Prisma Satellites - SSC Experiment: PROX GPS

(Duration of 18 days)

The PROX GPS is the first experiment group of two demonstrating Proximity Operations and Final Approach and Recede Maneuvers (PROX/FARM) using GPS as well as the Vision Based Sensor (VBS). These experiments will exercise forced motion flight of the Mango spacecraft in a virtual structure spanned around the Tango spacecraft. The Mango spacecraft navigates in a flight map consisting of a series of nodes and allowed transition between these nodes. The flight map is fixed to the orbital frame in the case of GPS based navigation while in the VBS based case, it is fixed to the Tango body. As in the AFF experiments, the orbit controller is based a Model Predictive Control framework but here, with shorter prediction horizon settings tailored for the forced motion case. The PROX/FARM experiments are defined to represent navigation around a large space structure in situations typically occurring in on-orbit servicing, inspection, and assembly.

The PROX GPS experiment exercises the GPS based part of the PROX/FARM experiments. The relative distances range from approximately 100 m down to 10 m. The performance of the VBS at medium and close range distances will also be assessed.

Reports:

Re-run of PROX GPS Approach to 2 meter

Finally on the 25th of january we had some time to perform an other very close approach based on GPS navigation. This time the DVS, supplied by Techno System, gave us the desired pictures. Make sure not to miss the video below.

At the start of the video Mango is in a position just in front of Tango, distance about 20m. Mango is then using it's thrusters to slowly move in closer to Tango, holding a position at 2 meter distance for about 5 minutes and then receding back to a safer orbit.

During the last weeks Mango, Tango and the operations team have had a bit more relaxed schedule. These weeks have been scheduled for platform and software maintenance as well as analysis of previous results. The outcome was a patch of the onboard software which was successfully uploaded and applied the week before the close approach. The pace is now picking up again with more exciting experiments to come, and with more experiments the reports will start appearing more frequent again!

Written by

Robin Larsson

2011-02-11 / 22:11:27

PROX GPS Approach to 2 meters

Tonight we performed not one, but two successful Mango approaches towards Tango where the closest distance between the two satellites was as small as 2 meters. Mango was controlled to within 3 decimeters from the commanded approach and recede trajectory. This was a part of the SSC PROX GPS experiment, where the guidance and control have been developed by SSC and the GPS navigation by DLR.

Pictures and videos are coming soon.

First pictures downloaded on last passage; here is one from 6 meters distance, captured during the second approach:

For the technical reader I can add that Mango is approaching from a position ahead of Tango, because of this Mango is in a slightly higher orbit during the approach and explains why Tango is in the bottom half of the image. The DVS camera is also mounted on the outside of Mango, away from the sun. This explains why Tango is on the left half of the image. Target pointing attitude guidance was not active on Mango during this exercise. The DVS camera is simply pointed opposite of the velocity direction.

More pictures and videos are coming after the weekend.

UPDATE 22/12-2010

We had some problems downloading the image sequences following the image posted above, after investigation we must sadly conclude that the rest of the images have been corrupted. Luckily the data gathered with the VBS is nominal and confirms the two successful approach and recede manoeuvres. But sadly the pretty pictures have been lost. If the timeline allows we might schedule in a remake of the very close range approach beginning 2011.

Written by

Robin Larsson

2011-02-11 / 22:11:27

PROX GPS Inspection Tour

Late last Friday Mango successfully inspected four different locations around a virtual space station as part of the SSC experiment PROX GPS.

Having only the small Tango satellite to normally fly around, in this experiment the avoidance region around Tango had been beefed up considerably. During this experiment Mango did not see Tango as the 1x1x0.4 m sized satellite, as in reality, but a 60x80x30 m space station. The super sized Tango is a more representative target for future missions involving on-orbit servicing, inspection, and/or assembly.

Inspection tour of the Tango space station as performed in-orbit last Friday. Click on the figure to open the full sized version in a new window.

Displayed is the inspection tour from three different angles. Commanded was four different inspection points, in the figures marked as A, B, C & D. Each point was held for ten to thirty minutes before Mango was commanded to find a suitable path and move on to the next inspection point.

Mango stayed out of the avoidance regions during all transfers and held the inspection points to within only a few decimetres, well within the commanded control box. The PROX inspection tour resulted in a travel distance of 145 m during two full orbits and spent 35 cm/s on orbit manoeuvres. The smallest distance to the real Tango was 7m (during the transfer between A & B).

The inspection tour is displayed in blue, also displayed in the figure is the entrance and exit trajectories in green and red performed by AFF.

Compared to the validation simulation performed earlier, the performance was slightly better in orbit. In orbit the inspection tour required about 10% less delta-V, but then a slightly different trajectory had been selected by Mango for some of the transfers.

