The Hemispherical Resonator Gyro for precision pointing applications A. Matthews and D. A. Bauer

Transcription

1 The Hemispherical Resonator Gyro for precision pointing applications A. Matthews and D. A. Bauer Hughes Delco S,vsteins Operations Goleta, California ABSTRACT The solid-state Hemispherical Resonator Gyroscope (HRG) uniquely offers the benefits of small size, extrernely long life, and ultrahigh reliability for spacecraft pointing and control application. The HRG also offers precision performance which can be scaled through control loop and electronics design tradeoffs. Delco Systems Operations has developed the HRG-based Space Inertial Reference Unit (SIRU), which has been designed to maximize the benefits of small size and high reliability and meet the medium grade performance of most spacecraft. The precision perfonuance required for many imaging spacecraft has been achieved through selected modifications to the SIRU design. The result is a precision pointing IRU which maintains the inherent small-size, high reliability characteristics offered by the solid-state HRG. These attributes enable the weight reduction and life extension of precision imaging spacecraft. Modifications to the SIRU design for precision pointing applications have been implemented and characterized. These modifications and test results are described herein. Keywords: Gyro, HRG, precision, pointing, stabilization, inertial, control. BACKGROUND To establish a baseline for discussion of HRG performance enhancements for precision pointing, a brief introduction to the HRG operation and the SIRU system architecture is described below.. HRG description..1 HRG principle of operation The HRG operation is based on the inertial sensing property of a standing vibration wave on a hemispherical body which was discovered and analyzed by British physicist G. H. Bryan in the late 18's. This property is illustrated in Figure 1. The figure shows a standing wave pattern at the rim of a hemispherical body, with two nodal and two antinodal axes and a fixed reference point on the body. When the body is rotated about the input axis (Figure 1), the wave pattern location precesses with respect to the fixed location on the body by a factor of.3. That is, an angular rotation of 9 degrees about the input axis will result in a precession of the vibration axes by 7 degrees. The amount of angular rotation about the input axis of the gyro is determined by sensing the change in the vibration pattern location... HRG construction The HRG has been developed by Delco Systems Operations, along with a range of HRG-based inertial systems. Delco currently produces the HRGI3OY which is shown in Figure. The HRGI3OY contains three primary 18 ISPIE Vol. 466 O $6.OO

2 Standing Point Antinode (1) L1_out Transfer Function Flexing Pattern Precession Angle Figure 1. HRG principle of operation EEEII:: * Figure. HRGJ3OY 94-1 Figure 3. Primary functional components ofthe HRG functional components: the hemispherical resonator, the forcer, and the pickoff, all three made of quartz. These components are illustrated in Figure 3. The resonator, which is the rotation sensing element, is positioned between the forcer and pickoffwith a small gap separating their surfaces. These quartz pieces are metallized so that small capacitors are formed between the resonator, and the forcer, and These capacitors are used for capacitive readout and electrostatic control of the resonator. SPIE Vol. 466 I 19

