SPHERES ISS TS001 Presentation - ssl.mit.edussl.mit.edu/spheres/iss/reports/SPHERES ISS TS001 Presentation.pdf · SPHERES Payload Systems Inc MIT Space Systems Laboratory Cambridge,

SPHERES

Payload Systems Inc

MIT Space Systems LaboratoryCambridge, MA

[email protected]

Synchronized Position Hold, Engage, Reorient, Experimental Satellites

ISS Test Session 1 Results

SPHERES

Payload Systems Inc2

Outline• Test Session Objectives• Timeline Summary• Setup• Results Analysis

– Test 1: Quick checkout– Tests 2 & 3: Open-loop rotations using different mixer algorithms– Test 6: Closed-loop tests involving gyroscopes only– Tests 4, 8 & 8.3: Closed-loop tests involving the ultrasonic navigation system

• Consumables consumption• Conclusions• Lessons Learned• Future Actions• Points of Contact• Revision History• Appendix A. Converted IMU data• Appendix B. Tests Operational Results

SPHERES


Test Session Objectives• Primary objective: ”Hardware Checkout”

– Powering the satellite– Establishing communications– Uploading the test session program– Running tests– Collecting data

• Science objective: demonstrate estimation and control algorithmsleading to autonomous docking– Thruster mixers performance - to demonstrate the ability to perform open-

loop maneuvers– 3DOF quaternion rotations - to initialize the satellite pointing towards its

target– Formation flight - to validate closed-loop control with the estimators– Docking - to demonstrate the full docking process

SPHERES


Test Description Start time IntervalT 1 Quick checkout with IMU data download 17:01:28 06:01T 2 Open loop rotations, old mixer 17:07:29 07:34T 3 Open-loop rotations, new mixer 17:15:03 09:23T 4 Beacon track attitude PD 17:24:26 07:05T 6 Closed-loop XYZ rotation 17:31:31 02:53T 6 Closed-loop XYZ rotation 17:34:24 05:41T 8 Dock Free Short S#1 PD 17:40:05 04:17T 8 Dock Free Short S#1 PD 17:44:22 03:52T 6 Closed-loop XYZ rotation 17:48:14 02:48T 6 Closed-loop XYZ rotation 17:51:02 02:46T 2 Open-loop rotations, old mixer 17:53:48 05:43T 3 Open-loop rotations, new mixer 17:59:31 04:12T 3 Open-loop rotations, new mixer 18:03:43 03:58T 8.3 Dock Range only S#1 18:07:41 02:31T 1 Quick checkout with IMU data download 18:10:12 03:00

Timeline Summary• Scheduled start: 15:05GMT (10:05CDT) Thursday 18-May-2006• Actual start: 15:20GMT• Hardware locate: ~30m• GUI Configuration: ~40m• Program plan review & program upload: ~20m• First test: ~17:00GMT• Tests: 15 total, 1h 12m operating time, 4m 27s avg. time per test

SPHERES


Setup• Setup for the first test Session took approximately 90 minutes

before the first test was executed– 30 minutes used to locate all SPHERES hardware

• The SPHERES beacon tester, previously believed misplaced, was found

– 40 minutes used to resolve a problem with the SPHERES GUIconfiguration

• The configuration file (gui.ini) was saved as “read-only”, which prevented theGUI from starting

• The SPHERES team found the problem and determined a solution in real-time• The crew executed the changes in that specific SSC

– 20 minutes used to read the test plan and load the satellite• The satellite loaded successfully within five minutes

SPHERES


Results Analysis Overview• The tests ran during this session correspond to the mission

objectives as follows:

– Hardware Checkout Test 1: Quick Checkout

– Thruster Mixers Tests 2 & 3: Open-looprotations using different mixers

– 3 DOF Rotations Test 6: Closed-loop rotationsusing only gyroscopes

– Formation Flight, Docking Tests 4, 8, 8.3: Closed-looptests using ultrasound navigationsystem

SPHERES


Test 1: Quick Checkout• Objectives

– Test all thrusters & IMU sensor– Collect data for mass-identification algorithms

• Quick checkout was run twice, at the start and at the end of the test session• Successfully collected necessary data, demonstrated thruster operations, and

closed-loop control using the gyroscopes

• False positive on demonstration of closed-loop rotation control…more on section for Test 6

Initial quick checkout experiment results Final quick checkout experiment results