Written by

Robin Larsson

2011-02-11 / 22:11:27

PROX GPS part 2 started

This week the second part of PROX GPS has started. It is a twelve day long campaign and will continue where the first part ended. There are two main activities to be performed.

The forced motion capabilities of Mango will be put to further tests when Mango is asked to inspect specific points around a virtual structure that is centred around Tango. The virtual structure is used to mimic a mission where a much larger object is subject to inspection.

On every second day we also run specific tests for the VBS, these tests typically include passive orbits very close to Tango (distance of 10 to 20 m) and with Mango constantly pointing the VBS towards Tango. This creates a very challenging situation for the GPS navigation as the GPS antennas on Mango and Tango are most of the time pointing in very different directions. This was the case during the last orbit, illustrated below, were the GPS antennas where pointing in directions which were 90 degrees away from each other.

Image generated with Celestia from data downloaded during the last passage. Mango is just passing directly under Tango with a distance of 10m (the small green arrow points to the centre of Earth and the dashed green line shows the relative flight trajectory). The orbit is maintained by AFF to within 1m in radial direction. The VBS can be seen pointing directly to Tango.

The image was captured when Mango and Tango were just about to enter eclipse and the background is thus very dark, but to the left is a large part of Japan, above Tango a small part of South Korea and to the right a part of Russia and further back China is visible.

Written by

Robin Larsson

2011-02-11 / 22:11:27

SSC PROX Experiment, R-bar approach, the Moon and Taiwan relative orbits

This post will give a different more complete view on the formation flying done during the last day of the PROX GPS part 1 experiment than what was possible with the videos posted earlier.

The following image presents the position of Mango relative to Tango from four different viewpoints. Position of Tango is marked by a blue star (*). The data spans 24h, from 5am on the 26^th until 5am on the 27^th of October. Only 9 hours (6 orbits) is used for PROX as the satellite is not visible for many hours during daytime and during the late hours of the night no personnel is available. Because of this Mango is left in a loosely controlled stand-by orbit (AFF) which includes some minimum eccentricity/inclination separation to ensure safety in case of some anomaly.

The scenario presented is a typical PROX day. Mango is waiting in a stand-by orbit maintained by AFF, in the image these orbits have been colored green. AFF is then given the command to change the orbit to a starting point for PROX, in this scenario 20m ahead of Tango on the along-track axis. This is done by AFF autonomously un-assisted by ground. AFF is only given the final reference point, and has to plan the handover trajectory in-orbit.

When this is completed one orbit later the PROX experiment can start, the PROX phase is colored blue. On this day the PROX experiment involved first a forced motion 360 degree inspection of Tango, an approach and recede on R-bar (from 10m down to 7m), after that aligning Mango, Tango and the Moon, and finally aligning Mango, Tango and Taiwan. All this was done with Mango pointing the VBS and DVS at Tango to gather calibration data and take pictures of Tango. The navigation used is based on GPS, which is forced to operate under very tough conditions. GPS antenna on Mango and Tango are under long durations even pointing in opposite directions.

After the PROX phase Mango is switched back in AFF mode and commanded to initialize the same stand-by orbit as was used in the beginning of the day. This transition back to the stand-by orbit is done again by AFF autonomously un-assisted by ground. AFF is only given the final reference trajectory and has to plan the handover trajectory in-orbit.

Written by

Robin Larsson

2011-02-11 / 22:11:27

Team Pic(k) of the month

During the impressive GPS based PROX activities in October, there were quite a number of opportunities for DVS imaging, so we used them to their full extent. The results are some of the nice sequences in the videos below. A team favourite, though, that has made its way on to the walls and computer desktops of most of the engineers here are SSC is the lovely picture below, showing Tango some 15m away, Earth limb in the background, and the dark of night approaching on the right.

Written by

Nils Pokrupa

2011-02-11 / 22:11:27

SSC PROX Experiment, the Moon and Taiwan

On the last day of the Swedish Space Corporation experiment PROX GPS part 1 two forced motion phases where included. In these Mango and Tangos position where aligned in a specific direction. This represents a typical scenario of future formation flying missions where a distributed telescope is created using two satellites.The images were taken using the DVS supplied by Techno System.

For the first experiment we aligned Mango, Tango and the Moon such that Mango could take a few pictures of both Tango and the Moon. For this to be possible Mango were commanded to hold a forced position 6 meter off in cross-track and 2.5 meters in radial during 5 minutes. The distance between the two satellites is about 10 meters.

