer (12a) is filled again with mean values from a corresponding number of further measured values (P), f) that the averaging and filling continue for so long is set until, at the end of the first circular travel of the vehicle, all measured values (P) with the same averaging are buffered in the data buffer (12a) as final averages (m), which points on the elliptical locus (O) of the magnetometer (14) are effective Represent magnetic field and g) that in the evaluation circuit from the mean values (m) of the data buffer (12a) as parameters the large and small semiaxis (a, b), the rotation (φ) and the displacement of the elliptical locus (O) with respect to the coordinate origin determined

become.

5. The method, preferably according to one of the preceding claims, characterized in that a first from the mean values (m)

Center point (M) of the locus (O) and its displacement

the origin of the coordinates is determined that the locus (O) is then broken down into a certain number of equally large angle segments (w) and the mean values (m) stored in the data buffer (12a) are assigned to the angle segments (w) and that the mean values ( m) a real value (e) of the locus (O) of each angular segment (w) is formed and temporarily stored.

6. The method according to claim 5, characterized in that a further center point (Mo) is determined from the real values (e) of the locus (O), then the locus (0) is again broken down into its angular segments (w) and from the mean values assigned to them (m) new real values (e) of the locus (O) are formed in the data buffer (12a) and temporarily stored.

7. The method according to claim 6, characterized in that the determination of a further center point (Mo) and new real values (e) of the locus (O) is repeated until the center point (Mo) no longer shifts during two successive calculation runs.

8. The method preferably according to one of the preceding claims, characterized in that the center point (M) of the locus (O) and its displacement from the coordinate origin each by

an averaging of the measured values (P), the temporarily stored mean values (m) or the real values (e) is determined.

9. The method preferably according to one of the preceding claims, characterized in that the calculation of the large and dee small semiaxes (a, b) and the angular rotation (φ) of the elliptical locus (O) compared to the origin of the coordinates, four points (P1, P2, P3, P4) offset from each other on the locus (O), the semiaxes ( a1, a2, b1, b2) calculates these points to the center point (M), determines a difference value (D) from the semiaxes of these points and this difference value with a difference value of semiaxes for a further four points of the locus curve, each offset by 90 ° ( O) is compared until there is a minimum difference value (D) which is assigned to the four points (E1, E2, E3, E4) on the large and small semiaxis (a, b) of the elliptical orbit curve (O).

10. The method according to claim 9, characterized in that in each case the difference value (P) of the four points (P1 to P4) of the locus (O) according to an algorithm ι; Z τ

is formed, where x1 and x2 are weighting factors and a1, a2, b1, b2 represent the semiaxes of the four points to the center (M) of the locus (O).