SPHERES


Tests 2 & 3: Open-Loop RotationsOverview

• Objectives– Allow comparison between a basic mixer algorithm and an

advanced mixer algorithm (mixers determine the length of thrusterpulses for a requested force/torque)

– Basic mixer assumes all thrusters are identical and pressure isconstant

– Advanced mixer accounts for thruster differences and drop inpressure when multiple thrusters are open

• Both tests performed a series of three open-loop 180 degreerotations, one about each body axis– Calculated desired torque to obtain a ∆ω of 0.3142rad/s (rotate

180 degrees in 10s)– Commanded that torque to the mixer

SPHERES


Test 2: Open-loop RotationsBasic Mixer: Data

• Test ran two times– During the initial tests the tank was partially inserted– The tank was fully inserted for the second run– No noticeable differences in the resulting ∆ω, therefore concluded that the first test

provided valid data• X and Y axis rotations had a smaller ∆ω than requested• Z axis rotation had a larger ∆ω than request

– Behavior makes sense physically, as the inertia is smallest about the Z axis– Indicates that the models used for the basic mixer do not estimate the inertia and/or

torques appropriately

First run of Test 2 Results

Second run of Test 2 Results

SPHERES


Test 2: Open-loop RotationsBasic Mixer: Plots

• The data and crew notes show that the X & Y rotations did not complete to180 degrees; the Z rotation overshot

– This can be seen from the larger magnitude of the Z pulses in the plots• Some coupling is noticeable on the Y axis during X & Z rotations• During the second test the -X pulse (stop rotation) did not stop the rotation

completely; reason under investigation

Second run of Test 2 ResultsFirst run of Test 2 Results

SPHERES


Test 3: Open-loop RotationsAdvanced Mixer

• The mixer commanded larger pulses, as expected• The Z axis performance improved noticeably

– The start and stop pulses were different due to the differences in thruster strength• The X & Y performance did not improve as much as expected

SPHERES


Test 6: Closed-Loop Rotations• Objective: demonstrate closed-loop attitude control• Same 180 degree sequences as for tests 2 & 3• Tests did not complete successfully due to FLASH memory corruption

– FLASH was only partially corrupted, therefore checkout test succeeded, but thistest failed

– Closed-loop path differences between checkout and full tests:

• The FLASH corruption error was found after the test session using telemetrydata made available within six hours of the test session

GyroscopeA/D

Bias

ConversionFactor

PlantCommand Mixer+

-

+

-

Checkout path

FLASH

Gain

SPHERES


Tests 4, 8, 8.3: Closed-loopTranslations

• Objectives:– Test the ultrasound navigation system– Demonstrate control algorithms to maneuver the satellite– Demonstrate maneuvers for autonomous docking

• Closed-loop control was not possible due to the FLASH corruption• Ultrasound navigation system provided encouraging results

– Range, range rate (velocity), and quaternion (direction) calculated to within desired performance– Convergence within 5s– Range/quaternion curves correlated with motion of beacon or satellite seen in video

Results for tests 4 and 8 (including corrupted rotation rates)

SPHERES


Consumables Consumption• Efficient battery usage:

– 0h 13m Setup and program load– 1h 12m Running Tests

• Tank usage was not as efficient:– 15% Running Tests– 39% Inadvertent tank vents

• Tank vent maneuver entered by depressing Enable pushbutton for more than10 seconds

• Since LED flashes for five seconds, the video shows the crew usuallydepressing the pushbutton for extended periods of time

• Need to improve feedback to the crew on SPHERES control panel

• Available resources after Test Session 1– Approximately one hour of battery– Approximately 50% of tank

SPHERES


Conclusions• Primary objective “Hardware Checkout”: SUCCESS

– Powering satellite: Success– Establishing communication: Success (nominal packet drops)– Uploading test session program: Success– Running tests: Success– Collecting data: Success

• Science Objectives: Partial– Thruster mixer performance: Success

• Need further algorithm development

– Closed-loop rotations: Failed due to FLASH corruption– Formation Fight and Docking: Partial

• Demonstrated estimator performance• Unable to perform closed-loop control due to FLASH corruption

SPHERES


Lessons Learned• ISS environment is highly benevolent for SPHERES

– No air flow effects– No infrared or ultrasound interference– Crew highly skilled to deploy satellites

• Quick Checkout procedure not sufficient to test closed-loopcontrol of both hardware and software

