As spacecraft begin to utilize higher performance flight computers, more processing is being performed on-orbit. This paper discusses four tpics that support on-orbit star processing using multi-star star trackers. The first topic addresses the considerations for creating an application specific star catalog and presents a newly developed, computational and memory efficient star catalog storage technique named the Sperical Rectangles Method. The Spherical Rectangles Method is used to extract a subset of stars from a very large (5000 or more stars) star catalog. This subset of stars is used as the candidate set of stars to support star identification and directed search commanding. The second topic examines star identification techniques and discusses their applicability to multi-star star trackers. Each of the star identification techniques discussed has advantages and disadvantages depending upon the star tracker operational mode and tracking status. No single technique is optimal for all scenarios, so a combination of the techniques is used, depending upon tracking status, to ensure a higher probability of successful star identification. The third topic is a comparison of directed search commanding of the star tracker versus full field acquisition. Also, a directed search technique is presented that determines which stars in the star tracker field of view should be tracked to support a more accurate attitude determination solution. In this technique, consideration is given to star brightness, star separation, and computational efficiency. The fourth topic identifies several measurement errors such as data latency, velocity aberration, and parallax that should be compensated for depending upon attitude knowledge requirements.

The Cassini spacecraft will perform a detailed examination of the Saturnian system, including the release of a probe to study Saturn's largest satellite, Titan. The star tracker for the Cassini mission must provide accurate data during the entire flight including four years of measurement in a harsh radiation environment. The star tracker will provide autonomous star identification over the entire celestial sphere using a 4,000 entry on-board star catalog. Three axis attitude reference will be determined by measurements of two to five stars in the tracker field of view which will allow the gyroscopes to be powered off during the cruise phase of the flight. When the gyros are operational, attitude updates will be provided. The Cassini star tracker consists of a CCD based star camera, called the stellar reference unit (SRU), which is being designed and built by Officine Galileo. The operation of the SRU, including functional modes, exposure times, and areas of the CCD to digitize is under the control of the Cassini Attitude and articulation control subsystem (AACS) flight computer (AFC). The raw digital pixel data is transmitted from the SRU through a dedicated direct memory access (DMA) interface to the AFC memory for subsequent processing. All pixel processing and centroiding is performed within the AFC. Once the initial attitude has been determined, the AFC algorithms will choose which stars within the SRU field of view to track in order to maintain attitude knowledge. The SRU will have a 15 degree field of view and will provide 60 (mu) rad (3 (sigma) ) 2-axis position measurement accuracy for stars of approximately 6.05 visual magnitude and brighter. The required 1 mrad (3 (sigma) ) twist accuracy is provided by star separation.

This paper summarizes our efforts towards the realizaton of an intelligent autonomous tracking and pointing system for space applications. A powerful 3D graphic software testbed has also been developed to simulate a likely scenario of autonomous tracking and pointing operations during a planetary flyby mission.

The three fine guidance sensors on-board the Hubble Space Telescope are the first white-light amplitude shearing interferometers to be used for space platform guidance, control, and astrometry. Two fine guidance sensors (FGS) under fine lock control now maintain spacecraft pointing precision to within 7 milliseconds of arc rms over the majority of each orbit. Fine guidance sensor control optimization techniques have yielded significant improvement in tracking stability, integrated performance with the pointing control system, loss-of-lock statistics and astrometric accuracy. We describe the optical interferometer, based on the Koester's prism design. We include a discussion of the instrument calibration status, the FGS fine lock performance design enhancements, pointing control system design enhancements, and ground software techniques appropriate to jitter removal in astrometric data. The combination results in marc sec precision astrometry.

F-Sat is a standard spacecraft bus which will support future LMSC programs, both commercial and government. The F-Sat bus has been designed with the adaptability to satisfy any applications with medium-weight payloads. This paper discusses F-Sat's design and a recently completed test which demonstrated the integration of flight electronics with attitude determination and control system sensors and actuators. With the successful completion of this demonstration. F-Sat has taken one more step toward realizing the goal of low cost access to space.

This paper shows the use and capability of magnetic desaturation for momentum management of relatively low altitude orbiting satellites. A multi-dipole model is presented which is a convenient representation of the Earth's magnetic field. An analytic representation of desaturation capability is derived that is intended for the sizing of magnetic systems. A unique method of magnetic desaturation is presented that uses the full capability of the magnetic system actuators. Simulation results are presented that show the effectiveness of employing this magnetic desaturation method on a satellite subjected to various external disturbances. These results are then compared to the results predicted by the analytical method.

This paper describes the design and implementation of digital filter algorithms to achieve precision pointing and tracking performance with dithered ring laser gyros. Test results from the real time implementation of these algorithms are presented.

The purpose of this paper is to describe the design and use of Star Field Simulator technology for ground-based check-out of star tracker systems in a closed loop environment. An operational overview, system setup, specifications and acceptance test results, as well as the observed system limitations are reported.

The paper describes a semi-autonomous control system on board a small mobile robot intended to explore planetary environments with minimal human supervision. The algorithm performs dead reckoning, hazard mapping, trajectory planning, motion control, health monitoring, and supervisor telemetry. A remote operator station provides a graphical interface for real time telemetry, guidance, and performance tuning. Design emphasis was placed on processing efficiency and robust performance for implementation with inexpensive sensors and microcontrollers. Of novel contribution are a high-level behavioral task architecture and a reactive trajectory planning routine that explicitly incorporates steering and safety constraints. Simulation and field test results under various operating conditions are presented.