3 The resonator, forcer, and pickoffare bonded and contained within a sealed vacuum housing. Because the gyro operates in a vacuum, space is a natural operating environment for the HRG. A buffer amplifier circuit is attached to the sealed HRG housing to amplify the capacitive pickoff signals. The complete HRG13OY assembly, including the buffer amplifier, is shown in Figure. The HRG13OY weighs approximately.3 pounds and measures approximately 6 centimeters in height and width. There are no moving parts in the HRG except for the very small amplitude vibration of the resonator, hence there are no mechanisms to wear out...3 HRG readout and control The typical HRG readout and control electronics consist of analog signal conditioning circuits, an analog! digital interface, and a digital signal processor.' A simplified diagram ofthe readout and control mechanization is shown in Figure 4. Four control loops are used in the "force-to-rebalance" mechanization. (There are optional loop mechanizations one of which, the "force-to-rebalance" mechanization, is used in the SIRU system.) The four loop mechanizations are: (1) the phase lock loop to track the natural resonance frequency; () the amplitude control loop for sustaining and controlling the nominal resonator flex amplitude; (3) the quadrature control loop to correct for small mass unbalances on the resonator; and (4) the rate loop to apply a "rebalance" torque to hold the vibration pattern nodal point stationary. I HRG I 13OYHRG I I I ELECTRODE ARRANGEMENT I Forcer 1 II Antinode ELECTRONICS - I DIGITAL CONTROL ALGORITHMS (TTT'\Resonator Il F Node I Phase Lock Loop Control Amplitude Control S Quadrature Control Rate Loop Control L.. _...3 Core SIRU architecture Figure 4. Basic HRG readout and control mechanization The extremely long life and high reliability offered by the HRG make it particularly beneficial for long duration space missions. Delco's Core SIRU uses the HRG to provide 15 years ofcontinuous, highly reliable three-axis IRU operation. The electronics block diagram of the fault tolerant SIRU is illustrated in Figure 5. The SIRU system electronics design is optimized to minimize size, weight, and cost and provides the moderate accuracy which meets most spacecraft requirements. The basic bias and noise characteristics of the SIRU are as follows: Bias Stability Angle Random Walk Angle White Noise.5 deg!h (1G).1 deg/root-h. arc-s!root-hz Enhanced SIRU performance is achieved by optimizing the electronics design for low noise and high resolution. 13 / SPIE Vol. 466

4 Figure 5. SiRUfunctional block diagram 3. PRECISION POINTING REQUIREMENTS While the SIRU performance satisfies the requirements of most spacecraft applications, certain missions require greater precision than the basic SIRU production system offers. In particular, remote sensing spacecraft and space-based astronomy missions often demand extremely low-noise, high resolution gyro performance. The Hubble Space Telescope (HST) is an excellent example of a very precise pointing spacecraft. The gyro performance requirements for the Hubble Space Telescope are very stringent, and are described in Figure 6. Note that the required HST noise characteristics are two orders ofmagnitude below those ofdelco's basic production SIRU discussed previously. Until recently, only spinning-mass gyros have been able to provide the performance demanded by such missions. Due to the limited life and reliability of the spinning-mass gyros, a strong need exists for a reliable solidstate gyro offering the equivalent performance. Analysis has shown that the HRG is capable of HST performance, with appropriate optimization of the system electronics. Angle White Noise.5 5ili Angle Random Walk.11 degali Rate Flicker <.11 degfh Bandwidth Hz Resolution.5 Maximum Rate.5 degls Low Rate deg/h ARW Hz Figure 6. Hubble Space Telescope gyro performance requirements SPIE Vol. 466 / 131

5 4. PRECISION HRG To demonstrate such precise performance of the HRG, Delco has implemented and evaluated a series of electronics and loop enhancements to the Core SIRU design. The HST performance requirements were used as the performance goal for this Precision HRG development and demonstration effort, which has resulted in extremely low noise, high resolution HRG performance without any modifications to the HRG instrument. The precision HRG performance is approaching the HST goal, and further development is continuing. The Precision HRG development approach and interim results are described below. 4.1 Analysis techniques Development of a gyro and associated electronics to meet the requirements shown in Figure 6 requires the utilization of signal analysis techniques and test instrumentation systems that can identify individual noise sources so that these noise sources can be eliminated or attenuated. During the development of the Precision HRG, extensive use was made ofpower Spectral Density and Allan Variance signal analysis techniques (both in the angular rate and angle domain) to separate sensor noise sources, and a back-to-back gyro test configuration to separate out the effects of external motions that fall within the spectrum of interest Power Spectral Density and Allan Variance noise analysis techniques A data record from the gyro or system under test can be analyzed both in the frequency and the time domain. In the frequency domain a Power Spectral Density (PSD) is computed using a Fast Fourier Transform (FFT). When the PSD is graphed, the gyro output forms a structure to extrapolate individual noise characteristics. The same data record can also be analyzed using the Allan Variance technique to compute the uncertainty (variance) as a function ofa varying sampling interval. In a graph ofthe Allan Variance (AV), the gyro noise signal again forms a characteristic structure. In a manner similar to PSD, individual noise sources can be determined from the slope of the noise structure. Figure 7 shows a rate PSD and AV showing the characteristic noise slopes for Angle White Noise (AWN), Angle Random Walk (ARW), and Rate Flicker (RF). Reference 3 gives a detailed account ofboth PSD and Allan Variance techniques and gyro noise sources. Table 1 lists noise sources encountered during the testing of the Precision HRG along with their characteristic slopes when graphed as PSD's or Allan Variances. Table 1 also includes both the gyro rate signal and the integral ofrate (angle) signal. When analyzing gyro noise signals, both rate and angle signals should be taken into account, because on certain noise error sources, it is possible to differentiate between noise error sources that would otherwise have the same characteristics in the angular rate domains. An example of this differentiation is AWN and Quantization Noise (QN). In the rate domain, both have a characteristic slope of+ in PSD asymptotic and -1 in Allan Variance asymptotic. When analyzed in the angle domain, AWN in Allan Variance asymptotic has a slope of-1/ and QN has a slope ofzero (). Thus, AWN can readily be distinguished from QN in the angle domain, but not in the rate domain. This distinction is particularly useful in separating A/D quantization errors from electronic white noise. The authors also prefer angular data analysis techniques, because they give a true indication of the attitude errors that would be seen after the attitude integration process. The above techniques proved invaluable during the development of the Precision HRG for the identification and modifications of noise error sources. 13/SPIEVo!. 466

