This paper presents an algorithm to correlate detections from a two-color low-resolution optical sensor. The algorithm is designed for a mosaic sensor consisting of two color detector arrays separated by a fixed distance. The scanning of the sensor provides two sets of detections separated in time by the quotient of the color detector array separation and the scan-rate. The correlation algorithm combines these two sets of data into a composite scan consisting of correlated pairs of two-color detections. The algorithm chooses candidate pairs for correlation by comparing the spatial separation of color pairs with the maximum computed separation. This maximum computed separation is a function of the maximum expected velocities, crossing angles, range, sensor scan-rate, and sensor color detector array separation. The algorithm then applies a redundant correlation resolution technique to eliminate cross-correlations from the list of candidate pairs. After all identifiable cross-correlations have been eliminated, the remaining pairs form the detections list. A star resolution algorithm that is based on an abs'ence of object motion is then applied to identify stars in the detections list. The algorithm is tested against statistical and deterministic point-source detection models. Algorithm precision in number of detections correctly correlated, cross-correlated, and wrongly correlated, and algorithm run-time on a VAX 11/780 are measured.
Estimation of the angular position of a target feature is considered with a special emphasis on a case where a much stronger signature component may "mask" the presence of the feature. Judiciously applied amplitude apodization is shown to be an effective technique for tracking a weak image signal in the presence of a much stronger but undesired signal component. A method is suggested for constructing position estimation algorithms employing apodization techniques.
The components and performance of a prototype high angle sensor with a measured angular resolution of 1 nanoradian at an update rate of 1 KHz are described and experimental results are presented. Calibration methods are discussed and the performance of a differential gas refractometer is presented.
A major concern when considering the design of a ground based precision tracking system, is the optical jitter that will result from various environmental disturbances. This paper considers disturbances including wind, angular seismic, linear seismic, gyroscope noise, electronics noise, and atmospheric turbulence. The disturbances may couple into a tracking system by applying forces or torques directly on the optical components, or by transmitting through the control loops which in turn place jitter into the optical path. The tracking system design can be altered by using different control system configurations to reject the expected disturbances. For instance, if seismic disturbances are large and expected to cause jitter in the optical train it may be desirable to add an autoalignment system to maintain dynamic optical train alignment. A further level of complexity is the selection of a reference for the autoalignment system. It might be a gyroscope stabilized mirror or simply a softly suspended mirror. This paper reports a trade study between the various disturbances and the various hardware configurations that result in some general guidelines when selecting the components in the design of a precision tracking system.
The Hubble Space Telescope (HST) is designed to carry five major scientific instruments to collect imagery, spectrographic, and photometric astronomical data. The Pointing Control System is to achieve pointing accuracies and line of sight jitter levels an order of magni-tude less than can be achieved with ground mounted telescopes. In addition, the HST must be able to acquire and track solar system targets with apparent motion up to 0."21/s. Such targets include planetary satellites, planetary surface features and comets. It is to perform this tracking with an accuracy under 0."03 at the maximum rate. Tracking of solar objects by the Space Telescope accounts for the effects of velocity aberration and parallax, as well as solar targeting a celestial object in a science instrument aperture. The design of the Pointing Control System solar object tracking features is discussed, with emphasis on the special timing and granulation problems inherent with a sampled data, multi-rate digital control system.
The Instrument Pointing System is a versatile three axis gimbal system providing pointing capability in the arc-second range to any type of experiment, requiring precision orientation towards celestial objects. Aside from this prime objective, the system is designed to execute automated acquisition procedures and to follow a pre-defined trajectory, e.g. for landmark tracking. A summary description of the IPS mechanical and electrical hardware configuration with special considerations of the optical sensor package (OSP) consisting of three Fixed Head Star/Sun Tracker (FHST) is given, followed by a presentation of the IPS operational modes, the control loop design, and functional modes and performance data of a FHST. The further chapters of this paper deal mainly with the inertial attitude acquisition of the IPS strapdown system, i.e. the initial attitude acquisition and the operational star identification. The method used for star identification realized in an IPS flight software application program and the selection criteria of candidate stars for the IPS star catalogue are discussed. The operations necessary for initial attitude acquisition as control unit switch on sequence, IPS deployment, slewing to celestial objective direction, and the subsequent attitude acquisition with the related software functions are presented. The approach of the initial attitude achievement using unique bright stars for line of sight (LOS) and roll attitude acquisition is described in detail together with back up methods. Finally the present tracking capabilities of the system are outlined and methods for improvements are indicated.
