12th Annual American Institute of Aeronautics and Astronautics Utah State University Conference on Small Satellites Logan, UT August 31 - September 3, 1998 Lawrence Livermore National Laboratory UCRL-JC-131321 Autonomous, Agile, Micro-Satellites and Supporting Technologies for Use in Low-Earth Orbit Missions Arno G. Ledebuhr, Joseph F. Kordas Lawrence C. Ng, Mark S. Jones Oliver D. Edwards, John C. Whitehead Richard J. Gaughan, Michael D. Dittman July 20, 1998 This is a preprint of a paper intended for publication in a journal or proceedings. Since changes may be made before publication, this preprint is made available with the understanding that it will not be cited or reproduced without the permission of the author. PREPRINT This paper was prepared for submittal to the
Transcript
12th Annual American Institute of Aeronautics and AstronauticsUtah State University Conference on Small Satellites
Logan, UTAugust 31 - September 3, 1998
Lawre
nce�
Liver
more
�
Nat
ional
�
Labora
tory
UCRL-JC-131321
Autonomous, Agile, Micro-Satellites and Supporting Technologies for Use in
Low-Earth Orbit Missions
Arno G. Ledebuhr, Joseph F. KordasLawrence C. Ng, Mark S. Jones
Oliver D. Edwards, John C. WhiteheadRichard J. Gaughan, Michael D. Dittman
July 20, 1998
This is a preprint of a paper intended for publication in a journal or proceedings. Since changes may be made before publication, this preprint is made available with the understanding that it will not be cited or reproduced without the permission of the author.
PREPRINT
This paper was prepared for submittal to the
DISCLAIMER
This document was prepared as an account of work sponsored by an agency ofthe United States Government. Neither the United States Government nor theUniversity of California nor any of their employees, makes any warranty, expressor implied, or assumes any legal liability or responsibility for the accuracy,completeness, or usefulness of any information, apparatus, product, or processdisclosed, or represents that its use would not infringe privately owned rights.Reference herein to any specific commercial product, process, or service by tradename, trademark, manufacturer, or otherwise, does not necessarily constitute orimply its endorsement, recommendation, or favoring by the United StatesGovernment or the University of California. The views and opinions of authorsexpressed herein do not necessarily state or reflect those of the United StatesGovernment or the University of California, and shall not be used for advertisingor product endorsement purposes.
SSC98-V-1
1Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites
Autonomous, Agile, Micro-Satellites and Supporting Technologies for Use inLow-Earth Orbit Missions
A. G. Ledebuhr, J. F. Kordas, L. C. Ng, M. S. Jones,O. Edwards, J. C. Whitehead, R. J. Gaughan, and M. D. Dittman
Lawrence Livermore National LaboratoryP.O. Box 808, L-043Livermore, CA 94550
Abstract. This paper summarizes the work at Lawrence Livermore National Laboratory in thedevelopment, integration and testing of the critical enabling technologies needed for the realizationof agile micro-satellites (or MicroSats). Our objective is to develop autonomous, agile MicroSatsweighing between 20 to 40 kilograms, with at least 300 m/s of Dv, that are capable of performingprecision maneuvers in space, including satellite rendezvous, inspection, proximity operations,docking, and servicing missions. The MicroSat carries on-board a host of light-weight sensors andactuators, inertial navigation instruments, and advanced avionics. The avionics architecture isbased on the CompactPCI bus and PowerPC processor family. This modular design leveragescommercial-off-the-shelf technologies, allowing early integration and testing. The CompactPCIbus is a high-performance, processor independent I/O bus that minimizes the effects of futureprocessor upgrades. PowerPCs are powerful RISC processors with significant inherent radiationtolerance. The MicroSat software development environment uses the space flight proven Vx-Works, a commonly used, well tested, real-time operating system that provides a rapiddevelopment environment for integration of new software modules. The MicroSat is a 3-axisstabilized vehicle which uses cold gas N2 for ACS and a novel pressure-fed, non-toxic, mono-propellant hydrogen peroxide propulsion system for maneuvering.
Introduction
Competition in the consumer electronicsmarketplace continues to drive componentminiaturization and consolidation reducingcost, mass and power consumption whileimproving system performance. Thistechnology push should provide thecapability to build significantly smaller andsmarter micro-satellites (MicroSats) with wetmasses between 10 to 100 kilograms. Thispaper describes on-going efforts at LawrenceLivermore National Laboratory to develop thecritical enabling technologies needed for therealization of autonomous agile MicroSatsthat are highly autonomous in function, lessthan 50 kg in mass, and possess a robustorbital maneuvering capability of >300 m/s ofDv. LLNL has developed a preliminarydesign of a near-term MicroSat for a varietyof Rescue Mission applications and hasconstructed two prototype vehicles forground test experiments.1 This paper willdiscuss representative mission applications,
enabling vehicle technologies, and integratedvehicle testing approaches.
