 |
Description  |
|
|
BACKGROUND OF THE INVENTION
This invention relates to laser inertial integrating rate sensors and, more specifically, to sensors in which a dither bias is used to obviate the effects of lock-in inherent in such sensors.
The behavior of ring laser inertial integrating rate sensors is well understood by those skilled in the art. Inherent in such sensors is the phenomenon known as "lock-in" in which counter propagating laser beams tend to lock together to a common
frequency. Lock-in arises in ring laser inertial integrating rate sensors at low rates of rotation. At such low rates of rotation, the frequency differential between the two beams is relatively small and the beams tend to couple or resonate together so
that the two beams oscillate at only one frequency. Because of this lock-in phenomenon, the frequency differential is no longer proportional to the rate of angular rotation, causing performance errors which have deleterious effects in navigation
systems.
To avoid or reduce the effects of lock-in, the laser inertial integrating rate sensor may be biased by dither techniques such as those shown and described in U.S. Pat. No. 3,373,650 to J. E. Killpatrick, assigned to the assignee of the present
invention and incorporated herein by reference. The biasing technique usually referred to as dithering may be implemented, typically using mechanical schemes, in a variety of ways. Since these biasing techniques directly affect the behavior of the
counter propagating laser beams, the sensor output signal contains not only rate information signals due to inertial motion but also contains a signal component related to the biasing of the sensor. This is true whether the output signal generator is
mounted directly on the sensor (block mounted) or off the sensor (case mounted).
The sensor output signal dither contribution further includes a base motion component due to dithering, herein referred to as ".theta..sub.B," and a sensor or gyro motion component due to dithering, herein referred to as ".theta..sub.G." The
sensor motor component of the dither signal represents the reaction of the sensor to the torque applied by a motive means attached to the base on which the sensor is mounted. The base motion component of the dither signal represents the motion of the
base on which the sensor is mounted which results from a reaction torque from the motor means mounted to the base. This torque acting on the base is of equal magnitude but in a direction opposite the direction of a dither torque applied to produce the
dither motion in the sensor. The torque acting on the base produces the base motion that is synchronous with the dither motion of the sensor and is opposite in direction.
Frequently, more than one inertial integrating rate sensor is mounted to an object for providing rotational information about the object. Sometimes only one or two inertial integrating rate sensors are mounted to the object as a means for
providing angular rotation information to a system. For example, inertial integrating rate sensors may be used to provide rotational information for an optical telescope having one or more axes of rotation. This rotational information is provided to a
control system that is capable of selectively activating servo motors to reposition the telescope thus insuring the telescope remains pointed at a selected target.
If dithered inertial integrating rate sensors are used, then the control system is provided rotational information from one or more sensors that are attached to a base which is in turn shock mounted to the telescope. Dithering the inertial
integrating rate sensor causes the sensor to react against the base producing base motion. Because the base is moving relative to the telescope, due to dithering, the control system receives rotational information about the telescope that is in error by
the amount of base motion due to dithering. Therefore, the control system is not able to maintain a pointing accuracy that is less than the angular rotation of the base due to dither.
If, however, the base motion due to dithering could be determined and supplied to the control system then the error due to base motion can be subtracted from the rotational information. The control system would then be capable of determining the
actual angular movement of the telescope. Once the actual angular movement of the telescope are determined by the control system then it is possible for the control system to correct or adjust for angular movements of the telescope that are smaller than
angular movements of the base due to dithering.
Thus, there is a need in these circumstances for a means for accurately determining the base motion due to dither. It is necessary that this base motion determining means maintain accuracy during aging and environmental effects such as
temperature variation.