Quality of an image movement sensor may be evaluated by two parameters: precision (measurement error) and nominal speed (frequency bandwidth), both depending on the image moving and technical characteristics of the sensor. Measurement error being given, those characteristics may be optimized to provide the widest frequency band. In this paper such an optimization problem is considered for the optoelectronic sensor, which is a constituent part of an image stabilization system for a spaceborn high-resolution earth's surveillance telescope, and some results of computer modeling of the sensor are presented.

A microprocessor controlled star tracker system using an area array charge coupled device (CCD) as detector has been developed for the attitude determination of remote sensing satellites. A pair of sensors with field of view (FOV) of 8 deg x 6 deg mounted in skewed configuration in the 3 axis stabilized satellite is capable of detecting stars up to 5.0 visual magnitude giving an attitude determination accuracy of better than 0.01 deg. This system is developed at the Laboratory for Electro-Optics Systems (LEOS), Bangalore and intended to be flown aboard Indian Remote Sensing Satellite (IRS-1C) scheduled to be launched by the middle of 1995.

On-board computer for Star tracker is designed for providing high accuracy attitude information to improve the location accuracy of the high resolution imagery of the Indian Remote Sensing Satellite (IRS-1C) scheduled to be launched by the middle of 1995. The on-board computer is based on a 16 bit microprocessor CMOS 80C86, which has an address capability of 1 MByte, operating at 3 MHz. The processor operates in minimum mode of configuration without numeric data processor (NDP). The on-board computer controls dual star sensors based on area array charge coupled device (CCD). The on-board computer searches for star in the field of view (FOV) by dynamic exposure sequencing and auto- thresholding. A background estimation is also done to avoid stray light in the sky. Once a star is searched and is within the FOV of the sensor, the star is continuously tracked until it leaves the FOV of the sensor, while taking care of the new star entry. The processor dynamically relocates the track window as the spacecraft moves in the orbit. The processing consists of centroid computation of star image by means of interpolation technique, valid star identification, attitude determination, data formatting for telemetry interface etc. The design details of the on-board computer is presented in this paper.

A portable, ground based angle-only IR tracking system, designated GBATS (ground-based acquisition and tracking system), is described. GBATS is designed to passively provide 2D angular track data on boosting targets of interest. Upon target launch, GBATS automatically performs target acquisition, track, and thrust termination (burn-out) time and location. Multiple target acquisition, target obscuration treatment, and optional radar cueing are features of GBATS. System and sub-system descriptions for all major hardware and software components are given in this paper. Test results on targets launched from the White Sands Missile Range as well as future plans are also discussed.

The design of a real-time, ground-based, infrared tracking system with proven field success in tracking boost vehicles through burnout is presented with emphasis on the software design. The system was originally developed to deliver relative angular positions during boost, and thrust termination time to a sensor fusion station in real-time. Autonomous target acquisition and angle-only tracking features were developed to ensure success under stressing conditions. A unique feature of the system is the incorporation of multiple copies of a Kalman filter tracking algorithm running in parallel in order to minimize run-time. The system is capable of updating the state vector for an object at measurement rates approaching 90 Hz. This paper will address the top-level software design, details of the algorithms employed, system performance history in the field, and possible future upgrades.

The Midcourse Space Experiment (MSX) satellite is a DoD sponsored spacecraft currently under development that will collect phenomenology in a variety of wavebands. MSX is required to perform on-orbit closed loop tracking of various objects of interest including not only dedicated targets but also targets of opportunity and atmospheric phenomena such as stars, auroral surges, and cloud structures. The onboard tracking system consists of, in part, (a) redundant tracking processors that determine, from sensor data, the trajectory of the object(s) and the desired attitude and rate to maintain observation, and (b) several tracking sensors, in particular, a suite of ultraviolet, visible, and spectrographic imagers and image processor incorporated into one instrument (called UVISI). This instrument is capable of isolating potential targets from various and possibly cluttered backgrounds and of providing both scientific and tracking data. We discuss the image processing algorithms implemented in the UVISI instrument image processor subsystem onboard MSX. We discuss details of the MSX tracking processor (TP) algorithms that provide either open-loop pointing or closed- loop tracking. Included are discussions of TP's star removal algorithm via frame-to-frame data association that renders the probable target based on inertial motion.

The low cost space structure pointing experiment investigates whether significant elements of acquisition, tracking, and pointing can be demonstrated using a representation of a spaceborne optical telescope. Specifically, the feasibility of accurately emulating a large space optical telescope (constituting a spatially correct sparse optical payload) has been investigated. This result is an experiment composed of a few representative critical `sparse' elements incorporated in a small satellite launched from a space shuttle as a hitchhiker payload. Use of sparse elements minimizes the cost, weight, and size of the experiment, while the hitchhiker approach substantially reduces launch costs. After orbital insertion, the stowed sparse components will be deployed to the correct distances with representative structural dynamics. They represent a spacecraft- borne large optical telescope that can be pointed and controlled with the desired accuracy. The experiment will demonstrate the accurate precision pointing and control of four meter class large optical systems in space. It is only in space that one can obtain a realistic test of precision pointing and control and structural dynamics.