6 4. SIRU baseline Delco's Core SIRU system was used as the basis for implementation and evaluation ofperformance enhancements for the Precision HRG. A SIRU system was reconfigured with two back-to-back HRG's in place of the normal SIRU gyro platform, and was tested to establish a performance baseline prior to implementing any performance enhancements. Figure 8 shows the rate PSD for the baseline Core SIRU. 4.3 Modifications to SIRU to give precision pointing capability As mentioned earlier, the SIRU was specifically designed to meet medium accuracy pointing requirements. To PSD AV f t Figure 7. Noise characteristics in PSD and AVgraphs Nomenclature Angle White Noise Angle Quantization Flicker Angle Mnemonic AWN AQ AF Characteristic PSD Slope Characteristic Allan Variance Slope Angle Rate Angle Rate / RateWhiteNoise (also known as angle random walk) RWN (ARW) - 1/ -1/ Rate Quantization RQ - 1 Flicker Rate RF Angular Acceleration White Noise (also known as rate random walk) AAWN (RRW) -4-1> 1/ Angular Acceleration Quantization AAQ -4-1 Angular Acceleration Flicker AAF Table 1. Noise sources and associated characteristic slopes, including gyro rate signal and the integral ofrate (angle) signal SPIE Vol. 466 I 133

7 !1 1 ;c;-' 1 Hz Figure 8. Baseline SIR Uperforinance 1 minimize cost, size, weight, and power, system trade studies led to a system electronics design which does not fully utilize many ofthe features inherent in the HRG. This design did not take advantage ofthe full HRG performance. To achieve precision pointing performance these unused features of the HRG were re-introduced, yet the basic Core SIRU mechanization and features were maintained. A number of relatively simple features were added to the Core SIRU, each providing incremental improvement in performance. These changes are listed in Table. Feature Core SIRU Precision Pointing SIRU Remarks Number of rate (nodal) pickoff pads utilized One pair () Two pairs (4) Improves SNR by Analog signal conditioning circuit Differential circuit to combine two pairs of pickoff pads Improves common mode noise rejection Amplitude of standing wave t in 4 1t in Improves SNR by Resonator bias voltage -1 V - V Improves SNR by Rate signal demodulation Analog rate demodulator Digital rate demodulator Improves bandwidth Maximum torquing rate 1 deg/s (36, deg/h) deg/h Lowers gain ofhrg to increase rate sensitivity Pickoffbuffer amplifier Low cost buffer Improved low noise pickoff buffer Reduces electronic noise Table. HRG/SIR U enhancementsfor precision pointing performance 134 ISPIE Vol. 466