Inertial navigation systems usually make use of three or more integrating rate gyros or sensors attached to the same base. Typically, three such inertial integrating rate sensors are coupled to the base having an orientation such that the axes
of sensitivity of the inertial integrating rate sensors are substantially mutually orthogonal to one another, as seen in FIG. 1. The sensor configuration shown in FIG. 1 includes a mounting base, 9, having three inertial integrating rate sensors, 10A,
10B, and 10C, attached thereto. Isolation mounts (not shown) tend to prevent the transfer of vibration between the mounting base 9 and the case (not shown) on which it is mounted. Each inertial integrating rate sensor 10A, 10B and 10C produces an
output signal that is indicative of the sum total of all the instantaneous angular motion changes the sensor has undergone in its input axis, or axis of sensitivity. The inertial navigation system transforms the sum of all instantaneous angular motions
produced by each inertial integrating rate sensor into navigation parameters.
The use of mechanically dithered inertial integrating rate sensors in these navigation systems results in base motion due to the base motion components, .theta..sub.B, of each individual inertial integrating rate sensor. As shown in FIG. 1, the
base motion component, .theta..sub.B, of sensor 10A produces a resultant motion of sensor 10C in a plane substantially orthogonal to the axis of sensitivity of sensor 10C. In a similar manner, the base motion component, .theta..sub.B, of sensor 10B
produces a resultant motion of sensor 10C in a plane substantially orthogonal to the axis of sensitivity of sensor 10C. Similarly, the base motion components of sensors 10A and 10C each produce a resultant motion of sensor 10B, and the base motion
components of sensors 10B and 10C each produce a resultant motion of sensor 10A.
This resultant motion of each of the inertial integrating rate sensors due to the base motion components of each of the other two inertial integrating rate sensors is such that a point in a plane substantially orthogonal to its sensitive axis
follows a Lissajous figure. The Lissajous figure produced for one sensor may be a straight line, ellipse or circle depending on the dither phases of each of the other two inertial integrating rate sensors, assuming each is dithered at the same rate.
The axis of rotation, or input or sensitive axis, of this third inertial integrating rate sensor more or less follows a cone due to such motion of the base, and so such input motion is generally referred to as "coning." This axis cone motion represents a
real input rate to the third gyro in addition to any inertial rotation about this axis and its dither reaction and, if not made insignificant or corrected for, is an error term in the inertial output signal.
The Lissajous figure for one sensor is the result of base motion components in each of two perpendicular axes due to the other sensors. These base motion components are represented by the angular displacement of sensor 10C due to sensor 10A and
10B dither motion, as seen in FIG. 1. The angular displacement of sensor 10C due to sensor 10A dithering can be represented by the following equation:
In equation 1, A.sub.A represents the amplitude of the base motion or oscillations due to sensor 10A dithering. The term fi represents the frequency of the base motion due to sensor 10A dither. The angular displacement of sensor 10C due to
sensor 10B dithering can be represented by the following equation:
In equation 2, A.sub.B represents the amplitude of the base motion due to sensor 10B dithering. The term .OMEGA. represents the frequency of the base motion due to sensor 10B dither. The term .xi. represents the phase shift between the
dithering of sensor 10A and the dithering of sensor 10B.
If the motions represented by equations 1 and 2 are assumed, then through standard mathematical translation a known mathematical representation for the mean angular rate sensed by sensor 10C can be derived. The mean angular rate that sensor 10C
senses in an axis of sensitivity orthogonal to the axis of sensitivity of both sensors 10A and 10B is represented by the following equation: ##EQU1##
In equation 3, the mean angular rate sensed by sensor 10C as a result of the dither motions of sensors 10A and 10B represents the coning error in the sensor 10C output signal. This coning error is constant with respect to time as long as both
the phase shift and frequency of the base motion are constant. The larger of terms A.sub.A and A.sub.B represents the major axis and the smaller term represents the minor axis of an ellipsoid Lissajous figure. If the dither motions produced by sensor
10A and sensor 10B are in phase, the term .xi. is zero and the coning error rate is zero.
If, alternatively, two orthogonally oriented gyros have different dither frequencies, the Lissajous figure in the plane of a third gyro is not constant but goes through both positive and negative phase relations. The volume swept out by the
input axis of this third gyro tends to be zero because the area swept out when the phase is positive is cancelled out by the area swept out when the phase is negative. An instantaneous coning or input error exists, but this error does not build up over
time. Therefore by running the gyros at different frequencies this coning error is greatly reduced, but not eliminated.