The satellite imaging experiment (SIE) is a tracking and imaging experiment conducted during the last half of 1993 (June through December). We have obtained results from a high fidelity simulation called the time-domain analysis simulation for advanced tracking (TASAT) and also from the experiment itself that demonstrate closed loop passive tracking of stars and satellites. TASAT accurately predicts the residual track error for these objects by modeling the detailed physics of tracking through the atmosphere. In particular, an `orbit' appropriate to a star or satellite, an image rendering function, atmospheric point spread functions in the presence of adaptive optics, detailed sensors with noise, and high bandwidth active control loops all combine inside TASAT in a coupled, realistic fashion to predict active and passive cross sections, atmospheric tilt and higher order degradations, and residual track errors. We will discuss the present state of the simulation, results from TASAT that are germane to the SIE, and results from the experiment itself.

This paper analyzes the impact of satellite vibration on acquisition time and acquisition probability for a charge coupled device based acquisition scheme. An approximation is derived for mean time to acquisition. Numerical examples of acquisition time and acquisition probability as a function of the signal-to-noise ratio are presented. The required beacon power onboard the spacecraft is also evaluated based on expected space to space link characteristics.

Loss of lock on background regions or target-like objects (clutter) is a major problem for imaging target trackers. Although gated trackers naturally resist clutter interference until the threat is within the track gate, they do not typically predict the impending perturbation. Also, classical trackers exercise a drastic response to the eventual detection of clutter within the track gate--they coast, totally ignoring current position measurements while propagating an old target rate. If the target accelerates during coast, loss of lock is very likely. We present an algorithm for detecting and tracking clutter objects in the tracked scene and modifying the loop gains in the primary target tracker. This method rejects breaklock due to clutter interference while remaining responsive to target maneuvers.

A novel wavelet transform based clutter measurement technique correlates remarkably well with imaging tracker performance. To determine the robustness of automated tracking algorithms we need a quantitative measure of ground clutter complexity. Current tracking and detection system performance is often specified against signal-to-noise ratio, which is sensor related, and provides no information regarding the scene content. A signal-to- clutter ratio (SCR) can specify the degree to which the target signal is discernible from the background. In the context of electro-optical imaging systems, clutter is defined as objects or scene phenomena that interfere with target detection and tracking. Scene clutter is endemic to imaging systems, yet limited useful work has been performed to measure it. Clutter complexity is usually determined subjectively, and in reference to the object of interest. We devised a clutter measure that is quantitative and also independent of the object of interest. Based on this measure, we also developed a SCR, which is used to analyze detection and tracking performance, and allows prediction of target tracking performance in given clutter levels for particular target signatures. We present results of imaging target tracking performance with respect to target signatures for a given target in various clutter levels.

This paper describes the measuring and modeling of jitter effects for a manportable missile system. We present a tracking approach which reduces sensitivity to jitter induced image degradations. The tracking approach described consists of small and large target tracking algorithms. The small target algorithm reduces interlace effects by extracting target features which are relatively insensitive to interlace errors. Adaptive threshold selection and target feature extraction will be described. A sparse correlation large target tracking algorithm will also be described.

Today's rapid technological developments and limited defense resources require the armed forces to upgrade their current weapons with the newly available sensors. These sensors often have lower production costs, higher reliability, as well as improved performance. Charged coupled device (CCD) passive imaging sensors provide these improvements over the older style vidicon imaging tubes. In addition, the device's sensitivity in the infrared portion of the spectrum offers a potential benefit of penetrating haze and aerosols inherent in a battlefield situation. The theoretical improvements cannot be realized without carefully integrating the new technology with the existing aircraft displays as well as considering the human factors impact of the weapon system officers interpretation of the very near-infrared image. This paper describes the approach, problems encountered, and results of evaluation of a new CCD seeker for the AGM-130 weapon system.

The initial development of a passive, automated system to track bullet trajectories near a target to determine the `miss distance,' and the corresponding correction necessary to bring the following round `on target' is discussed. The system consists of a visible wavelength CCD sensor, long focal length optics, and a separate IR sensor to detect the muzzle flash of the firing event; this is coupled to a `PC' based image processing and automatic tracking system designed to follow the projectile trajectory by intelligently comparing frame to frame variation of the projectile tracer image. An error analysis indicates that the device is particularly sensitive to variation of the projectile time of flight to the target, and requires development of algorithms to estimate this value from the 2D images employed by the sensor to monitor the projectile trajectory. Initial results obtained by using a brassboard prototype to track training ammunition are promising.

This paper proposes a track fusion algorithm for similar and dissimilar synchronous sensors. It is shown that under communication requirements between the fusion center and the local remote station, the proposed algorithm is optimal in the minimum mean square sense.

An experimental airborne target acquisition and tracking system is developed and tested from an aircraft platform. In this paper, assessment and estimation of the system performance characteristics in flight, in regard to (a) detection range and TV sensor resolution, (b) tracking rates (elevation and azimuth) encountered and tracking accuracy achieved, and (c) residual sightline jitter, are presented based on the recorded aerial imagery. The above estimates are made by using trigonometric expressions and details such as flight altitude and speed, sightline orientation, sensor's field-of-view, steering limits, time-on-target, target dimensions and its image fluctuations.

The extracting of multiple targets in a field of view (FOV) is one of the key problems of the acquisition, tracking, and pointing technique, where a large number of computation and complicated operations must be done. In this paper, a new real- time multiple target centroid extractor is presented. The basic idea is that by using hardware circuit performing image segmentation and horizontal neighborhood analysis, the interesting target pixels are extracted from the raw image, then a digital signal processor carries out the vertical neighborhood analysis of target pixels and calculation of multiple target parameters. An algorithm of target pixels neighborhood analysis is presented for both convex and concave target shapes. A prototype system has been built which can extract parameters of about 100 targets in each FOV, offering centroid and average intensity for each target in the FOV from video signal in real- time.

