JPS62206413A - Gyro-bias error calibration system of strap-down type inertia reference apparatus - Google Patents

Abstract

PURPOSE:To curtail setting of a special reference-axis gyro, by constructing a 3-axis Cartesian coordinate system using 2 gyros with 2-axial degrees of freedom and using the remaining one axis for calibration of bias errors. CONSTITUTION:2-axis degree of freedom gyros 1, 2 are installed. 3-axis Cartesian coordinates are constructed with input axes X2, Y2 of the gyro 2 and input axis Y1 of one side of a gyro 1 and the remaining input axis X1 is used for calibration of bias errors. Upon detecting the bias errors, the gyro 2 is rotated through an angle of approximately 90 deg. respectively around the axis Y1 of the gyrol with respect to the gyrol and detection rate of the gyro obtained in each rotation position is introduced into a processing apparatus. The processing apparatus determines a bias error of a required axis based on the detection rate.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for calibrating (calculating) gyro bias error in a strap-down type inertial reference device.

[Summary of the invention]

The present invention uses two two-axis degree-of-freedom gyros to construct three-axis orthogonal coordinates using three axes out of a total of four axes. The other gyro is rotated by approximately 90'' with respect to the one gyro about the axis of the gyro, and the detection rate of the gyro axis obtained at each rotational position is processed by a processing device to obtain the desired value of the gyro. This method is designed to calculate the axis bias error.

[Conventional technology]

In a strap-down inertial reference device, such as a strap-down gyro compass, if there is a bias error in the gyro (angular velocity sensor) used in the device, an error will occur in the azimuth that is the output of the gyro compass. The amount of azimuth error is proportional to the eastward component of the gyro bias error. Therefore, in order to obtain accurate azimuth, find the gyro bias error and
It is necessary to correct the azimuth including the error.

An example of a gyro bias error calibration method for a conventional strap-down type inertial reference device for this purpose will be described with reference to FIGS. 5 and 6.

In this method, the input shaft (X2) of the gyro (G).

(Y2) by 180° around the axis (Z) perpendicular to both, both axes (X2) at the 0° position in A of the same figure
, (Y2) and the detection rate of both axes (X2) and (Y2) at the 180° position to find the bias error of the gyro (G).

A specific example of the method for determining the bias error of this gyro will be explained with reference to FIG. At the 0° position of the gyro (G) shown in Figure 5A, the axis of the gyro (G) (X2
), (Y2) respectively detection rate (angular velocity) ω fold, (O
) and ω'/2(0) are stored in the memory (23). The detection rate ωx2(0) of one axis (X2) is the external action rate ωX and the bias error BX2 of the axis (X2).
(ωX+BX2), and the other axis (¥2)
Detection rate ω'? 2 (0) is a value (ωY + B) that includes the external action rate ωY and the bias error BY2 of the axis (Y2)
Y2). Next, the gyro (G) is on the axis (X2).

180 around the axis (Z) perpendicular to the plane containing (Y2)
It rotates through 180° and stops at the 180° position shown in FIG. 5B.
The detection rate ω fold z (180) of the axis (X2) of the gyro (G) at the 180° position includes the external action rate −ωX and the bias error BX2 of the axis (X2) (−ωX+
BX2), the detection rate ω% (180) of the axis (Y2) is
External action rate - ωY and axis (Y2) bias error B
Y2 (-ωY+BY2). Therefore, the detection rate ω5z(0) of the axis (X2) at the 0° position stored in the memory (23) is retrieved from the memory (23), and this is used as the detection rate ω of the axis (X2) at the 180° position.
(180) and the adder (24). Since the result of this addition is 2BX2, the output of the adder (24) is multiplied by 1i2 by the multiplier (25) to obtain the bias error BX2 of the axis (X2).

On the other hand, the detection rate ω' of the axis (Y2) at the 0° position stored in the memory (23)? 2 (0) to memory (23)
, and calculate this as the detection rate ω' of the axis (Y2) at the 180° position. 2 (180) and an adder (26). The result of this addition is 2BY2, so the adder (2BY2)
By multiplying the output of step 6) by 1i2 by the multiplier (27), the bias error BY2 of the axis (Y2) is obtained.