8 The results of these modifications drastically improved the HRG performance from the basic Core SIRU. The performance ofthe enhanced SIRU, after introducing the modifications listed above, is shown in the PSD of Figure 9. This improvement in performance is between I to orders ofrnagnitude improvement over a Core SIRU. This improvement was achieved without any changes to the FIRG instrument. This data illustrates the scalability of HRG performance, which allows for a high level of flexibility in system design tradeoffs for various mission requirements with a common HRG instrument. Figure 9 also shows a photograph of the Precision HRG development/demonstration unit which contains two back-to-back HRG's and all support electronics. 4.4 Back-to-back gyro testing Considerable testing is required to demonstrate that a precision pointing gyro meets specifications. This testing is performed in a tenestrial environment; motion associated with the terrestrial environment will also be measured when testing very low noise gyros. Ground motion makes it extremely difficult to demonstrate the gyro's true performance. One solution to baselining the ground test environment is to conduct final performance testing at a test site which is known to exhibit low seismic motion, such as the Central Inertial Guidance Test Facility (CIGTF) at 1-lolloman Air Force Base. However, this site may require that the performance testing of the gyro be performed far away from its source of manufacture, assembly, and test, which adds cost and time. Performing back-to-back gyro testing eliminates the requirement for a low terrestrial noise requirement. Delco's back-to-back testing approach for Precision HRG development testing is described below. io io5 - io i io-3 i- -1 Hz 5174 Figure 9. Precision HRG development/demonstration unit and PSD of enhanced HRG performance SPIE Vol. 466 / 135

9 Gyro Noise G,7A Gyro Noise GB input axis L4 lb Gyro A - external motion EM Gyro A Output GA Gyro B Output GB GflA, GflB GA GB GA+GB GA-GB gyro uncorrelated random noise. Gfor gyro family. GflA-EM GflB+EM...() GflA+GflB (EM\ + GflA x G GflBJ G -EM _+ Figure 1: Gyro back-to-back testing (It is assumed that GflA and GflB are uncorrelated noise.) Figure 1 describes the back-to-back test technique where the two gyros under test are mounted with their input sensing axis in opposite directions. Thus, terrestrial motion will appear as an opposite sign between the two gyros. If the data from both gyros is collected synchronously, then, by summing the gyro outputs, terrestrial noise up to the sampling time interval is removed. Measurement of the actual terrestrial motion (TM) is obtained by differencing the two gyro outputs. Obviously, the noise inherent in each gyro sensor will show up in the sum and difference data. Figure 1 1 shows the PSD's ofback-to-back Gyros A and B and their composite (A+B)1 from a typical test. Notice that the high frequency noise components shown in the Gyro A and Gyro B outputs have been completely removed by (A+B)/ processing. The (A+B)/ PSD shows the combined noise characteristic of Gyros A and B, and these are usually matched. Therefore, this PSD is one-halfofthe gyro family noise characteristics, because 136 ISPIE Vol. 466

10 these components are not random noise events, but are probably associated with induction motors (air conditioning, blowers, and so on). A plot of (A-B)/ can be used as a measurement of the environmental noise plus I I-f ofthe combined Gyro A and B random noise. Since the gyro has noise characteristics that should follow a "family" characteristic, certain levels of environmental motions are easy to identify. The back-to-back test technique was used during development testing of the Precision HRG at Delco, which is located in seismically active California. The back-to-back test technique was validated during a test conducted at the low-noise facility of CIGTF, Holloman Air Force Base, and the HRG exhibited performance similar to the backto-back results at Delco. Delco's conclusion is that considerable time and effort can be saved by using the back-toback test technique for precision gyro development and testing. c'1 N ) C, 1' 1.1 1' 1' 1' t :...: I ) :i I :i:i i : i i::i: : :. : : ' c,1 N ) C, -D 1' :i:ii..!.:.:....!! I : ::: i : :::.! 1' 1 1' 1' 1' 1 1 1' Hz Hz GyroA Gyro B i i i : --:!-:-r;1 ::j z :;i I I -u:r-. TTTT: ' 1' io- 1 1' 1' 1 ' 1W' (M-B)/ Figure 11. PSD ofback-to-back Gyros A and B aizd their composite SPIE Vol. 466 I 137