We have developed a high fidelity seeker simulation including point spread function, pixel fill factor, read noise, background radiation, fixed pattern noise, and smearing due to relative platform/target motion, which is fully described in a previous paper. In this paper we report on enhancements to this model and integration of the seeker model with a six degree of freedom (6dof) missile model. The enhanced seeker model is applied to a photon starved scenario and used to evaluate the tradeoff between a high quantum efficiency, low noise CCD and a photon counter exhibiting only photon noise but significantly lower detection efficiency. It is shown that even very small amounts of read noise bring significant track accuracy penalties in low background situations. The 6dof simulation employs an object oriented architecture and is implemented in C++ for Windows. It simulates target position and missile (or observing platform) position and attitude versus time, as well as the missile guidance and control systems. This allows interactive tradeoffs between seeker design and guidance and control system design to be evaluated early in the design cycle. An example of such a tradeoff is presented.

This paper describes an extension to the formulation of the equations of motion for a system of connected rigid bodies introduced previously. Formerly, the multibodies were restricted to either hinge or slide connections. Where applicable, the formulation has been extended to include full six degree-of- freedom (DOF) motion capability between connected bodies. In particular, consecutive bodies or lumped masses can be connected with 6 DOF springs and/or dampers. This capability will allow the user to integrate key flexible body dynamics with multiple rigid body dynamics. The connection topology is still limited to be that of a tree. Extending the topology to a full array connectivity will be investigated in the future. Once the topology restriction is removed, this work will incorporate full finite element modeling capability. This paper develops the modified vector operator equations of motion which give a set of nonlinear differential equations that are amenable to standard numerical integration techniques. This updated formulation has been implemented in FORTRAN code and embedded in several simulation packages being used to generate multibody spacecraft dynamics. For example, using a control system-based simulation software package interfaced with this multibody dynamics code, the tool can be used to include several key flexible body attributes. This tool allows one to rapidly input the dynamic model of a multibody system by simply entering the mass, geometry, spring/damper and connection (location of hinges and/or slides) properties. This provides the engineer the capability required to implement detailed simulation work with rapid turnaround. By removing the need for the timely task of developing the dynamic models of a spacecraft, the control system engineer can concentrate on designing the control system. Many different spacecraft configurations have been studied with this tool. An example of a spinning communication satellite deploying its wings (communication equipment) will be presented. Two modes will be included for each of the solar arrays. These results will be used to verify the effects of the solar array modes during the wings deployment.

Movement of an optical element from its nominal mounting position can result in shifts in the pointing direction and image orientation of an optical system. This movement can be either static (due to mounting misalignment), or dynamic (due to structural vibration). The structural influence coefficients map the translational and rotational motions of lenses, mirrors, prisms, and detectors to changes in line-of-sight direction and image orientation. Influence coefficients are used by the structural dynamicist to predict image jitter from structural finite-element models and by the mechanical designer to tolerance the element-support structure. A method for calculating the influence coefficients for plane-mirror systems is presented. The method includes a working definition of an influence coefficient, a general process for coefficient extraction, and an algorithmic approach to numerical calculations. The method described allows ready calculation of influence coefficients without tedious ray- trace perturbations or dedicated optical analysis software.

A primary cause of degraded performance in pointing and tracking systems is the jitter in its line-of-sight. This jitter is caused by the residual angular motion of the stabilized platform within the system. A major contributor to this residual motion if the gyroscope noise. Thus, to further reduce angular jitter, lower noise gyroscopes need to be selected, generally at a premium cost. Another approach is to electronically enhance the accuracy of the gyroscopes (by suppressing measurement noise) before their outputs are fed into the stabilized platform control system. Optimal filtering techniques can be used for this purpose. The goal is to estimate the platform motion so that the calculated value is closer to the actual value than the measurement is. Enhanced performance is obtained at the expense of added complexity, but in many cases this approach may prove to be more economical than resorting to more precise and costly lower-noise gyroscopes. This paper presents a novel Kalman filtering method that provides more accurate angular motion estimates than the measured values. The effectiveness of this method is evaluated through a computer simulation case study. The simulation demonstrates that the new approach yields excellent 3D angular velocity estimates, very small mean-square-estimation errors, and over a 5 to 1 improvement (in the mean-square sense) over angular velocity measurements obtained from 3 orthogonal gyroscopes. The enhanced 3D angular velocity estimates can be fed into the platform stabilization control system, rather than feeding raw gyroscope measurements, significantly reducing the contribution of gyroscope noise toward the overall jitter in a stabilized platform. This would permit a relaxation on gyroscope noise specifications, which could lead to substantial savings, while maintaining the same error budget.

Tip/tilt mirrors are widely used to stabilize astronomical images during integration because tip/tilt image stabilization systems provide a large gain in image quality with a relatively simple control system[KM93,Mc93,0193]. In solar vector magnetographs the polarization analysis section generally precedes the fip/tilt mirror to avoid systematic polarization errors[Ru88,R091]. This causes a magnification of the apparent pointing errors so that the dynamic range requirements for a tip/tilt mirror are multiplied by the magnification. We have used tip/tilt mirrors based on high voltage piezo electric stacks. These units have relatively limited throw of +/- 65 arcsec, and require stack voltages of 0 to 1000 volts. In addition these units displayed a high Q resonance around 200 Hz which limited the stable closed loop image stabilization bandwidth to around 20 to 30 Hz. The tip/tilt system in our ground based instrument at Sacramento Peak Observatory has always been limited by the limited dynamic range of the tip/tilt mirror[St90,St92].