Potential Missions
Potential missions for MicroSats center onspace ÒlogisticsÓ missions such as rescue andservicing, that will require vehicles with theability to perform a variety of missionfunctions autonomously or semi-autonomously. These include, rendezvous,inspection, proximity-operations (formationflying), docking and robotic servicingfunctions (refueling, repowering orrepairing). Each of these mission functionsrequire key technical capabilities. Forexample, rendezvous with a space asset byperforming orbit matching requires precisionDv maneuvers. Inspection of a space asset byflying to different view points of aninspection geometry requires precisionMicroSat positioning, precision pointing,tracking and imaging. A satellite rescue might
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites2
involve docking, repairing or refueling thesatellite, followed by a departure, and post-rescue inspection. The rescue missionrequires precision guidance, navigation, andcontrol; precision ranging; high resolutionimaging; and some type of micro-roboticmanipulator. For example, a variety ofrobotic arms could be used to enable theMicroSat to perform a physical dock with atarget satellite. Once docked, a precision 6degrees-of-freedom manipulator (actuator)could be used to align and plug an externalconnector into a targeted satelliteÕs umbilicalconnector for data collection, diagnosticmeasurements or repowering. Other spacelogistic operations such as the collection andde-orbiting of hazardous space debris (junk)or the interdiction of asteroids or comets in aplanetary defense system, requires precision
vehicle guidance, navigation and control anda precision endgame homing strategy.Formation flying, flying in concert with aspace object or another MicroSat, requiresstation keeping and positioning, andprecision state vector estimation.
Recently LLNL has studied the technologiesand requirements for a satellite RescueMission.1 Table 1 contains a preliminaryanalysis of a timeline for this representativerescue mission. It assumes a Pegasus launchfor the MicroSat with orbit injection errors asoutlined under Day 1Õs General Comments.The estimated Dv for each MicroSatmaneuver is listed. It is assumed that theservicing required is a battery recharge andprocessor restart.
Table 1 Preliminary Rescue Mission timeline for Rendezvous, Inspection, Docking and Departure.Day Event DV
(m/s)General comments
1¥ Pegasus launches and deploys MicroSat¥ MicroSat performs practice rescuemission with payload interface adapter¥ MicroSat orbit determination
15Pegasus with HAPS (hydrazine auxiliarypropulsion system) can place MicroSat withthe following (3s) injection errors:± 5 km @ apogee; ± 5 km @ perigee;0.5° in ascending node; 0.05° in inclination
Correct perigee error to within 1 km of targetsatellite and avoid collision
6¥ MicroSat proceeds to a point 100mbehind target satellite 50
Image target and conduct station keeping;relay imagery to ground
1 4 0Total Dv required for orbit match andrendezvous including a completepractice docking
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites3
7¥ MicroSat initiates a 100m circularinspection of target¥ Preliminary estimate of target axis ofrotation and rotation rate
25Relay imagery and estimates to ground fordetailed analysis
8¥ MicroSat proceeds via a circular transferorbit to 10m range from target satellite¥ Refine target spin axis estimation¥ Initiate a 10m circular maneuver abouttarget spin axis and track the landing site
8
2
Image target and relay pertinent data to groundstation
9¥ MicroSat proceeds to 3-5m range fromtarget¥ Execute a circular maneuver about target
1
2
MicroSat tracks landing site, analysesimagery, and orients itself for landing
10¥ MicroSat closes in and docks
2MicroSat grabs on target satellite flangestructure
11¥ Align MicroSat to target connector 1
Send imagery to ground and wait for up-linkcommand to proceed
12¥ Insert MicroSat adapter into targetsatellite connector 2
MicroSat uses a six degree of freedom fixtureto insert pin connector and vision basedimaging software to guide adapter
13¥ Mate and revive lost satellite Trickle charge satellite battery, re-activate
satellite computer, and report reactivationdiagnostics to ground
14 ¥ Departure from satellite 2 Move away to 50 - 100m behind satellite
15¥ Observe target satellite deploymentmaneuver 5
Mission accomplished
5 0Total Dv for Inspection, Docking,and Departure operations
Spacecraft Technologies The LLNL MicroSat is an adaptation of thelightweight spacecraft technology developedfor the Clementine I & II missions. It featuresa lightweight, Seeker Head, state-of-the-artavionics, H2O2 divert propulsion systemwith N2 gas for attitude control, VxWorksbased real-time controller with a collection oftechnology based software for precisionguidance and control and image analysis.Key modifications from the Clementinemissions include the addition of stereo
cameras for passive ranging and 3D imaging,a micro-impulse radar for precision docking,miniature grappling arms for docking and a6DOF robotic arm for satellite servicing. Alsoadded is a mating camera to provide visionfeedback to guide the robotic arm, a GPSreceiver for precision orbit determination, andspot lights for illumination. Figure 1 shows aconceptual design of this system based oncurrent prototype MicroSat vehicles. Thefollowing paragraphs briefly describe theenabling technology for the development ofthe agile MicroSat.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites4
Figure 1 A conceptual MicroSat designed for a rescue mission.
MicroSat Spacecraft
Figure 2 shows an exploded view of thecurrent design of the MicroSat. The SeekerHead Assembly consists of a suite of imagingsensors and provides the GN&C input to thevehicle. These sensors include a highresolution acquisition camera for long rangedetection and standoff inspection, a mediumfield-of-view inspection camera for closer-ininspection, one of two small wide-field-of-view (WFOV) color docking cameras forstereo imaging and passive ranging, a WFOVStar Tracker for inertial navigation, alightweight IMU, a compact GPS unit and aminiature micro-impulse Docking Radar. TheIMU operates in conjunction with the StarTracker to provide attitude quaternions forprecision vehicle guidance and control. TheMicroSatÕs modular state-of-the-art avionicssuite is based on a compactPCI bus and usesa PowerPC 603e processor and several I/Omodules that support the rest of the vehiclesub-systems. Other sub-systems include a
pair of rechargeable NiCad battery packs,solar array panels for battery charging,coupled ACS jets for in-place rotation, H2O2divert thrusters and N2 jets for precisiontranslation control. Also included is a mini-SGLS transponder that is AFSCNcompatible, a set of grappling arms fordocking and a 6DOF robotic actuator forsatellite servicing (arms and actuator notshown in this view).