11 4.5 Motion detect test In the quest to reduce gyro noise sources, overzealous signal processing techniques can be developed that will reduce noise, but can greatly distort the true inertial angular motion signal. During HRG testing at Hollornan, AFB, a test was conducted to determine the minimum sensitivity of the HRG. A motion sensitivity test was developed by NASA-Goddard Space Flight Center, which consisted ofa rotating unbalanced motor mounted on the baseplate supporting the gyros under test. By varying the frequency of rotation and the amount of unbalance of the motor io 1-1_ io _ N ) C) 1-1 _ 1 1 N io 1-s IIl v Gyro A io-3 i- -1 i6 Hz 5175 io d - 1 Hz 5176 Gyro B 1_ 1-1 N 1 - ) C) o7 (A+B)/ io 1c.1 Hz 1 io 5177 Figure 1: PSD ofthe Precision HRG hit/i I inilliarc-second input 138 ISPIE Vol. 466

12 shaft, a varying sinusoidal motion disturbance was applied to the system. Seismic tiltrneters were used to independently measure the individual disturbance; however, these devices were limited in accuracy to approximately 1 to 4 rnilliarc-seconds. Figure 1 shows the PSD for a sinusoidal disturbance ofapproxirnately 1 rnilliarc-second. Notice that both Gyros A and B detect the motion and that the (A+B)/ processing removes the disturbance, thus confirming the effectiveness ofthe back-to-back test techniques. This test verified that the Precision HRG has very high resolution and can sense an inertial input ofless than 1 milliarc-second. 5. SUMMARY Beginning in July 1994 Delco initiated the development ofan HRG-based inertial reference unit capable of meeting the precision pointing accuracy requirements of a spacecraft such as the Hubble Space Telescope. Using the mature design ofdelco's Core SIRU product as a baseline, substantial improvement in performance was made by a series ofsmall incremental modifications to the gyro operating characteristics and its support electronics, without any changes to the HRG instrument itself. The result ofthe development has demonstrated that the solidstate HRG can provide high precision performance for remote sensing, Earth observation, and scientific satellites while continuing to offer HRG's inherent benefits ofhigh reliability, long life, small size, and low power. In addition, back-to-back gyro testing has been demonstrated effective for precision gyro testing and substantially reduces the development test time and costs. 6. ACKNOWLEDGMENTS The authors acknowledge NASA-Goddard Space Flight Center (GSFC) and Lockheed Missiles and Space Company (LMSC) for definition ofthe Hubble Space Telescope gyro performance requirements and for support in conducting the Precision HRG study. The authors acknowledge the Central Inertial Guidance Test Facility (CIGTF) for test facilities and instrumentation support for Precision HRG testing at Holloman AFB. The authors recognize the Precision HRG engineering team of Hughes Delco Systems Operations for their outstanding effort in the development of the Precision HRG. 7. REFERENCES 1. D. Wright and D. Bunke, "The HRG as Applied to a Satellite Attitude Reference System," Guidance and Control 1 994, by R. D. Cuip and R. D. Rausch, Volume 86: "Advances in the Astronautical Sciences," pages 57 to 63, American Astronautical Society (AAS), NASA-Goddard Space Flight Center, Hubble Space Telescope Flight Systems and Servicing Project, "Key Performance Requirements Matrix for '99 Servicing Mission Rate Gyro Assemblies," 18 March G. W. Erickson, "An Overview of Dynamic and Stochastic Modeling of Gyros," Institute of Navigation 1995 National Technical Meeting Proceedings, January SPIE Vol. 466 / 139

13 本文献由 学霸图书馆 - 文献云下载 收集自网络, 仅供学习交流使用 学霸图书馆 ( 是一个 整合众多图书馆数据库资源, 提供一站式文献检索和下载服务 的 4 小时在线不限 IP 图书馆 图书馆致力于便利 促进学习与科研, 提供最强文献下载服务 图书馆导航 : 图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具