Another technique for reducing coning error has been to make the mass of the base very large relative to the mass of the sensor mounted thereon. In this manner, the base motion due to dither is reduced. The base motion due to dither, however,
is not eliminated.
The device base motion due to dither cannot readily be determined precisely from the dither pickoff signal provided by the piezoelectric output device because of the wide variation of the device output due to aging, temperature, and various
environmental effects. Hence, here too, there is a need for a means for determining the motion of the base due to dithering. Once the motion of the base due to dithering is known, it can be provided to the remainder of the navigation system which can
in turn correct for the error component in the sensor output signal that is due to this base motion.
In addition, the system can correct for any coning error that might be present in a sensor output signal if the actual base motion due to dithering is known. For sensors that are configured such that a coning error input is produced in one of
the sensor outputs, the system can compute or determine this error from the base motion components due to dithering resulting from the two orthogonal sensors. Once the coning error that is produced by each sensor is computed, it can then be subtracted
from the sensor output thereby eliminating coning errors from the inertial navigation system.
SUMMARY OF THE INVENTION
The present invention is an angular rotation sensing system for sensing rotational motion about a primary axis with respect to a base compliantly mountable to a supporting means. The angular rotation sensing system is capable of providing a
sensing system output signal that is indicative of angular rotation.
An angular motion sensor is included that is capable of providing a sensor output signal indicative of angular rotation about a primary axis. The sensing system further includes an angular motion sensor mounting means for mounting the angular
motion sensor to the base. The angular motion sensor mounting means has compliance so that the angular motion sensor can be rotationally oscillated.
An oscillation means is provided that is capable of producing a first torque acting on the angular motion sensor with respect to the base. Producing the first torque results in a second torque on the angular motion sensor acting in a direction
opposite the first torque causing both the angular motion sensor and the base to rotationally oscillate.
The sensor output signal indicative of angular rotation has an inertial component due to any inertial rotation of the angular motion sensor about the primary axis. The sensor output signal also has an oscillation component due to the rotational
oscillation of the angular motion sensor. The oscillation component further includes a base motion contribution and an angular motion sensor motion contribution.
An estimation means is included for providing an estimated value of the base motion contribution. A stripping means is included for determining the angular motion sensor contribution from both the oscillation component and the estimated value.
The stripping means provides the sensing system output signal with the base motion contribution therein having removed the angular motion sensor motion contribution.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a front plan view of a mechanical representation of a mounting base having three inertial integrating rate sensors thereon.
FIG. 2 is a system block diagram illustrating a closed loop system providing for removing dither signal components other than the base motion component from the ring laser output.
FIG. 3 is a diagrammatic sectional view of a dithered inertial integrating rate sensor assembly that is isolation mounted to a case.
FIG. 4 is a top plan view of a sensor block in isolation which is spring mounted to a sensor block support pin.
FIG. 5 illustrates a mechanical schematic representation of a dithered inertial integrating rate sensor assembly that is mounted to a base supported by an isolation mounting.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Referring to FIG. 2, there is shown a ring laser inertial integrating rate sensor, 10. Sensor 10 includes an output signal generator, 11, mounted to a block, 12. Block 12 provides the propagation path for counter-propagating laser beams.
Sensor output signal generator 11 is responsive to a portion of counter-propagating laser beams and provides output signals, 13 and 14, representative of the frequency difference between the beams as will subsequently be described in further detail.
Output signal generator 11, as stated above, includes a means for combining a portion of each of the counter-propagating laser beams to generate an interference pattern representing the results of the interference therebetween. Output signal
generator 11 usually includes two photodetectors responsive to different locations in the interference pattern for each providing one of a pair of sensor output signals 13 and 14 indicative of the intensity variation in the fringe pattern that impinges
the detector. Angular rotation of the sensor 10 in either direction about an axis of sensitivity causes a frequency differential to occur between the laser beam frequencies, which differential is proportional to the rate of angular rotation. The fringe
pattern represents the beat frequency of the heterodyned optical frequencies of the two beams. This fringe pattern consists of alternate light and dark bands of light that move one way or the other depending on the direction of rotation of the inertial
integrating rate sensor 10. The photodetector output frequency, or rate of fringe movement, is proportional to the rate of rotation of the sensor 10.