In order to achieve centimeter level ofposition control of the MicroSat, duringdocking maneuvers, we make use of thesmall minimum impulse bit that is availablefrom the cold gas propellant. Our currentdesign goals are: 4 N-s for H2O2 diverts,0.01 N-s for cold gas diverts. The currentMicroSat ground test vehicle is <25 kg andour final design for a flight qualifiedMicroSat with a full complement ofsubsystems is expected to weigh less than 40kilograms.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites5
HiResSeeker
Star Tracker
MicroRadar
Batteries
Divertthrusters
H2O2tanks
ACStanks
ACS thrusters
Avionics &Processors
Antenna &Transmitter
Solarpanels
IMU
Axialthrusters
StereoCameras
Figure 2 An exploded view of the MicroSat design for the rescue mission (Some components notshown).
The original worst case minimum thrusterburn period was assumed to be one second(assumes a cold start) for the H2O2 divertsystem and five milliseconds for the nitrogensystem. Table 2 below computes apreliminary set of requirements for theminimum impulse bit (MIB), minimumresolution in velocity, minimum resolution inposition, and maximum vehicle acceleration.based on these timing assumptions. Thecoupled cold gas ACS can provide precisiontranslation, by firing two matching 1N ACS
jets simultaneously, a mission requirementused primarily during precision orbitmatching, docking, un-docking and departuremaneuvers. Preliminary testing of the H2O2divert system indicates that once the thrustershave been warmed-up the response times aretypically 10X shorter. For example ourprototype 20 N vacuum thruster provided 14N thrust at sea-level with less than a 0.1second response time when warm. Thisyielded an MIB of approximately 1 N-s forthe first generation H2O2 divert system2.
Table 2 Precision vehicle control requirements.Thrust pulsewidth MIB Dv Dp Accel comments(N) (sec) (N-sec) (m/s) (m) (mg)
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites6
LLNL's current approach to propulsion formicro-spacecraft is to work with nontoxicpropellant. We have demonstrated thathydrogen peroxide, when used on a scaleappropriate for micro-spacecraft, permitscost-effective bench-top testing of custom-developed hardware. Unlike conventionalspace propulsion systems, a hydrogenperoxide (H2O2) micro-propulsion systemcan readily be tested at the system level, evenafter spacecraft integration.
While H2O2 has a lower specific impulse thanhydrazine, 2.5 times the Isp of cold nitrogenis readily achieved. In addition, H2O2 hasover 3 times the density of 5,000 psinitrogen. The combined effect of Isp anddensity makes monopropellant H2O2 manytimes more capable than cold gas used alone,which is currently the only choice forcustom-designed micro-propulsion systems.
Recently during the second half of FY97,LLNL researchers designed, fabricated,tested, debugged, and demonstrated aminiature divert propulsion system (seeFigure 3 below) in support of the ClementineII asteroid intercept mission plan1,2. Theresults of ongoing R&D (e.g. long term tankstorage tests) indicate that custom-designedH2O2 propulsion is well suited to LEOmissions, for spacecraft weighing 10 kg to50 kg total.
There are several options for using H2O2,including self-pressurizing tanks which feedliquid to catalytic thrusters, while alsodelivering a steam-oxygen mixture to warmgas attitude control jets. A relatively smallamount of kerosene can be carried to doublethe specific impulse (nontoxic bipropellants).Figure 4 shows the block diagram of a firstgeneration pressure-fed H2O2 propulsionsubsystem.
The total maneuvering requirement is 250m/s, including a maneuvering reserve. Thepropulsion system is being designed to 300m/s, for a 20% propulsion margin. Thebreakdown is 275 m/s with the liquid systemand 25 m/s with nitrogen (ACS impulserequirements have been converted to anequivalent delta-v based on vehicle mass afterliquid burns are complete). The microspacecraft total mass allowance is 40 kg, ofwhich 9 kg of 85% H2O2 would be requiredto achieve 275 m/s at Isp=130 s. Delivering25 m/s after liquid burnout requires 1.89 kgof nitrogen at Isp=50 s. This includes anadditional 150 grams of nitrogen topressurize the liquid tanks. Initial hardwaremass estimates are 5 kg for the "tank-as-structure" liquid system (alreadydemonstrated with prototype hardware), and4 kg for the gas components. Thus, total wetpropulsion mass is just over 20 kg.
A preliminary estimate of the vehicle massbudget is shown in Table 3 below. A totalvehicle wet weight of 39.9 kg with a designmargin of 7.2 kg is shown at the currentconceptual design stage. Our goal is toachieve a total wet weight of less than 40kg.The table shows that we should be able tomeet this goal and that there is also adequatemargin to mass balance the vehicle.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites7
4 x Pressurant Tank9 ksi MEOP ratedlimit to 3.0 ksi MEOPproof @4.5 ksi
The sensor suite or Seeker Head consists of aStar Tracker for attitude determination usingstellar navigation, a high resolutionAcquisition camera for long range (20 km to0.1 km) target detection and stand-offinspection, a medium resolution Inspectioncamera for closer-in inspection (in the 100 mto 10 m range), one of a pair of wide-field-of-view color Docking cameras forproximity-operations (in the 10 m to 0.1 mrange), a microRadar for active close-in (< 10meter distances) providing precision rangeand range-rate measurements, a compactGPS receiver for coarse positionmeasurements, and an IMU for inertialnavigation. These sensors are lightweight andcompact. The imaging cameras (Acquisition,Inspection and Star Tracker) were originallydeveloped for ballistic missile defenseapplications. A schematic of the MicroSatsensor suite is shown in Figure 5. Versionsof the Acquisition (HiRes) camera,Inspection (UV/Vis) camera and the StarTracker cameras, have previously flown inthe successful Clementine I Lunar mappingmission3,4,5.