The new flexure-beam micromirror (FBM) spatial light modulator devices developed by Texas Instruments Inc. have characteristics that enable superior acquisition, tracking, and pointing in communications and other applications. FBM devices can have tens of thousands of square micromirror elements, each as small as 20 microns on a side, each spaced relative to neighbors so that optical efficiency exceeds 90 percent, and each individually controlled with response times as small as 10 microseconds for piston-like motions that cover more than one-half optical wavelength. These devices may enable order-of-magnitude improvements in space-bandwidth product, efficiency, and speed relative to other spatial light modulator devices that could be used to generate arbitrary coherent light patterns in real time. However, the amplitude and phase of each mirror element cannot be specified separately because there is only one control voltage for each element. This issue can be addressed by adjusting the control voltages so that constructive and destructive interference in the coherent light reflected from many elements produces the desired far field coherent light pattern. Appropriate control voltages are best determined using a robust software optimization procedure such as simulated annealing. Simulated annealing yields excellent results, but it is not real time (it may require hours of execution time on workstation-class computers). An approach that permits real-time applications stores control voltages determined off-line by simulated annealing that produce key desired far field coherent light beam shapes. These stored results are then used as training data for radial basis function neural networks that interpolate in real time between the training cases.

A method of calculating numerically the optical transfer function appropriate to any type of image motion and vibration, including random ones, has been developed. This method has been verified experimentally, and the close agreement justifies implementation in image restoration for blurring deriving from any type of image motion. The goal of this research is to recover the original image from its degraded version. There are many methods of image restoration based on the point spread function. One of the common methods is the Wiener filter. Here, some image restorations of synthetic degradation and physically degraded images are presented, based on a constrained least squares improvement of the original Wiener filter. The key to restoration is determination of the optical transfer function unique to each particular image motion and vibration.

An experimental tracking system of fast steering mirror (FSM) is described in the paper. The system consists of a FSM, an E-O imaging sensor, an image processor and a host computer. Data of tilt are extracted from video signals of the imaging sensor by the image processor. These data are then led to the host computer. This computer executes the control program and leads the output control signal to the FSM through D/A converters. A PZT driving FSM is designed. The mirror diameter is 100 mm, tracking range is +/- 582 (mu) rad and the structure resonance frequency is 430 Hz. Digital PID control algorithm is used. In order to improve the characteristics of the acquired procedure, the control parameters are varied by the tracking error and time, and these parameters are self-optimized on line. Quadratic curve interpolation algorithm is used to expand the system bandwidth. The image processor is DSP based. Data of target centroid are extracted for estimating the tip-tilt. In order to reduce the processing, pyramid structure is used. Subpixel locating techniques are adopted. Close-loop experiments with RS-170 camera and a Reticon camera as sensor have been carried out separately. These experimental results are analyzed in the paper.

This paper presents a solution to a benchmark problem for tracking maneuvering targets. The benchmark problem involves beam pointing control of a phased array (i.e., an agile beam) radar against highly maneuvering targets. The proposed solution utilizes an interacting multiple model (IMM) algorithm that includes a constant velocity model, a constant thrust model, and a constant speed turn model. The output error covariance of the IMM algorithm is used to compute the time for the next measurement so that a given level of tracking performance is maintained. Using this on-line measure of tracking performance automatically takes into account target range, target maneuvers, missed detections, and strength of the returns. A testbed simulation program that includes the effects of target amplitude fluctuations, beamshape, missed detections, finite resolution, target maneuvers, and track loss is used to evaluate the performance of the proposed algorithm. The `best' tracking algorithm as defined by the benchmark problem is the one that requires a minimum number or radar dwells while satisfying a constraint of 4% on the maximum number of lost tracks. The proposed technique lost less than 2% of the tracks and provided average sample periods of 3.6 s for the commercial aircraft trajectory and 1.9 s for targets maneuvering with as much as 7 g's.

One of the more difficult problems in target tracking involves the use of a phased array radar to track an aircraft performing high speed maneuvers. Most tracking algorithms use single motion model track filters whose performance can degrade significantly when the target maneuvers. Multiple model algorithms can be used to improve the tracking accuracy and avoid the decision-directed techniques of single model track filters for maneuvering response. The interacting multiple model (IMM) algorithm uses multiple models that interact through state mixing to track a target maneuvering through an arbitrary trajectory. When tracking highly maneuvering targets with a phased array radar (i.e., agile beam), the issue of radar beam pointing is critical because poor pointing can lead to missed detections and eventually declaring lost tracks. The IMM algorithm provides a better method for beam pointing when compared to single model filters. This paper compares single and multiple model track filters with track loss as a measure of effectiveness. The effects of target maneuvers, data rate, track filter configuration, and radar beam pointing on the percentage of tracks lost are discussed. A phased array radar simulation that includes a fluctuating target, probability of detection, radar beamshape effects, and monopulse processing was used to assess the track loss performance of each algorithm.

This paper presents a simulation model for an air-to-air missile to measure the power losses due to specular and diffuse scattering on various terrains. This includes a range of surfaces from a sea surface of different root-mean-square surface roughness slopes to desert sand. This paper also presents the correlation between theoretical and empirical data for specular scattering on dry land and moist sand.