[Problem that the invention seeks to solve]

However, in the gyro bias error calibration method of such a conventional strap-down type inertial reference device, as described above, the absolute values of the external action rates ωX and ωY on the gyro (G) are different from the 0° position and the 180° position. °
Since it is assumed that the condition does not change depending on the position,
When there is oscillation, such as on a ship equipped with a gyro, the oscillation causes the external action rate to change to the 0° position and 1
There is a problem in that the bias error changes at the 80° position and causes an error, making it impossible to obtain an accurate bias error.

[Means for solving problems]

The present invention aims to eliminate the above-mentioned problem, and the means thereof is to use two two-axis degree-of-freedom gyros (11, f21, for a total of four axes (XI), (Yl), (X2).

Of (Y2), 31[1i1 axes are used to construct three-axis orthogonal coordinates (X, Y, Z), and one gyro (1) is orthogonal to both axes of the other gyro (2). The other gyro (2) is rotated approximately 90 degrees with respect to the one gyro fil around the axis (Yl), and the processing unit (3) processes the detection rate of the axis of the gyro obtained at each rotational position.

(5)~(13), (14), (15)~
This is a gyro bias error calibration method for a strap-down type inertial reference device in which the bias error of a desired axis of the gyro is determined by processing in (22).

[Effect]

The present invention uses two two-axis degree-of-freedom gyros to construct three-axis orthogonal coordinates using three axes out of a total of four axes. ft.
The other gyro is rotated approximately 90° relative to the other gyro around the 1il axis, and the detection rate of the gyro axis obtained at each rotational position is processed by a processing device, thereby making it independent of the external action rate. to find the bias error of the desired axis.

〔Example〕

Hereinafter, the present invention will be explained based on the drawings. 1st and 2nd
The figure is a diagram for explaining one embodiment of the present invention.

In Fig. 1, (1, 1, and (2) indicate two-axis degree-of-freedom gyros, respectively. (XI, ), (Yl), and (X2)
.

(Y2) represents the input shafts of the gyros (1) and (2), respectively. Input shaft (X2) of gyro (2), (
Y2) and both axes (X2), and one input axis (Yl) of the gyro (1) perpendicular to (Y2) constitute a three-axis orthogonal coordinate (X, Y, Z), and the gyro (1) The remaining input axis (Xl) is used for bias error calibration. The relative angles of the axis (Xl) and axis (X2) of the gyro (2) to the gyro (11) are A, B, C,
As shown in D, O', 180°, 90°.

It rotates around the axis (Z), that is, around the axis (Yl) of the gyro +1) so that it can be stopped at each position of 270°.

At the 0° position in Figure 1A, the axes (XI) and (X2) of the gyro (11, (21) are in the same direction,
) and (2), the detection rates (angular velocities) of the axes (Xl) and (X2) are expressed by the following equations (1) and (2).

ω51to)−ωXto)+B Xl−・・・■
ω'i%2(Ol−ω×fO1+Bx2””■).

Here, ωki(0) and ωρ2(0) are the axes (Xi) of the gyros (1) and (2) at the 0° position.

(X2) detection rate, ωx (0) is the external action rate, BX1+ BX2 are the axes (Xi), (X
2) represents the bias error.

By taking the difference between formula (■) and formula (2), the following formula is obtained.

BXI Bx2-ωki(o)-0%2(ol
”@Also, at the 180° position in Figure 1B, the axes (Xl) and (X2) are in opposite directions, and the gyros (1) and (2)
The detection rate of each axis (Xl) and (X2) is “L (1
80)-ωx (180) +-Bxz...
・■ωS2 (180) =-ωx (180) +
BX2 -...-■.

Now, by taking the sum of 2〇 and 2〇, we get the following formula.

Bxz+Bxz-ωxt (180) +ω%, (1
80) ``By solving ■Equations■ and Equations■ for BX□ and BX2, Bx1=+ (ω5t(0) a+whisper2(ol+ω also(
180) +ω■2(180) )...■ BX2-+ (-ω whisper, (01+ω%fOJ10ωL (
180) +0%2 (180) 1...■, and the gyro fl) and the axis (Xl) of (2) and (
The bias error of X2) is determined.

Similarly, at the 90'' position in Figure 1C, the axis (Xl) and (Y
2) is in the opposite direction, and the detection rate of the axis (XI) and (Y2) is the following formula (■) and [phase].