WFOV Star Tracker
A key sensor in the MicroSat is a wide-field-of-view (WFOV) Star Tracker whichprovides inertial orientation of the vehicle.The Star Tracker camera in conjunction withStellar Compass software can provide aquaternion pointing accuracy of 450 mrad.The Star Tracker field of view is largeenough to contain several bright stars in anyorientation. Single images are processed toidentify unique stellar patterns and providethe determination of the inertial orientation ofthe MicroSat in real-time. The Star Trackerlens was specified and built to have a 42° x28° field of view. This sizing results in apixel IFOV of 1.3 mrad, which is smallenough to provide the requisite quaternionaccuracy. The collection aperture of the lensis maximized for the greatest possible lightgathering capability. At F/1.25, mv = 4.5 G0stars provide an integrated star signal that is15 times the electronic noise from the focalplane. This level of signal gatheringcapability, matched with the wide field ofview, ensures a 99.9% probability that 5stars above minimum threshold will beavailable for the algorithm set for all possiblequaternion pointing vectors. This allows theStar Tracker to handle the Òlost in spaceÓcondition with a single star image frame andno other a priori knowledge of attitude. Apreliminary analysis of the imaging sensorrequirements to support the various phases ofa typical rescue mission is given next.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites9
Hi-Res Camera
Stereo Camera
GPS
Radar
IMU
Star Tracker
UV/Vis Camera
Accelerometers
Battery Pack
Figure 5 Multi-function MicroSat sensor suite.
Sensor Requirements
The top-level optical system requirements,shown in Table 4 below, support operationalphases of our representative rescue mission.In the Rendezvous phase from 20 km to 100meters, the High Resolution Acquisitionsensor will be required to identify the targetand provide centroids to the guidance system.This sensor will provide the first detailedimages to the ground at a few kilometers out.At several hundred meters out a transitionwill be made to the Medium ResolutionInspection camera which will be used todetermine the target spacecraftÕs rotationalparameters. The 100 m to 10 m closingportion of the Inspection phase will againrequire input from the optical sensors, andthe 10 m inspection will produce moredetailed images for ground evaluation. At 10meters the vehicle will rotate 90 degrees(pitch or yaw) and will then switch to its pair
of color Docking cameras. The sensors willagain be required to provide information tothe guidance and navigation control systemduring the 10 m to 10 cm closing portion ofthe Docking phase . For the dockingmaneuver the optical sensors will allowfeatures to be identified, and in the terminalphase they will determine the precise locationon the target where the hard dock will bemade (launch vehicle Marmon interfaceflange). After the docking is complete, thesensor system will provide images for the Servicing phase of the mission includingproviding images that support the mating ofthe MicroSat to the target satelliteÕs umbilicalconnector and confirmation when theconnection is completed. A small Matingcamera (Endoscopic), attached to the MatingActuator (manipulator mechanism) willprovide imagery of the precise alignment andorientation of the individual pins in theconnector. After the servicing is complete and
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites10
Table 4 Top Level Optical System Requirements.Mission Phase
LabelRange Sensor Function Operational Phases
A 20 km - 100 m acquisition/tracking RendezvousB 100 m - 10 m imaging/tracking/guidance InspectionC 10 m - 10 cm imaging/guidance DockingD 10 cm - 1 mm imaging/guidance/alignment ServicingE 10 cm - 100 m imaging/tracking/guidance Departure
the MicroSat switches to the Departure phase ,it will un-dock and depart the satellite. Thesensor system now will produce newinspection images to confirm the successfulservicing operation.
Each mission phase levies a particular set ofrequirements on the optical sensor system.These requirements are best met with asensor system optimized for each missionphase. An alternate approach would be toutilize a zoom lens system to provideoverlapping capability. Future designs willexplore this option. Our current opticalsystem mission support is summarized inTable 5. The Star Tracker provides MicroSatattitude information in all mission phases,and is used to support both the identificationof the satellite spin axis in inertial space, andMicroSat guidance and navigation. TheAcquisition camera provides the long rangeacquisition of the target spacecraft and
provides initial imagery of the target. Thiscamera supports the terminal portion of theRendezvous phase and can provide data tosupport an autonomous rendezvousoperation. The Inspection camera providesadditional target information to the guidanceand navigation computer, and also suppliesdetailed target images at 100 meter to 10meter ranges. The Docking camera systemproduces close-in imagery of the payloadinterface flange (docking region) of the targetsatellite. This system may be used in eitherstereoscopic or monoscopic mode. Stereoimagery provides passive range data for theGN&C system. The docking camera is alsoused to assist in mating the umbilicalconnector. The Mating camera is mounted onthe 6DOF manipulator and will support thefine alignment of the two connectors. TheDeparture phase will essentially be a timereversal of the Docking and Inspectionphases and can be met with these sensors.