Coherent Doppler processing techniques add a fourth dimension to the data product of range instrumentation radars. Image generation from radar range and Doppler data is discussed and range/velocity tracking techniques using video tracker algorithms are explored. Test results and imagery from a Coherent-on-Receive Monopulse system are examined.

The 'Within-the-Beam Miss Distance Analysis' presentation is based on the study and development of an electronic subsystem, referred to as the Miss Distance Processor (MDP) and Post Mission Processor (PMP) software package. The MDP subsystem is used to sense the presence of multiple targets within the beam, i.e. field of view, of a single target tracking radar. Off-boresight angle and range errors gathered from each target within the defined window are used by the PMP to reconstruct and predict trajectories, which ultimately lead to the computation of the minimum distance between each target, as well as spatial position of each target. The complete subsystem will have a significant effect on the capabilities of single object tracking radars used on test ranges around the world, giving them the capability to provide spatial information on secondary targets within its beam without the use of multiple tracking stations. The subsystem's primary use is to assist in the accurate measurements of surface- to-air, air-to-air, and sea-to-air ballistic launches against airborne objects. These measurements can be obtained within minutes of the mission, as opposed to conventional techniques used with optical systems. Future capabilities of the system include real time reconstruction and miss distance calculations, construction of submunition trajectories after ejection from the primary projectile, and the use of multiple target video trackers for angle sensing in lieu of radar.

By means of return wave algorithm, acquisition performance equations for multiple glints are derived. It is found that phase conjugation adaptive optical system can lock to the glint of the highest reflectivity. Computer simulation of the acquisition process of multiple glints in the view field has been implemented. The acquisition behavior in the temporal and spatial domain is achieved.

The high altitude balloon experiment (HABE) is developing technologies important to resolving critical acquisition, tracking and pointing issues. An introduction to the technology papers in this session is given along with an update on the progress of the experiment. Contributions in the areas of controls, sensors, optics, systems analysis, and mission planning are presented along with progress in system and subsystem testing.

The high altitude balloon experiment (HABE) optical system is designed to address target tracking and pointing issues. The HABE optical system is based on a 60 cm telescope and involves an IR (4.4 micrometers ) tracking camera with a 2.3 mrad field of view and a visible (0.532 micrometers ) tracking camera with a 0.257 mrad field of view. An inertial reference alignment laser controls a fast- steering mirror to point a marker laser at the target. The basic optical layout and the optical design issues are discussed. Diffraction-limited performance was achieved for the design of the imaging cameras.

The high altitude ballon experiment (HABE) optical suite has been defined and specified. This paper describes the optical sensors on the payload and the test that has been accomplished. This paper covers the visible and IR sensors and alignment sensors. It also reports the radiometric calibration, the model fitting of the sensor responsivity data, and uncertainty analysis for the end to end radiometric calibration.

The High-altitude Balloon Experiment (HABE) is being developed to collect data, validate technologies, and resolve critical acquisition, tracking and pointing (ATP), and fire control issues for future tracking and pointing experiments. The cornerstone of the experiment is the control design architecture mating hardware and software to achieve HABE ATP objectives. This paper discusses the hardware and software components that define the control subsystems followed by their integration into the specific control modes. A discussion of the ATP performance goals, major design issues and implemented solutions in the HABE control mode architecture are also presented. The paper concludes with simulation results taking the HABE platform from open loop pointing to acquire the target within the initial track camera to the point ahead calculations of the marker system to place a beam on the target. The simulation details the expected performance of the HABE tracking system through its mission profile.

This paper presents the design of high altitude balloon experiment (HABE) attitude determination and control architecture. Inertial pointing forms the foundation of the HABE tracking system where the precise attitude control system is required to ensure target acquisition. The basic concept involves constructing an inertial line of sight (LOS) vector with target and HABE trajectory measurements and controlling the gimbal attitude to create a parallel LOS. Algorithms are developed incorporating the gimbal dynamics and inertial sensor information to optimize the pointing control system's flexibility and accuracy. Coordinate transformations between inertial, sensor, and vehicle frames intensify the complexities in realizing the pointing control system design. Kalman filtering is incorporated, providing trajectory prediction and compensating measurements for extraneous noise and latencies. The complete architecture of the pointing control system is analyzed with both frequency domain models and time domain simulation to characterize performance predictions.

The primary pointing and tracking telescope on the HABE optical payload is a 60 cm, clear-aperture, Cassegrain telescope. This paper presents the requirements for the telescope, a description of the optical, structural, and thermal design, the manufacturing process, and the performance test results. The telescope is required to have very high optical quality, a cryogenic system for actively cooling the primary mirror, a remote-controlled mechanism to refocus the secondary mirror, a baffle system to effectively reject off-axis radiation, and a wide-band optical coating. In addition, the telescope needs to be lightweight and stable over a wide temperature range during operation. The telescope has been successfully designed, fabricated, and delivered.

The High Altitude Balloon Experiment optical payload uses four on-board fast steering mirrors to accomplish its acquisition, tracking, and pointing function against boosting missile targets. The payload function requiring these high-performance mirrors include inertial line-of-sight stabilization in conjunction with an inertial sensor assembly, internal optical beam path auto- alignment, illuminator and marker laser boresight, alignment, and point-ahead. These mirrors are required to have high bandwidth, low noise characteristics, be very lightweight, and operate in various analog and digital closed loop modes. A goad of design commonality among the mirrors is set to facilitate interchangeability during field operations. All five fast steering mirrors, including one spare, have been successfully designed, fabricated, tested, and delivered to Phillips Laboratory.