Experiment scaling is described for sub-resolution, active tracking. It is shown that scaling is feasible for emulating long range active tracking. This allows laboratory scale demonstration of nearly every aspect of the full scale situation. The apertures for the scaled laboratory experiments must be pinhole-size of order 1 mm diameter; this allows the experiments to be virtually insensitive to atmospheric turbulence over controlled laboratory air paths up to 100 meters in length. A discussion of several methods for speckle reduction is also included. If an effective space-to-space field experiment is to be carried out, near-term scaled, laboratory experiments should begin. The field experiment can be fully emulated in a scaled laboratory environment, using much of the actual field equipment, e.g., detectors, intensifier, signal processing, controlled steering mirrors.
An advanced experimental satellite laser ranging station has been established at the Goddard Space Flight Center's 48-Inch Telescope Facility located in Beltsville, Maryland. The goal is to develop a laser ranging system capable of ranging to both low and high earth orbiting satellites with a few millimeters accuracy. To accomplish this goal requires precise pointing of the massive (15 ton) azimuth-elevation mounted Coude telescope since the spread of the laser beam is typically 12 arcseconds. Because tracking is performed under open loop computer control with no feedback from the laser returns (except via operator corrections), the initial acquisition data and the algorithms for computation of the orbit must be accurate enough to maintain the satellite within the narrow laser beam width. The system is currently being used to track the LAGEOS and STARLETTE satellites. In actual operation the pointing is accurate enough so that minimal operator intervention is required to acquire the satellite and maintain the track. This system has performed extremely well in terms of both data quantity and quality. On a clear night the return to fire ratio is typically 80 percent or better throughout a large portion of the pass, with single pulse laser output energies of less than 100 millijoules, and the shot to shot range RMS for most passes is near the one centimeter level.
Testing of the high speed, high resolution optical tracking system intended for use in the Spaceborne Geodynamic Ranging System (SGRS) has been completed at the Goddard Space Flight Center (GSFC). This two-axis gimbaled pointing system, designed to operate from a Space Shuttle platform, is required to point to, acquire, and then track multiple retro-reflector ground targets with arcsecond accuracy. This application requires accurate tracking, high slew rates, and rapid settling times. Laboratory testing of the stepping characteristics of the pointing system for various step sizes and directions showed, in most cases, arcsecond repeatability with little wasted motion, overshoot, or ringing. Worst case RMS tracking jitter was 1 and 2 arcseconds in the roll and pitch axes respectively at the maximum tracking rate of 2° per second. At tracking rates below .1° per second, the RMS tracking jitter was .1 arcsecond or less in both axes. Tracking offsets of .1 arcseconds in the roll axis and .5 arcseconds in the pitch axis were observed throughout the testing. Servo response to tracking commands were recorded for various tracking rates, step sizes, sine wave commands, and simulated SGRS pass data, and will be presented here.
An experiment employing a precision pointing two-axis gimbal has been constructed to demonstrate and substantiate various aspects of precision pointing of a gimballed laser mirror. The main objective of the experiment is to demonstrate the capabilities of achieving simultaneous pointing accuracy and jitter stability requirements. Results of the experiment have shown that both typical and extreme spacecraft disturbances may be significantly attenuated by a fast servo system.
This paper summarizes the technical background and analysis relevant to the performance verification of high-performance airborne telescopes usually operating in the IR spectrum. The purpose of this paper is to communicate an understanding of the state-of-the-art in control systems and performance prediction methodology relevant to ultra-high accuracy optical platforms. This introduction reviews the general control requirements of such systems the motivation for the analysis, the genesis of the techniques used, and summarizes the conclusions of the paper.
The Solar Optical Telescope (SOT) is designed to operate as an attached payload mounted on the Instrument Pointing System (IPS) in the cargo bay of the Shuttle Orbiter. Pointing and control of SOT is accomplished by an active Articulated Primary Mirror (APM), an active Tertiary Mirror (TM), an elaborate set of optical sensors, electromechanical actuators and programmable controllers. The structural interactions of this complex control system are significant factors in the stability of the SOT. The preliminary stability study results of the SOT dynamical system are presented. Structural transfer functions obtained from the NASTRAN model of the structure were used. These studies apply to a single degree of freedom (elevation). Fully integrated model studies will be conducted in the future.