ω”?＜1(90)−ωx (90) 10Bx1”・
・■'? 2 (90) -1ωx (90) 10BY2 ・=l@1ko\, BY2 is the axis (Y2)
This is the bias error.

Take the sum of formula ■ and formula [phase], and get B xl + B y2-ωS1 (90) 10ω912
(90)... Obtain ■.

Furthermore, at the 270° position in the first diagram, the axis (Xl) and (
Y2) is in the same direction, and their detection rate is the following formula @
and becomes 0.

0%1 (270) -ωX (270) +BX1
"@ω'? r2 (270) - ωx (270) 10, BY2...Take the difference between @formula@ and formula [phase], Bx1Bv2-(')? +1 (270) (=1'
! 2 (270)...■ is obtained.

By solving equations ■ and ■ for BX11BY2, we get Bxl-+ (4s (90) +42(90) +ω
Qi (270) a+%2 (270) )...[
Phase] BY2-+ (ω9□(90)+ωS2 (90)-ω
'A1 (270) + ωS2 (270) 1...[
phase], and the bias errors of the axes (XI) and (Y2) are found.

The processing device (P) that performs the above calculation will be explained with reference to FIG.

At the 0° position of the gyro (2) shown in Figure 1, the axis (XI) of the gyro (], ), (2),' (X2
) bias error BX□, output ωx1 including BX2
(0) and ωρ2(0), i.e. (ωx (0) +B
x1) and (ωx (0) +BY2) in memory (
3).

Next, the gyro (2) rotates 90° relative to the gyro (]) and stops at the 90° position shown in FIG. 1C. This 90
In the ° position, the gyro (11, (21 axis (XI)
, the output ω'L including the bias error of (Y2)
(90) and “S2 (90), that is, (ωx (9
0) +Bx□) and (-ωx (90) +BY2
) in memo January 3).

Next, the gyro (2) further rotates 90 degrees relative to the gyro 41.1 and stops at the 180 degree position shown in FIG. 1B. At this 180° position, the gyro (11, (
2) Outputs ω9□ (180) and ω'? that include bias errors on the axes (XI) and (X2). <2 (180)
, that is, (ωx (180) +Bx1) and (−ωx
(180) +BX2) are added together in an adder (4) to obtain (BX1+BX2). Output ωx1 (0) and ω52 (0) at 0° position stored in memo January 3)
is taken out from the memo January 3), and the latter is subtracted from the former using the subtractor (5) to obtain (Bx□-BX2). As is clear from this, the outputs of the adder (4) and the subtracter (5) are values in which the external action rate is canceled, and the bias error T
3x1. Contains only BX2.

Therefore, by adding the outputs of the adder (4) and the subtracter (5) in the adder (6) to obtain 2Bx1, and multiplying this by 1/2 in the multiplier (7), the gyro (1) axis (Xl)
The bias error BX1 is obtained. Furthermore, adder (4)
By subtracting the output of the subtracter (5) from the output of the subtracter (8) to obtain 28X2, and multiplying this by 1/2 using the multiplier (9), the axis (X2) of the gyro (2) is Obtain bias error BX2.

Furthermore, gyro (2) is 90° with respect to gyro (1).
It rotates and stops at the 270° position shown in the first diagram. At this 270° position, the gyro (11, (2+ axis)
XI), output ω including bias error of (Y2)
ρ□(270) and ω'? 2 (270), i.e. (ω
, <(270) + Bxx) and (ωx (270
) + BY2) and the latter is subtracted from the former by the subtractor (10) to obtain (BXI BY2). The output ω at the 90° position stored in the memo January 3) is also (90) and ω% (
90) from the memo January 3) and add both of them to the adder (1
1) to obtain (Bxt+By2). As is clear from this, the re-outputs of the subtracter (10) and adder (11) are values with the external action rates canceled and contain only bias errors. Therefore, the output of the subtracter (10) is subtracted from the output of the adder (11) by the subtracter (12) to yield 2BY2.
By multiplying this by 1/2 by the multiplier (13), the bias error BY2 of the axis (Y2) of the gyro (2) is obtained.
get.

As described above, in an aspect of the present invention, even if the rate of external action on the gyro changes at each position, the bias error of each axis of the gyro can be detected regardless of this change.

FIG. 3 is a route map for explaining another embodiment of the present invention.