• The crew needs more feedback when the satellite downloadsdata

• The “Enable LED” operation must be simplified• Pre-mature test termination due to communication failure must

be minimized– Satellite should continue to download data unless commanded to

stop by the crew or because of safety requirements

SPHERES


Future Actions• All the critical values in the closed-loop path which were stored in FLASH are

now uploaded every time a program is loaded• All the programs uploaded to the ISS have been reviewed; none of them

contain the FLASH corruption bug• The internal MIT/PSI procedures for delivery of a satellite to NASA have been

updated to check all FLASH values• SPHERES GUI executable will be updated (potentially by Exp 14) to indicate

to the crew when data download is taking place• Unclear what actions to take to solve communications issues after this session• SPHERES GUI executable updated to allow to function even if the files it uses

are read-only when starting the program– Modified the code to check all files, not just gui.ini, for this issue

• SPHERES Core software has been updated so that the “Enable LED” turnssolid on during all times after the “enable” command (pushbutton > 0.5s) hasbeen received

SPHERES


SPHERES Points of Contact

Payload Integration:John MerkPayload Systems Inc.(617) 868-8086 [email protected]

Principal Investigator:Prof. David MillerDirector, MIT Space Systems Laboratory(617) [email protected]

Space Test Program (Code WR1):Maj Matthew Budde, USAF, (281) 483-7576Mark Adams, SAIC, (281) 483-3520

Science Lead:Dr. Alvar Saenz-OteroMIT Space Systems Lab.(617) [email protected]

Graduate Students:Simon Nolet (PhD)[email protected]

Swati Mohan (MS)[email protected]

Nicholas Hoff (MS)[email protected]

[email protected]

SPHERES


Revision HistoryDate Version Notes Released by2006/Aug/08 1.0 Initial release alvarso

SPHERES


• The raw (bit counts) data collected during tests 1, 2, and 3 isconverted into metric units– metric = (raw - bias) * scaling factor

• Results:– Accelerometer biases need correction– Unable to validate accelerometer scaling factors without further

analysis– Gyroscope data with reasonable margins

Appendix A: Converted IMU DataOverview

SPHERES


Appendix A: Converted IMU DataPlots

First Quick Checkout Results Second Quick Checkout Results First Test 3 Results

Second Test 3 Results Test r Results

SPHERES


Appendix B: Operational Results• Summary of results from the GUI and crew notesTest Description GUI result Crew notesT 1 Quick checkout with IMU data

download1 = success

T 2 Open-loop rotations, old mixer 3 = stopped by crew 1. First 2 rotations were less than 180 deg (~150-160)2. Test was manually stopped after ~3 minutes of "running" on GUI.

T 3 Open-loop rotations, new mixer 1 = success 2nd axis appeared to be 20-30 deg less than 1803rd axis appeared to be 20-30 deg greater than 180

T 4 Beacon track attitude PD 0 = lost comm Satellite performed a series of ~10deg changes in one axis insteadof the expected response

T 6 Closed-loop XYZ rotation 0 = lost comm No jet firings at all during test.T 6 Closed-loop XYZ rotation 1 = success* No thruster firings. Comm light confirmed green during test.T 8 Dock Free Short S#1 PD 3 = stopped by crew Beacon was offT 8 Dock Free Short S#1 PD 3 = stopped by crew satellite performed plus/minus ~10 deg rotations in Y and drifted in

other axis.T 6 Closed-loop XYZ rotation 4 = lost comm no response from the satellite. Enable, Comm, Bat confirmed green

on GUI.T 6 Closed-loop XYZ rotation 1 = success*T 2 Open-loop rotations, old mixer 1 = success But first braking pulse was insufficient to stop the rotationT 3 Open-loop rotations, new mixer no resultsT 3 Open-loop rotations, new mixer 0 = lost comm Satellite performed 1st rotation with a 20-30 overshoot and then

initiatedthe 2nd rotation. Test ended without any other jet firings.

T 8.3 Dock Range only S#1 1 = success** No response from the satellite. Beacon green light flashed duringtest.

T 1 Quick checkout with IMU datadownload

1 = success 2nd part was a series of rapid fire jet firings with very little rotation

SPHERES ISS TS001 Presentation - ssl.mit.edussl.mit.edu/spheres/iss/reports/SPHERES ISS TS001 Presentation.pdf · SPHERES Payload Systems Inc MIT Space Systems Laboratory Cambridge,

Documents