Output signal generator 11 provides output signals 13 and 14 to an output signal processor, 15. Sensor output signals 13 and 14 are sinusoidal signals that each represent the intensity variations in the fringe pattern that impinges the
respective detectors. Each detector is positioned a quarter wavelength apart so that output signals 13 and 14 will be sinusoidal signals 90.degree. out of phase from each other. When sensor 10 rotates in one direction, one of the sensor output signals
will lead the other by 90.degree.. When sensor 10 is rotated in the opposite direction, the other sensor output signal will lead by 90.degree.. In this manner, output signal processor 15 is capable of direction discrimination by determining which
sensor output signal, 13 or 14, is leading the other by 90.degree. of phase.
Output signal processor 15 provides output signals, 16 and 17, to a pulse accumulator, 100. To better understand the function of output signal processor 15, consider plotting the sinusoidal output signals 13 and 14 on an x-y axis. Output signal
13 is plotted on the x axis and output signal 14 is plotted on the y axis such that a circle is swept out. The output signal processor 15 digitally processes sinusoidal output signals 13 and 14 in a known manner to produce digital output signals 16 and
17. The direction of angular rotation of the inertial integrating rate sensor 10 is determined by the phase of output signals 13 and 14 represented by the direction in which the circle is swept out on the x-y axis. This direction of rotation of sensor
10 is represented by the sign of the digital output signals 16 and 17. In addition, the number of pulses present in output signals 16 and 17 represents the number of times the circle, or portions thereof, is swept out on the x-y axis.
Therefore, output signals 16 and 17 are a series of pulses which represent the instantaneous angular displacement of sensor 10 for a given direction. For example, logic level "1" pulses present in output signal 16 represent the instantaneous
angular displacements of sensor 10 in a clockwise direction while logic level "1" pulses present in output signal 17 represent the instantaneous angular displacements of sensor 10 in the counterclockwise direction.
Sensor 10 is also provided with an input dither signal, 18, operating a dithering mechanism to provide the dither or rotational motion bias as previously described. For example, block 12 may be rotationally oscillated for dithering purposes, as
shown in U.S. Pat. No. 3,373,650. A dither spring having piezoelectric motion inducing devices attached thereto causes the spring to periodically flex resulting in block 12 rotationally oscillating. Furthermore, a piezoelectric motion sensing device
may also be attached to the spring for providing an output signal, 19, identified as ".theta." that is directly related to the relative motion between the sensor 10 and the base 9. The just-mentioned piezo-electric output signal ".theta." is sometimes
referred to as the "dither pickoff signal."
The dither pickoff signal is substantially a sinusoidal signal with phase and amplitude related to the total motion of sensor 10. The dither pickoff signal is measured such that it reflects the relative rotation of the sensor and the base.
Because the directions of rotation for the sensor 10 and the base 9 due to dithering are opposite, or 180 degrees out of phase, the dither pickoff signal has peak amplitudes that are greater than either the sensor motion or the base motion due to
dithering. The dither pickoff signal 19 may be obtained by other techniques depending on the dither scheme that is selected.
The mechanical biasing of the counter-propagating laser beams results in rotation motion that directly affects the number and rate of fringe changes passing the photodetector, and so directly affects output signals 16 and 17. Output signals 16
and 17, therefore, include both a dither induced signal component and an inertial rotation induced signal component.
Also shown in FIG. 2 is a closed loop system providing for the removal of the dither signal components that are not related to base motion from the sensor readout digital signal herein designated ".theta..sub.S." A corrected output signal is
provided by the closed loop system that is substantially equal to the base motion component .theta..sub.B due to dithering and the inertial rotation component. More specifically, the closed loop system removes a sensor motion component .theta..sub.G of
the dither signal from the output signal while allowing a base motion component of the dither signal .theta..sub.B to remain therein.