Table 5 Optical System Mission Support.Camera Effective Range Mission Phases Supported
Star Tracker N/A A,B,C,EAcquisition Camera 20 km - 100 m AInspection Camera 1 km - 8 m A,B
Docking Camera System 10 m - 10 cm C,D,EMating Camera System 10 cm - 1 mm D
Avionics
The MicroSat avionics architecture is basedon the latest PowerPC processor andCompactPCI bus as shown in Figure 6. ThePowerPC family has significant inherentradiation tolerance. The CompactPCI bus is ahigh-performance, processor independent I/O
bus which will minimize the effect of futureupgrades in processors. The system supportsmodern, well-used and well-tested embeddedsoftware development environments. Thisdesign allows rapid code development,debugging, and testing, as well as hardwareintegration. The chosen architecture andprocessor will provide a high performance
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites11
solution for multiple MicroSat missions. Itsmodular design leverages COTS
It provides a clear path to performanceupgrades. In addition, this approach and typeof architecture are being embraced by majoraerospace system providers.
Processor Module
The MicroSat processor module contains ahigh-performance PowerPC 603e RISC CPU(100MHz-300MHz) that utilizes the 33 MHz,32 bit data path (132 Mbyte/sec max)CompactPCI bus, a high bandwidth I/O busthat is an industry-standard ruggedizedversion of the desktop PCI bus used invirtually all desktop computers, MPC106PCI bridge/memory controller with built inmemory single bit error correction and doublebit error detection and low power modes, 32Mbytes DRAM, 4 Mbytes flash EEPROM, 8Kbytes PROM, a real time clock with 4Kbytes non-volatile RAM, two high-speed
UART channels, interval timers, andwatchdog circuit. The flight processor will bea COTS module with modifications forthermal management and radiation tolerantparts as needed.
The MicroSat local data store will beconnected to the processor's IDE controllerport as indicated above. This will be acommercially available Flash Disk that hasbuilt-in error correction capability. The sizeof the disk selected will be based on theworst-case telemetry data requirements. Thedocking phase presents the most stressingconditions. In Table 6 we show data storerequirements for a representative rescuemission. Estimates are provided for twodifferent docking camera designs, using twodifferent CCD size options (different pixelnumber, size and formats, designated A andB).
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites12
Table 6 Data storage requirement estimates for a representative Rescue Mission.Component/subsystem Unit Count Frame size Frames Pixels Bytes Bytes Comp. Data Store Minutes Storage Req.
Horiz Vert /sec /sec /p i xe l /sec Ratio :1 Bytes/sec Active Bytes
The digital frame buffer module is designedto provide a high-performance imageacquisition and data handling interfacebetween the CompactPCIª bus and high-speed digital cameras. The features of themodule are as follows: frame formats to4096x4096, 8 to 16 bit pixels, pixel clockrates to 20Mhz, two independent videochannels (simultaneous acquisition),multiplexed operation for cameras sharingchannels, Automated Imaging Association(AIA) digital camera compatibility,Synchronous Addressable Serial Interface(SASI) camera controllers, 32 MbytesSDRAM image storage, Image compression(Lossless » 2:1, Lossy > 10:1), and aCompactPCI bus interface with Direct Masteror Slave operation and two independent DMAcontrollers. The frame buffer is a LLNLdesign that will contain rad-tolerantcontrollers and will incorporate thermalmanagement.
RF Transceiver
The RF transceiver under development forMicroSat applications is a SGLS-signalingAFSCN-compatible unit that is beingdeveloped by the Naval Research Laboratory(NRL). This unit was started as part of theClementine II program and was originallydesigned to provide a telemetry link betweenthe asteroid impact probe and the Mothershipbus during its fly-by of a near earth (earth-
crossing) asteroid. This new transceiverdesign is expected to have dimensions of14.5 cm x 14.5 cm x 4.6 cm and a mass thatis under 2 kg. It is designed to operate onSGLS Channel 10, with a command up-linkdata rate of 2 kbps (1799.76 Mhz) and atelemetry down-link data rate of 1 Mbps(2247.5 Mhz). For a 1 Mbps telemetry ratethe total estimated radiated output power is 3watts and the total power consumption is 25watts. More detailed information may beobtained by consulting the NRL Naval Centerfor Space Technology specifications SSD-S-CM013 and SSD-S-CM017.
GPS Receiver
Several commercially available GPS receiversare available for this application and a lightweight (<2 kg) space qualified unit isavailable from Boeing North AmericaÕsRockwell division. Initial testing will becarried out with off-the-shelf commercialunits, one of which will be selected for flightqualification during the later developmentphase of a flight experiment program.Ground test experiments will be carried outwith differential GPS to determine theaccuracy and utility of the position andattitude data derived from this subsystem.Closed loop positioning (docking)experiments will be performed to determinethe overlap in functionality between thevarious sensor systems (Star Tracker, Micro-Radar and Docking cameras).
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites13
Power Distribution System
In order to minimize the size and mass of theMicroSat solar array assembly and batterypack, a power management scheme isrequired which allows components andsubsystems to be powered down when notneeded, and which allows for periods ofrecharge between periods of activity thatcause the battery charge level to fall belownominal value.
The master power distribution modulecontains a low power consumption controllerfor the purpose of spacecraft powermanagement. The controller is capable ofmonitoring and controlling the power to all ofthe spacecraft loads. A communications linkto the system processor allows the systemprocessor to make mode changes to thepower management controller, send powerdown commands, monitor the power systemcondition and provide a way for the controllerto signal the system processor of powersystem alarms. During the power downperiod the battery modules will be rechargedby the solar cell array. The recharging ratewill be a function of the particular missionorbit flown, which will determine the averagesolar incidence angle and effective exposuretime.