IPSRU (Inertial Pseudo Star Reference Unit) is an inertial line- of-sight pointing reference that has an ultralow jitter of less than 40 nrad rms from 0.4 to 312 Hz. IPSRU emits a collimated optical reference beam from a stabilized, sterrable two degree of freedom platform, and this reference beam is used to stabilize a payload's line-of-sight against vehicle vibrations. IPSRU's three critical performance parameters are isolation of host vehicle vibration from the reference beam, the intrinsic reference beam jitter due to self-generated platform noise, and open-loop inertial pointing accuracy of the reference beam. IPSRU, developed at Draper, is in final performance testing before delivery to Phillips Laboratory for the high-altitude balloon experiment (HABE). The challenges of measuring and validating IPSRU's performance, including a total beam jitter of less than 40 nrad, a base isolation transfer function that ranges from -93 dB at 1 Hz to -55 dB at 100 Hz, and an open-loop pointing accuracy of 21 ppm, are met with a specially designed pointing scoring facility that includes state of the art instrumentation and a quiet reference pier. Beam jitter and base motion isolation are measured with a Draper-enhanced quad-type detector with an rms noise of 6 nrad from 0.4 to 312 Hz and a field of view of 8 (mu) rad. Open-loop pointing maneuvers are measured with an enhanced-resolution laser interferometer (LI) (6.8 nrad per bit) with a 4 deg field of view. Additionally, IPSRU is scored in a thermal vacuum chamber that simulates the HABE environment or a satellite environment.

This paper presents and discusses laboratory performance measurements of a Roto-Lok drive system for the HABE azimuth and elevation gimbals. The HABE system is a 7,000 lb acquisition tracking and pointing (ATP) balloon-launched vehicle. The primary azimuth and elevation drive systems are zero-backlash torque multipliers referred to by the trade name Roto-Lok rotary drive and designed by Sagebrush Technology, Inc. The Roto-Lok used in the azimuth gimbal has a limited 320 deg of angular travel; therefore, it is supplemented with a secondary drive element to provide unlimited travel. This secondary drive is used to counteract the gross angles resulting from the freely rotating nature of the untethered balloon system. The Roto-Lok drive is used for the fine tracking and pointing of the gimbals. Both the azimuth and elevation Rota-Lok drives are tandem drives with an end-to-end ratio of 72:1. Performance specifications developed from the mission requirements are compared against the actual system performance measurements. The entire gimbaled azimuth and elevation systems are required to point in inertial space to less than 250 (mu) rad RMS over the band DC to 100 Hz for each axis. Performance measurements better than the specification were measured. The primary gimbal base-motion disturbances, however, are due to the motor cogging torque or torque ripple. A brief discussion of the measurement methods and the control system used to drive the gimbals is presented. Several system anomalies, such as the structural compliance between the drive element and the inertial rate sensors and the coarse gear backlash, are discussed in terms of their impact on the gimbal control system.

A significant objective the High-Altitude Balloon Experiment (HABE) is to conduct and develop experiments that focus on laser pointing and tracking. The target scoreboard systems is a scoring system that determines pointing accuracy, jitter, and beam drift. It also is designed to potentially provide direct feedback to HABE's payload tracker. In this paper, we describe the system's design, a model of critical parameters, and provide the operation data we have collected to date. The target scoreboard is composed of a marker laser, a portion of the payload's optics, an 11 x 11 photo diode array mounted in the nosecone of a WB-57 aircraft, centroid processing electronics that operate at 2,000 frames per second, telemetry, a global positioning system (GPS) receiver, motion sensors, a digital recorder, and an operator control console. The scoreboard system and design are unclassified.

The High-Altitude Balloon Experiment requires the injection of an inertial reference beam into the optical system by means of an extended corner cube (ECC). The optical system including the ECC can move up to +/- 1 mrad with respect to the inertial beam. The ECC must return the beam antiparallel to the inertial beam to within +/- 40 nrad. The analysis presented here indicates that the three mirrors of the ECC must be perpendicular to each other to within 2 arcsec to meet the specification.

The High-Altitude Balloon Experiment telescope was designed to operate at an ambient temperature of -55 degree(s)C and an altitude of 26 km, using a precooled primary mirror. Although at this altitude the air density is only 1.4 percent of the value at sea level, the temperature gradients within the telescope are high enough to deform the optical wavefront. This problem is considerably lessened by precooling the primary mirror to -35 degree(s)C. This paper describes the application of several codes to determine the range of wavefront deformation during a mission.

Requirements for image resolution can be used to set upper limits on the allowable line-of-site (LOS) motion of an acquisition, tracking, and pointing (ATP) system. Image resolution is important for image-based tracking algorithms and for typical ancillary requirements for target phenomenology data gathering. During the system design phase of an ATP platform, base-motion- disturbance details such as total rms power and spectral distribution of this power may not be known for primary disturbance sources such a gimbals, cooling systems, and steering mirrors. In this case, setting upper limits to allowable LOS jitter is an important criteria in the trade study analyses for these components. The effect of jitter is frequency dependent and can be partitioned into regimes based on the image sample rate of the system. The application of image-resolution requirements for the High Altitude Balloon Experiment are used to set allowable LOS motion for random, sinusoidal, and linear disturbances. Three frequency regimes are identified with different allowable-motion amplitudes. This top-level systems methodology can be applied to many imaging applications such as estimating the blur induced by wind loading of ground based telescopes.