In the second experiment, executed in the following orbit, we aligned Mango, Tango and Taiwan. Taiwan was selected as the next conference article on Prisma is presented here this week at the 6th International Workshop on Satellite Constellation and Formation Flying (IWSCFF 2010)http://iwscff2010.nspo.org.tw/IWSCFF2010/index.html

During 5 minutes Mango is commanded to hold a forced position 5.5 meters off in both radial and cross-track compared to Tango. The total distance is about 10 meters.

Unfortunately it was very cloudy this morning in Taiwan, only a part of the coastline can be seen if you look carefully to the right of Tango around 16s in the video.

In this video the DVS images have been embedded on top of a video generated with Celestia based on in-orbit navigation data. Here it is clearly visible that Mango, Tango and Taiwan are aligned.

Written by

Robin Larsson

2011-02-11 / 22:11:27

Two more videos from fly-around on 15 m distance

Two more images sequences have been uploaded, these ones shows part of two different forced motion fly-arounds, both on 15 m distance.

Written by

Robin Larsson

2011-02-11 / 22:11:27

PROX forced motion fly-around on 10m

Today we executed three fly-arounds on 10 m distance, two in-plane and one 45 degrees out–of-plane. These motions where aimed to gather VBS calibration data as well as exercising the relative orbit controller. Two other PROX experiments where also performed involving some smaller approach and recede maneuvers.

The figure below displays one of my monitoring pages after a contact period with the satellites. The data displayed is the essential telemetry which is down linked first in each passage. The essential telemetry is sampled once per minute and gives a quick look of what happened during the last orbit. The full telemetry is delayed a few orbits due to the limited downlink capacity.

Displayed here is the last fly-around of Tango. During the full maneuver Mango is pointing the VBS towards Tango to gather data, this also means non optimal conditions for the GPS navigation, as for a long period of time the GPS antenna on Mango is even pointing towards Earth. Even so the maneuver has been performed successfully.

The top graph shows the Feedback Orbit Controller position error in meters, the controller has been set to keep the control errors smaller than one meter and at the same time minimize the fuel consumption. The middle one relative position of Mango with respect to Tango also in meters and the bottom one shows the accumulated number of seconds of commanded thrust. For this maneuver a total of 0.14 m/s were used (=30s on time).

Written by

Robin Larsson

2011-02-11 / 22:11:27

Approach and recede down to 7m video

Finally here are the images taken from the DVS, supplied by Techno System, during the approach and the recede down to 7 m, here compiled into a short video.

Written by

Robin Larsson

2011-02-11 / 22:11:27

Approach and recede down to 7 m

We are very busy in the control room operating the two satellites at very close distances most of the time. All the time the SSC developed safety modules are making sure that the constellation is collision free and safe.

The newly developed SSC guidance and control modules which are now being tested have proven to be robust, flexible and capable to accurately control the formation in a delta-V lean way also in the very close proximity of Tango. The navigation used is the GPS navigation developed by DLR which has been configured for a scenario which involves more frequent thrusting than normal.

Read more about the guidance and control modules in this 2006 AAS Breckenridge article:

Fuel Efficient Relative Orbit Control Strategies for Formation Flying and Rendezvous within PRISMA
R. Larsson, S. Berge, P. Bodin, U. Jönsson

Yesterday night we performed a V-bar approach and recede from 20 m down to 7 m distance between Mango and Tango. At 7 m Mango and Tango was flying in exactly the same orbit, except Mango was flying 1 ms ahead of Tango! The complete approach, hold and recede were performed in 40 minutes. The relative velocities where about 1 cm/s.

During the complete maneuver Mango was taking images of Tango using the DVS. These images will arrive on the blog soon.

Written by

Robin Larsson

2011-02-11 / 22:11:27

First close range PROXimity experiments performed

During the evening we have successfully completed two PROX GPS experiments, consisting of proximity fly-around orbits in the range 10 to 20 m. The trajectory followed was a forced motion trajectory where the operator has a much higher freedom to select the type of relative motion to be performed. This freedom has a cost, which is higher fuel consumption. The two experiments performed tonight was a warm up for what is coming next, the forced motion is not so “tough” and Mango is always pointing the GPS antenna to zenith.

The two experiments are visualized in the following image generated by Celestia using flight TM:

For the technical reader here are some more details:

Relative trajectory during the two experiments. Tango is marked by "+" and the last position of Mango is marked by "o".

Delta-V consumption 0.17 m/s for the two PROX experiments performed during two orbits (excluding entrance and exit trajectory)

The controller used is the same as is used during AFF, but it has been configured with a shorter control cycle. The control box has also been reduced, the controller is set to keep the error smaller than three meters in X and one meter in Y and Z. The control error is shown in the following figure,

It is quite easy to see the difference in the control error behavior during the forced motion fly-around and during the longer hold phase.

Written by

Robin Larsson

2011-02-11 / 22:11:27