Battery Module
Representative power system design for anear-term first generation agile MicroSat is asfollows. Each battery module would containa 23 cell, rechargeable, high capacity Nickel-Cadmium battery pack. At a nominal outputvoltage of 28 Volts, the module will produce50 Watt-Hours (Wh) of power at a mass of1,100 grams. The high capacity Nickel-Cadmium cells have approximately 40%more capacity than that of standard cells.Discharge characteristics include nearlyconstant voltage, low internal impedance andshort term discharge rates of up to 100 timesthe amp-hour cell rating. The cells canwithstand over charging and discharging.They feature long storage and service life,and they are rugged and reliable. Nickel-Cadmium technology has a long history and
is well characterized. Future efforts will lookto qualify Li-ion technology for MicroSatapplications since these cells have a 2Xincrease in energy storage density (specificenergy in Wh/kg). Alternate technologiesinclude miniaturized fuel cells which promisea >4X improvement over Li-ion cells.
Solar Array Module
The solar array module would consist of anarray of high efficiency, advanced technologydual-junction gallium arsenide solar cells.Each panel will produce a maximum of 25Watts output per module (actual power outputwill of course be a function of the angle ofincidence of the sun). In this first generationdesign there would be four body fixedmodules that enclose the propulsion tankstructure of the MicroSat, see Figure 2. Thesolar array module consists of a honey combsubstrate to which the solar cells and theircover slides are attached. Each cell has anindividual protection diode and is wired toprovide a 28 volt output. A mass estimate forthis assembly, based on real hardware, isapproximately 224 grams per module. Thecover slide protects the cell from handlingand radiation. The benefit of this dualjunction cell technology is that each of the celljunctions, simultaneously convert a differentpart of the solar spectrum into electricalpower. Conversion efficiencies of 21.5%have been achieved.
Power Management
Table 7 below summarizes the electrical andpower budget for the MicroSat to support arepresentative rescue mission. While theabsolute maximum power consumption (withall equipment on and operating at maximumcapacity) is estimated to be 300 W, themaximum instantaneous consumption permission phase is about 170W. The actualconsumption per phase will be less when theduty cycles of various components areconsidered. For instance, the RF subsystemwill only be active when the MicroSat has anopportunity to communicate with a groundstation. The previously described on-boardbattery packs (two) will provide 100 Wh of
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites14
continuous usage. The solar array is designedto recharge the battery in about eight hours.Thus the MicroSat will need to recharge itsbattery packs after each maneuver. Thepower budget estimated in Figure 7 is basedon relatively near-term Òin-handÓtechnologies and sub-systems and iscurrently considered a conservative estimate.Future flight hardware will be able to takeadvantage of technology improvements thatreduce power consumption, providing greateroperational flexibility for future missions.
There are two means by which the MicroSatwill enter a power-down mode. The first isinitiated autonomously by the systemprocessor as a result of preprogrammedmission operations or as a result of detectingbattery depletion via the master powerdistribution module. The second means isreceipt and confirmation of a power-downcommand from the ground station.
Table 7 Electrical Power Budget and Profile.Subsystem Absolute Maximum per Mission Phase:
When powering down the spacecraft, thesystem processor will send a power-downcommand to the power managementcontroller. The power management controllerwill set the appropriate time into the wake-uptime comparator of the real time clock circuit,and then will remove power to all loads,make a battery voltage measurement, and putitself to sleep. After being awakened by thereal time clock at the programmed time, the
controller will make a battery voltagemeasurement, turn the system processor onand report the battery condition with anestimate of the spacecraft power available tothe system processor. During the powerdown period the battery will be recharged bythe solar cell array.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites15
Thermal Management
Thermal management on the MicroSat will beaccomplished by using industry standardpassive heat transfer mechanisms augmentedby tank heaters for the liquid fuel duringbattery recharge mode. When the MicroSat isperforming LEO operations it will encounterthe solar flux of nominally 1353 W/m2. TheMicroSat will produce an additional 150 Wfor about a 30 minute interval during typicaloperational modes. This power level isreduced to a few watts for approximately 8hours during battery charge mode and thenthe operational cycle can be repeated. Multi-Layer Insulation, (MLI), sun shades, heatsinks, radiators and surface coatings will beemployed for passive thermal control. MLIwill be used around the sensor suite as ageneral purpose insulator from incident sunloading and internal heat loss.
The Solar Array, module encapsulates theH2O2 and N2 tanks and provides insulationfrom incident sun loading and from internalheat loss. Cooling radiators, integral to theSolar Array module provide a passive meansto provide thermal control of that assembly.A radiator panel assembly will be used tothermally control the Avionics module.Unshielded areas will be covered with MLI.The MLI and radiator assembly also serve toprovide some radiation shielding for theelectronics.
GN&C
The core capability of the MicroSat toconduct precision rendezvous maneuvers and
docking with a satellite is defined by itsguidance, navigation, and control subsystem.This subsystem contains functional softwarethat interacts with the various on-boardhardware, processes the signal, selects theappropriate mission scenario, and issues a setof maneuver commands to actuators andvalve drivers to meet the specified missionobjectives. Figure 8 shows a general blockdiagram of the intrinsic GN&C function. Thehardware components include: GPS for orbitestimation, MIR radar for close in ranging,Stereo and HiRes cameras for imaging and3D positioning, Star Tracker for inertialorientation, RF COM subsystem to relay andreceive data, and an IMU for vehiclestabilization and positioning. In addition, theactuator hardware includes attitude controljets, cruciform and axial divert thrusters, andthe 6DOF vision based servo arms to providea hard dock and to insert an adapter to thesatellite service pin connector.