Instruments attached to the payload bay of the Space Shuttle require special attention where fine pointing is involved. Fine pointing, for purposes of this discussion, is defined as sub-arc second pointing. Instruments requiring such fine pointing (Solar Optical Telescope and Shuttle Infrared Telescope, for example) will require two stages of pointing (coarse and fine). Coarse pointing will be performed by a gimbal system such as the Instrument Pointing System (IPS). Fine pointing will be provided by image motion compensation (IMC). Most forms of IMC involve adjustable optical elements in the optical system to compensate for fast transient disturbances. This paper describes work performed on the Solar Optical Telescope (SOT) concept design that illustrates IMC as applied to SOT. The fine pointing requirements on SOT dictate use of IMC at about 20 Hz. bandwidth. It will be shown that the need for this high bandwidth is related to shuttle-induced disturbances. Shuttle-induced disturbances are primarily due to two sources; man push-offs and vernier thruster firings. Both disturbance sources have high-frequency content that drive the IMC bandwidth.
The Solar Optical Telescope (SOT) is a shuttle-based observatory designed to give more detailed views of the Sun than has been possible using Earth-based observatories. It will be able to resolve details on the Sun which subtend 0.1 arc seconds for continuous viewing periods of up to several hours. When budgeted into the various error sources which contribute to the overall image quality, jitter of the telescope can contribute only 0.03 arc seconds rms pointing error over the science observation periods. To meet this stringent pointing requirement, several layers of control systems are utilized.
During the last few years there has been a push to increase shaft resolution as well as pointing accuracy of positioning systems. In order to do this it is necessary to improve the methods of shaft angle encoding. In the process of meeting higher resolutions, several beneficial side results have appeared. This paper deals with these beneficial results as well as the methodology used to gain these results. The system is based on a microprocessor based two-speed combiner that provides a moderately fast conversion time along with a tachometer output. The use of a microprocessor in conjunction with two resolver to digital converters provides a high degree of flexibility. The processor accepts two speed data from the resolvers and produces a binary shaft position. Since this is under software control flexibility is acheived. With proper converters 21 bit resolution is obtainable with rotational rates greater than 100 degrees per second. Other benefits created by the use of a processor are the ability to do software error correction on the shaft angle data to correct for mechanical alignment problems. The approach also provides moderately fast conversion time and with proper converter choice a solid state tachometer is available that provides .1% linearity.
Unmanned free-flying platforms are part of the NASA space station program. Many scientific and operational meteorlogical earth-observing missions are best conducted from polar orbit. The technology required to develop a platform (large spacecraft) to carry a multiplicity of advanced instruments is presently under study. Ensuring the line-of-sight stability and accurate coallignment of instruments mounted on the platform is a fundamental requirement of the structure. Early spacecrafts were small and compact and designed to meet the stresses of the launch vehicle. For practical purposes, these structures could be considered as rigid bodies in orbit. In the space station era it is no longer feasible to design structures sufficiently rigid to make distortions and structural dynamic deflections negligible. For a platform of approximately 4 x 17 meters, carrying a variety of imaging and sounding payloads, an "intelligent structure" is required. We believe it is necessary to investigate active dynamic control of structural resonances. This paper will focus on actuators which must be lightweight, low power and capable of being integrated into the structure without degrading its integrity. The wide dynamic range from dc to 100 Hz may well require several types of actuators. At this time we are exploring novel actuator types and attempting to characterize them for computer simulation and modeling. How their unique features suit this application is described. The objective of providing broadband damping including higher order modes requires modeling the structure with a distributed array of sensors and actuators. Optimization of location and actuator's characteristics to achieve a high level of performance with minimum system weight and power demand is our goal.
This paper describes a potential pointing and tracking system for the Space Shuttle with possible future application to Space Station. In order to accomplish high precision pointing and tracking (at rates up to 2o/s) in the expected disturbance environment, a high bandwidth gimbaled pointing system is required. A soft mounted momentum compensated gimbal system is suggested for this role. A momentum compensated system is inertially reacting, decoupling the control system dynamics from the basebody structural dynamics. This allows a soft isolation stage to he added between the hasebody and the articulation stage, which attenuates high frequency distrubances. In this paper, three configurations are examined: a hard mounted system, a passive soft mounted system, and an active soft mounted system. Analysis demonstrates that the soft mounted systems have superior disturbance rejection properties. The active soft mount allows reduction of the isolation stiffness to zero, and so obtains the highest level of performance.
"Alignment" signifies the process of aligning the optical axes of two cameras. This can be effected in video real-time by means of image localization, as localization of a section of an image in a larger image is termed. An optimal localization procedure consists of filtering in order to decorrelate disturbing parts and of determining the error values or calculating a cross-correlation function. Suboptimal procedures are presented which yield feasible structures. Comparing the procedures shows that the highest localization frequency can be attained by combining standardized correlation with adaptive pre-processing.