Successful tracking of a target from the High Altitude Balloon Experiment (HABE) requires selecting an initial position for the HABE platform so the greatest quantity and quality of imagery can be gathered for the longest duration. During the engagement period, a target typically covers a large distance while HABE drifts with the wind. The parameters of the engagement used to measure potential success depend directly on the capabilities of the tracking hardware and software. An engagement analysis technique has been developed which allows optimization of important engagement parameters. This technique can support a wide variety of tracking platforms which require pointing to specific locations or targets. Two methodologies are described in this paper which predict those positions for HABE where the most desirable engagements will be achieved against a nominal target. The desired objective is a long period of continuous maintainable track.

Significant technical considerations prompted the start of an evolution, twenty years ago, from the use of the floated, rate- integrating gyro to the dynamically tuned gyro (DTG) for the precision, three-axis attitude control of spacecraft. DTG's have since then been widely applied in U.S. spacecraft. The anticipated advent of smaller, lower cost spacecraft prompted the development of a new generation, lower cost, low weight (3.6 lb/1.8 kg), moderate performance inertial reference unit (IRU) for such applications by making use of the miniature CONEX DTG. The resulting two-axis rate assembly has already been qualified and delivered to two programs. In parallel, development of a miniature, three-axis ring laser gyro has proceeded to the point where an inertial measurement unit has entered the early phase of production for a tactical missile program. Simple modification of the hardware will provide a production-based IRU or IMU design at a cost advantage for spacecraft applications.

A pointing control system has been designed based on a CCD array and a fine steering mirror. The control system accurately points a transmit laser ahead of an incoming beacon laser without requiring accurate prestabilization of the beacon line of sight. The digital design of the control loop and a laboratory demonstration of CCD-based laser pointing are presented.

A high frame rate CCD camera with interface to a digital signal processor has been developed for a pointing control loop for laser communications. The CCD array is operated in a windowed- read mode with the digital signal processor controlling the vertical and horizontal shifts in the image and storage planes. Development of the camera and interface hardware for a demonstration of precision beam pointing is presented.

An autonomous star tracker (AST) is basically a `star field in, attitude out' device capable of determining its attitude without requiring any a priori attitude knowledge. In addition to this attitude acquisition capability, an AST can perform attitude updates autonomously and is able to provide its attitude `continuously' while tracking a star field. The Lockheed Palo Alto Research Laboratory is developing a reliable, low-cost, miniature AST that has a one arcsec overall accuracy, weighs less than 1.5 kg, consumes less than 7 watts of power, and is sufficiently sensitive to be used at all sky locations. The device performs attitude acquisition in a fraction of a second and outputs its attitude at a 10 Hz rate when operating in its tracking mode. Besides providing the functionality needed for future advanced attitude control and navigation systems, an AST also improves spacecraft reliability, mass, power, cost, and operating expenses. The AST comprises a-thermalized, refractive optics, a frame-transfer CCD with a sensitive area of 1024 by 1024 pixels, camera electronics implemented with application- specific integrated circuits, a compact single board computer with a radiation hard 32 bit RISC processor, and an all-sky guide star database. Star identification is performed by a memory- efficient and highly robust algorithm that finds the largest group of observed stars matching a group of guide stars. An important milestone has recently been achieved with the validation of the attitude acquisition capability through correct and rapid identification of all 704 true-sky star fields obtained at the Lick Observatory, using an uncalibrated prototype AST with a 512 by 1024 pixel frame-transfer CCD and a 50 mm f/1.2 lens that provided an effective 6.5 by 13.2 degree field of view. The overlapping fields cover 47% of the sky, including both rich and sparse areas. The paper contains a description of the AST, a summary of the functions enabled or improved by the device, an overview of the identification algorithm, results obtained with a realistic simulation program, a description of the true-sky star field identification method and a presentation of the results obtained. The AST tolerates the presence of bright objects as was demonstrated by a field that included Jupiter.

In a typical planetary spacecraft mission it is of interest to autonomously detect features on a planetary surface and to be able to track them. Various algorithms exist to track a part of an image in a continuous sequence, for instance correlation algorithms. However, these algorithms tend to be computational intensive, and in addition, they lack autonomy. Furthermore they are not robust toward occlusion, changed lighting conditions, and changes in vantage point. Therefore a more reliable strategy to track planetary terrains is needed. This is one of the major research topics of the Autonomous Feature and Star Tracking (AFAST) project at Jet Propulsion Laboratory, Pasadena. This paper presents an alternative method of feature tracking, namely by extraction of such features in the images that are reoccurring in consecutive images and that are invariant to various parameters. With these candidate features, it is possible to navigate the spacecraft based on identified terrain data. A novel methodology is proposed to track planetary surfaces by recognizing feature constellations, a method similar to those employed in recognizing star constellations in a star tracker.

This paper describes the implementation and first on flight results of a GPS and a star imaging system (SIS) using a common transputer processing unit. The described transputer navigation unit has been integrated on PoSAT-1, the first Portuguese satellite, which was launched into LEO on September 26, 1993. The authors present the possible employment of these systems, the SIS, the GPS, and the parallel nature of transputer processing, as a powerful and efficient navigation unit for small and low earth orbiting satellites.

A six-feature all-sky star field identification algorithm has been developed for autonomous attitude determination. The minimum identifiable star pattern element consists of an oriented star triplet defined by three stars, their celestial coordinates and visual magnitudes. This algorithm has been integrated with a CCD- based imaging camera and tested in an observatory environment. The autonomous intelligent camera identifies in real time any star field without a priori knowledge. A set of observatory tests on star fields with this intelligent camera are described.