The GN&C function will need to control thespacecraft to better than one milliradian inorientation, five millimeters in translation,10mg acceleration for performing orbitalmaneuvers and 1mg for close in inspectionand docking. An initial set of the IMUperformance requirements is listed in Table 8.Several candidate IMUs can support theserequirements. The LN-200 seems to be agood candidate, however its 1milli-gee bias ismarginal during docking maneuvers. Thiscan be remedied by either procuring a moreaccurate set of accelerometers or making useof the millimeter position resolution of themicroRadar.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites16
MIRRadar
Vis Stereo /Seeker
Startracker
IMU
9 State TargetKalmanTrack Filter
7 StateKalman for Attitude Ref Filter
3 State Kalman for Gyro and Accel biasCorrection
TargetGuidancewith PWM
AttitudeControlwithPWM
DivertThrusters
ACSjets
Hardware
SoftwareGPS
Axialthrusters
Mission ControlOrbit matching
InspectionDocking
6DOFservo arm
Vision feedbackOrbit Estimation
COMLinks
Figure 8 MicroSat rescue mission guidance and control block diagram.
Table 8 IMU Requirements and performance comparison.
The MicroSat software developmentenvironment uses the VxWorks real-timeoperating system (RTOS). VxWorks is acommonly used, well tested, RTOS thatprovides a rapid development environmentfor integration of new software modules. It isalso portable among many processors. It hasbeen used in space applications includingJPL's Mars Pathfinder and the Clementine Ispacecraft. LLNL has more than ten years ofexperience in building software for spacecraftapplications. Many software modules forGN&C, imaging, target tracking, and otherreal-time operating codes can be adapted forthe representative satellite mission.
In general, LLNL's software developmentphilosophy is to maintain maximumflexibility for multiple reuses and to test theintegrated software with realistic hardware-in-the-loop experiments - as soon as and asoften as possible. Figure 9 shows thehierarchical organization of the missionsoftware. Table 9 summarizes the memoryand throughput requirement for theinspection/rescue mission. As indicated, wehave built in 50% contingency margins in theprocessor and memory allocation. The peaksoftware throughput requirement is 90 MIPSout of a 100, including a 50% contingencyfactor. The peak requirement occurs duringdocking, and is dominated by imageprocessing functions.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites18
Integrated Vehicle Testing
Ground performance testing is the key to thesuccess of a MicroSat mission. It is crucial tobe able to repeatedly practice and test theintegrated vehicleÕs ability to performprecision orientation and translationalmaneuvers. These tests should includemaneuvers to achieve orbit matching,endgame chase, inspection, docking, satelliteservicing, and un-docking. Ideally, onewould like to have a 6 Degrees of Freedom(DOF) test environment. However, in mostcases a 5 DOF or 4 DOF environment issufficient.
In order to support the testing of integratedMicroSats, LLNL has developed 4 DOF and5 DOF dynamic air bearing ground testingfacilities.6 The 4 DOF facility is an air railwith 3 degrees of rotational freedom and onedegree of translational freedom. The 5DOFfacility is an air table with 3 degrees ofrotational freedom and two degrees oftranslational freedom. These facilities enablelow cost repeatable end-to-end performancetesting of completely integrated MicroSattestbed vehicles, and full-up performanceacceptance testing of final flight hardware andsoftware before launch.
Air Table
One of our test vehicles on the air table ispictured in Figure 10. The vehicle sits on ahemispherical air bearing supported by 3linear air bearings. The hemispherical airbearing allows ±15o pitch, ± 360o yaw, and±15o roll while the 3 linear air bearingsenables the vehicle to translate in 2dimensions. This facility has been operationalfor over 10 months and has had over 500experimental runs performed on it including 3DOF tracking, 5 DOF tracking, 360o yawmaneuvers, and precision translationalmaneuvers.
Figure 10 Forerunner Test Vehicle on 5 DOFAir Table.
Air Rail
The Engineering Test Vehicle (ETV100) ispictured in Figure 11 on the air rail. The airrail as pictured allowed only one degree oftranslation and was used in this configurationto test the ETV100Õs hydrogen peroxidedivert capability. The air rail is beingupgraded to enable full 4 DOF movement andwill be used to practice high speed diverts,tracking, and docking maneuvers.
Figure 11 Engineering Test Vehicle(ETV100) on 4 DOF Air Rail.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites19
Figure 12 shows a full scale engineeringmodel of the agile MicroSat.
Summary
To date LLNL has developed a preliminaryconceptual design for an autonomous agileMicroSat for on-orbit logistics operations.Figure 12 shows a full scale engineeringmodel of this vehicle. On going efforts arefocused on integrated ground testing ofprototype MicroSat vehicle testbeds and thedevelopment of advanced versions of ournon-toxic propulsion systems that willprovide multi-kilometer per second levels ofon-orbit Dv capability.
Acknowledgments
The authors would like to thank the entireMicroSat Technologies Development Teamincluding Don Antelman, GermaineBrassinga, Eric Breitfeller, Ed English, DougHoward, Joe Lupton, Fred Mitlitsky, DarronNielsen, Elon Ormsby, Gloria Purpura, BillRice, Jeff Robinson, Bill Taylor, DeanUrone, and Bruce Wilson.
The research was part of the Clementine IIProgram and the MicroSat TechnologiesDevelopment Program at LLNL, bothsupported by the Air Force ResearchLaboratory. This work was sponsored by theU.S. Government and performed by theUniversity of California Lawrence LivermoreNational Laboratory under Contract W-7405-Eng-48 with the U.S. Department of Energy.
References
1. Ledebuhr, A.G., Ng, L.C., Kordas, J.F.et al., ÒMicroSat Rescue DemonstrationMission: A Feasibility StudyÓ, UCRL-ID-129880, January 30, 1998
2. Whitehead, C.H., ÒHydrogen PeroxidePropulsion for Small SatellitesÓ, SSC98-VIII-1, The 12th. Annual Utah StateUniversity Small Satellite Conference, 1998.