Shown in FIG. 2 is a pulse accumulator 100 for integrating (counting) the series of pulses contained in the output signals 16 and 17 to produce an output signal .theta..sub.S indicative of the angular rotation of the sensor 10 at any instant.
The output signal .theta..sub.S of pulse accumulator 100 is a digital signal provided to a signal combining means, 101. Output signal .theta..sub.S is a 16 bit digital word representing one of 2.sup.16 values, with hexadecimal values OOOO.sub.H
representing the lowest cumulative angular rotation and FFFF.sub.H representing the greatest cumulative angular rotation. Signal combining means 101 combines a corrected pickoff signal identified by the .theta..sub.CP, to be described below, and the
sensor output signal .theta..sub.S provided by pulse accumulator 100.
The output signal from combining means 101 is defined as the corrected or "stripped" output signal and is herein designated as ".theta..sub.CS." Signal .theta..sub.CS is the difference between the sensor output signal .theta..sub.S and the
corrected pickoff signal .theta..sub.CP, thereby providing a corrected sensor output signal omitting dither signal components unrelated to base motion. The signal combining means 101 performs essentially a subtraction function so as to subtract signal
.theta..sub.CP from signal .theta..sub.S.
The corrected output signal .theta..sub.CS provided by signal combining means 101 is presented to a second signal combining means, 102, which receives a second input signal .theta..sub.C to be described below. The signal .theta..sub.C represents
a computed base motion sinusoidal signal, and has a magnitude and phase estimated representative of the base motion component of the corrected pickoff signal .theta..sub.CP. Signal combining means 102 provides an output signal .theta..sub.E
representative of the corrected output signal .theta..sub.CS omitting the computed base motion component .theta..sub.C of the dither signal. Signal combining means 102 performs essentially a subtraction function so as to subtract signal .theta..sub.C
from signal .theta..sub.CS.
The output signal .theta..sub.E from the second signal combining means 102 is presented to a synchronous demodulator, 103. The digital value of the dither pickoff signal is also provided to synchronous demodulator 103 by an analog-to-digital
(A/D) converter, 105. Synchronous demodulator 103 produces an output signal that is representative of the of those components of its input signal, .theta..sub.E, that are present in the digital representation of the dither pickoff signal. The
synchronous demodulator 103 output signal therefore does not contain substantially any of the sensor motion component .theta..sub.G. The output signal of synchronous demodulator 103 is a digitally encoded value ranging from zero to a maximum value, with
zero representing that little or no dither signal components are present in the .theta..sub.E signal.
In addition to receiving the output signal from synchronous demodulator 103, a gain means, 104, also receives a second input signal, the digital value of the dither pickoff signal, from an A/D converter, 105. The gain means 104 produces an
output signal .theta..sub.CP, herein referred to as the corrected pickoff signal. Gain means 104 functions as a variable gain amplifier having a gain determined by the output signal of synchronous demodulator 103.
Because signal .theta..sub.E contains only dither signal components other than the base motion due to dithering .theta..sub.B, the output signal from synchronous demodulator 103 represents the dither signal components other than the base motion
component present in signal .theta..sub.E. The dither pickoff signal .theta..sub.P, represents the relative rotation between the sensor 10 and base 9, and has essentially the same phase and frequency as the dither signal component sensor output signal
.theta..sub.S. Then, by gain adjusting the dither pickoff signal, gain means 104 essentially reproduces the dither signal components other than the base motion component that are present in the sensor output signal .theta..sub.S. Therefore, the
corrected pickoff signal .theta..sub.CP has the same characteristics as the dither signal components, other than the base motion component, in the sensor output signal .theta..sub.S. The corrected pickoff signal .theta..sub.CP provided by gain means 104
is presented to both the signal combining means 101 and a multiplying means, 106.
Multiplying means 106 essentially multiplies the input signal .theta..sub.CP by a scale factor representing the ratio of the rotational inertia of the rotational sensor (or gyro), J.sub.G, to the sum of the rotational inertias of the sensor
J.sub.G and the base J.sub.B. Multiplying means 106 provides an output signal .theta..sub.C which represents an estimated value of the base motion component of the dither signal due to dithering based on the value of the sensor motion component
represented by signal .theta..sub.CP.