This calculates the bias error of the gyro's input axis at the three positions of 0°, 180°, and 90° shown in Figure 3 A, B, and C, which saves time compared to the example in Figure 1. .

That is, by solving the above 2〇, ■, and ■ for BX□, B×2, and BY2, the bias error of each axis is determined as shown in the following equation. That is, BX1=+ (ω support, (
0)-ω? 12(0) 10ωρ, (180) +ωdrop, (180) ')...0 BX2=+ (4t(ol++42(o)+ωL (1
80) +ωxt (90) +2ω?2 (9
0) a) StfO1+42 (01-ω also (180
)-0%2 (180) 1...[Phase] Referring to FIG. 4, a processing device (P
l) will be explained. At the 0° position of the gyro in FIG. 3A, the axis (Xi) of the gyro [11, [2].

The outputs ωx(0) and ω52(0) including the bias error of (X2), that is, (ωx(0) +Bxt) and (ωx(0)+BX2) are stored in the memory (14). Next, the gyro (2) rotates 90 degrees relative to the gyro (1) and stops at the 90 degree position shown in FIG. 3C.

At this 90° position, the gyro (], I1, (2
+ axis (XI).

Output ω%x (90
) and ωv2(90), i.e. (ωx (90) +
Bxt) and (-ωx (90) + B y2)
is stored in memory (14). Next, Jai 0 (23
further rotates 90 degrees with respect to the jig 0(j) and stops at the 180 degree position shown in FIG. 3B. At this 180° position, the gyro (11, (21 axis (Xi)), (
The output ω(180) and ω(180) including the bias error of X2), that is, (ωx(180)
+Bx1) and (-(13X (180) +Bx2
) are added by the adder (15) and the bias error (BX
1+BX2) only. Next, the outputs ωHiroshi (0) and ω'L (0) at the O° position stored in the memory (I4)
is taken out from the memory (14), the latter is subtracted from the former by the subtractor (16), and the bias error (BX BX2) is
Get only. By subtracting the output of the subtracter (I6) from the output of the adder (15) using the subtracter (17) to obtain 2BX2, and multiplying this by 1/2 using the multiplier (18), the output of the gyro (2) is Obtain bias error BX2. Furthermore, the outputs of both the subtracter (16) and the adder (15) are sent to the adder (I9).
28X1 is obtained, and this is multiplied by 1/ by the multiplier (20).
By doubling, a bias error BX1 of gyro +11 is obtained. Next, the outputs ω(90) and ω3□(90) at the 90° position stored in the memory (14) are added by an adder (21) with a bias error (Bx, +BY).
Obtain only 2). From the output of the adder (21) to the multiplier (2
The bias error BY2 of the gyro (2) is obtained by subtracting the output of the gyro (2) by the subtracter (22).

〔Effect of the invention〕

As described above, according to the present invention, even under dynamic conditions such as oscillation, a gyro bias error can be detected without being affected by oscillation or the like.

Furthermore, since two 2-axis degree-of-freedom gyros are used and the remaining axes forming the 3-axis orthogonal coordinates are used, there is no need to newly provide a gyro for the reference axis, so the structure is relatively simple.

[Brief explanation of drawings]

First FI! J is a route map for explaining the principle of one example of the present invention, FIG. 2 is a block diagram of a processing device implementing the principle, FIG. 3 is a route map for explaining the principle of another example of the present invention, and FIG. Figure 4 is a block diagram of a processing device that implements the principle.
5 and 6 are a route map and a block diagram for explaining the conventional example. In the figure, (1) and (2) are gyro, (31, (1
4) is memory, (41, (6), (11),
(25), (19), (21) are adders,
(51, +81. (10) , (12) ,
(16). (17), (22) are subtractors, (71, I91.
(13), (18). (20) respectively indicate multipliers.

Claims (1)

[Claims] Using two 2-axis degree-of-freedom gyros, 3 out of a total of 4 axes
A three-axis orthogonal coordinate system is constructed using three axes, and one gyro is centered on one axis perpendicular to both axes of the other gyro, and the other gyro is approximately 90 degrees relative to the one gyro.
A strap-down type inertial reference, characterized in that the bias error of a desired axis of the gyro is determined by rotating the gyro in degrees and processing the detection rate of the axis of the gyro obtained at each rotational position with a processing device. Gyro bias error calibration method for the device.