3. Kordas, J.F., Priest, R.E. et al., ÒStarTracker Stellar Compass for the ClementineMissionÓ, Proc. SPIE Vol. 2466, p70-83,June 1995.
4. Kordas, J.F., Lewis, I.T. et al.,ÒUV/Visible Camera for the ClementineMissionÓ, Proc. SPIE Vol. 2478, p175-186,June 1995.
5. Ledebuhr, A.G., Kordas, J.F. et al.,ÒHiRes Camera and LIDAR Ranging Systemfor the Clementine MissionÓ, Proc. SPIEVol. 2472 p62-81, June 1995.
6. Ng, L.C. and Ledebuhr, A.G., ÒDynamicAir Bearing Guided Intercept and Line-of-sight ExperimentsÓ, UCRL-JC-128922,February 28, 1998
Biography
Dr. Arno Ledebuhr earned an undergraduatedegree in Physics and Math in 1976 from theUniv. of Wisconsin and masters anddoctorate degrees in Physics from MichiganState University in 1982. Dr. Ledebuhr spentthe following four years at the HughesAircraft Company and earned 13 patents inprojection display technology. He has been atLawrence Livermore National Laboratorysince 1986 and led the development ofadvanced sensors for the Brilliant Pebblesinterceptor program and the design of theClementine sensor payload. In 1996 he wasthe Clementine II program leader and iscurrently the MicroSat Technologies Programleader.
Arno G. Ledebuhr 12th AIAA/USU Conference on Small Satellites20
Joseph Kordas earned his B.S. in Physicsfrom St. Procopius College and his M.S. inBiophysics from Michigan State Universityin 1974. He has worked at LawrenceLivermore National Laboratory since 1974.His assignments included lead experimenterfor separator systems Laser IsotopeSeparation Program, sensor engineer BrilliantPebbles Program, lead electronics engineerfor sensor development Clementine Program, and Clementine II Deputy Program Leader.Currently, he is the Deputy Project Leader forthe MicroSat Technologies DevelopmentProject.
Dr. Larry Ng received his B.S. and M.S.degrees in Aeronautics and Astronautics fromthe Massachusetts Institute of Technology in1973, and a PhD degree in ElectricalEngineering and Computer Sciences from theUniversity of Connecticut in 1983 under aNaval Undersea Warfare Center (NUWC)Fellowship. In addition, Dr. Ng received hiscommission as an Air Force officer in 1973and served at the Hanscom Air Force Base inBedford, MA. His work experience include:four years with General Dynamics ElectricBoat Division in Groton, CT., responsiblefor the development of the TRIDENTsubmarine digital control systems; sevenyears at the NUWC where he led theadvanced development of the advanced sonarsignal processing for the Seawolf submarine.Since 1986, he joined the LawrenceLivermore National Laboratory where he iscurrently the group leader of the signal/imageprocessing and control group and is focusinghis research in micro-spacecraft guidance andcontrol and integrated ground testing. Dr. Ngis a member of several professional societies,including honorary memberships in SigmaXi, Tau Beta Pi, and the National ResearchCouncil. He has published numerous papersin signal estimation, and precision vehicleguidance and control.
Mark Jones has worked in electronics since1977. He holds a B.S. degree in ComputerEngineering from the University of thePacific, Stockton, CA. Mark came to LLNLin 1984 and has worked on space-relatedprojects including Brilliant Pebbles, MSTI,Clementine, and Clementine II. Currently,
Mark leads the avionics effort for theMicroSat Technologies Development Project.
Oliver Edwards earned his BS in ChemicalEngineering from Worcester PolytechnicInstitute 1980 and his MS in AppliedMathematics from Carnegie Mellon in 1986.He has worked for Exxon ChemicalCompany doing chemical reactor simulationand Laboratory automation. Currently, heworks for Lawrence Livermore NationalLaboratory doing real-time software systemsand Software Project Management.
Dr. John Whitehead earned his undergraduatedegrees in both science and engineering fromCaltech. He received his doctorate inmechanical dynamics and controls from theUniversity of California, Davis in 1987. Atthe Lawrence Livermore NationalLaboratory, he led the development ofminiature pump-fed rocket engines. Since1995, he has contributed towardunderstanding unsolved propulsionproblems, such as SSTO and Mars departure.
Richard Gaughan is the optical systemsdesigner for the micro satellite developmentprogram. He has designed, built, and testeda wide variety of optical hardware for spaceflight systems.
Michael Dittman earned his undergraduatedegree in mechanical engineering fromCalifornia Polytechnic State University, SanLuis Obispo. He received his masters inmechanical engineering, controls anddynamics from San Jose State University in1993. Since 1987 at Space Systems/Loral hewas a lead mechanical engineer for manysatellite operations including alignments,assembly, integration and test, controlmechanism design and development andspacecraft systems. Beginning in 1997 atLawrence Livermore National Laboratory hehas contributed to the design anddevelopment of micro satellite technologiesincluding low pressure warm gas ACSsystems.
![Autonomous Precision Control of Satellite Formation Flight ...ruk.usc.edu/bio/udwadia/papers/Autonomous Precision...dynamical systems [20–26] and satellite formation systems [27–30].](https://static.documents.pub/doc/80x56/60f67a5f67bc8c763b272407/autonomous-precision-control-of-satellite-formation-flight-rukuscedubioudwadiapapersautonomous.jpg)