The multiplication or scaling performed by multiplication means 106 produces an estimation of the base motion component .theta..sub.B of the dither signal from the corrected pickoff signal .theta..sub.CP is the focus of this invention. This
estimate of the base motion component due to dithering is a good one because any coning error components present in the sensor output signal .theta..sub.S will be rejected by the synchronous demodulator 103. Since any coning error component that may be
present in the sensor output signal will not have the same phase as the dither pickoff signal .theta..sub.S, it will not be passed by synchronous demodulator 103 and thus will not affect the gain of gain means 104. In closed loop operation, signal
.theta..sub.CP continues to change until it is substantially equal to the dither signal components other than those due to base motion in the sensor output signal .theta..sub.S.
The closed loop system of the present invention is preferably implemented with a processing means, 110. This processing means 110 performs each of the functions previously described for the first signal combining means 101, second signal
combining means 102, synchronous demodulator 103, gain means 104, and multiplier means 106. Processing means 110 may be a microprocessor, digital computer, or some form of programmable logic device, all of which are well known. Processing means 110
periodically receives digital values from pulse accumulator 100 indicative of the angular rotation of sensor 10 at any instant. In addition, processing means 110 periodically receives a digital value representing the dither pickoff signal .theta..sub.P. A digital value representing the base motion component of the dither signal .theta..sub.B is periodically produced by processor 110.
FIGS. 3 and 4 illustrate the mechanical dithered inertial integrating rate sensor assembly, 119. Sensor assembly 119 includes a mounting base, 9, a case, 120 and one or more flexible, mechanical linkages, 121A, 121B, 122A, 122B, positioned
therebetween. The flexible, mechanical linkages shown in FIG. 3 are represented by dashpots 121A and 121B and springs 122A and 122B. These flexible, mechanical linkages form an isolation or shock mount tending to prevent low frequency vibrations from
the larger more massive case 120 from coupling through the flexible mechanical linkages to the base 9. In addition, these isolation mounts tend to prevent the high frequency dither motion of the base 9 from coupling through these flexible mechanical
linkages to cause case 120 to vibrate. Isolation mounts represented by dashpot 121A and 121B and springs 122A and 122B are well known.
Sensor assembly 119 further includes sensor 10 positioned between an upper sensor block housing, 123, and a lower sensor block housing, 124. Fastening screws, 125 and 126, hold the upper sensor block housing 123 together with the lower sensor
block housing 124 while at the same time attaching the upper and lower sensor block housings 123 and 124 rigidly to mounting base 9. A sensor block support pin, 127, is centrally located and extends between the upper sensor block housing 123 and the
lower sensor block housing 124.
Sensor 10 further includes sensor block 12, the previously discussed output signal generator which is not shown and a dither motor assembly also not shown. Sensor block 12 is supported from the sensor block support pin 127 by three sensor block
support springs, 128A, 128B and 128C. The dither motor assembly is located between the upper sensor block housing 123 and the lower sensor block housing 124 to provide a sinusoidal rotation input to the sensor block 12 for preventing the two counter
rotating beams from tending to lock together.
FIG. 5 illustrates a mechanical equivalent schematic representing the gyro assembly 119, shown in FIGS. 3 and 4. Shown there is the sensor block 12 attached to base, 9, by shock mount represented by a spring 128 and a dashpot, 129. Base 9, in
turn is mounted using a shock mount represented by a dashpot 121 and a spring 122 to a case 120. The shock mounting 121A, 121B, 122A and 122B shown in FIG. 3 for base 9 is represented in FIG. 5 by a spring 122 having rotational stiffness K.sub..phi.
and a rotational dashpot 121 having a friction f.sub..phi.. In a typical inertial sensor assembly, there are three such sensor block assemblies each mounted to the same base 9 such that each sensor assembly has a mutually orthogonal axis of rotation as
shown in FIG. 1.
< | | |
