Prisma Satellites - DLR Experiment: AFC 2

(Duration of 18 days)

Abstract
The second DLR's Autonomous Formation Control (AFC) experimental slot is dedicated to the fully autonomous formation keeping and reconfiguration based on GPS relative navigation. Several constellations will be acquired autonomously by AFC and maintained over prescribed time intervals.

Description
This slot represents a comprehensive demonstration of autonomous formation acquisition, keeping and reconfiguration based on GPS relative navigation. It can be logically split into several units, each dedicated to the acquisition and maintenance of a prescribed formation geometry. The MAIN and TARGET spacecraft attitude modes are chosen to guarantee the best GPS antenna orientation. The Delta-V attitude guidance function is enabled based on the results of the GPS Calibration campaign and on the experimental slot DLR AFC 1.
Typically a set of different constellation geometries is defined (ca. 3 days allocated to each constellation for a total of 6 configurations). Mean along-track separations will vary between 0 and 1 km. Minimum spacecraft separations will be around 50 m. The initial and final formation geometry corresponds to parallel relative eccentricity/inclination vectors. In the sequel a draft sequence of formation configurations is provided (the ultimate formation geometries will be selected shortly before the experiment). The relative orbital elements are expressed in meters:
1) aDa = 0, aDex = 0, aDey = 200, aDix = 0, aDiy = 200, aDu = 0.
Initial geometry. Ideal relative eccentricity/inclination vector separation (relative argument of perigee at 90 deg). Centered bounded relative motion. Duration: 3 d
2) aDa = 0, aDex = 20, aDey = 160, aDix = 20, aDiy = 160, aDu = 22.49.
Difference in orbit inclination estaslished. Parallel relative eccentricity/inclination vectors. Out-of-plane relative orbit control needs execution of maneuvers because of J2 secular perturbations. Centered bounded relative motion. Duration: 3 d
3) aDa = 0, aDex = 60, aDey = 130, aDix = 20, aDiy = 110, aDu = 215.46.
Decrease minimum separation perpendicular to the flight direction. Non-zero angle between relative eccentricity/inclination vectors. Mean along-track separation of 200 m. Duration: 3 d
4) aDa = 0, aDex = 70, aDey = 90, aDix = 50, aDiy = 90, aDu = 212.65.
Increase amplitudes of relative eccentricity/inclination vectors. Decrease relative argument of perigee and relative argument of latitude. Mean along-track separation of 200 m. Duration: 3 d
5) aDa = 0, aDex = 50, aDey = 40, aDix = 50, aDiy = 40, aDu = 5.62.
Minimum separation perpendicular to the flight direction. Zero mean along-track separation. Duration: 3 d
6) aDa = 0, aDex = 100, aDey = 0, aDix = 100, aDiy = 0, aDu = 0.
Final geometry. Parallel relative eccentricity/inclination vectors. Zero relative argument of perigee. Duration: 3 d
The DLR's AFC experiment can be considered complete after completion of this experimental slot.

Reports:

GSOC supports PRISMA operations

On March 14, 2011 or 272 days, 4 hours, 41 minutes and 31 seconds into the mission, PRISMA operations were handed over to GSOC⁴ for a limited five-month period. The purpose is to provide an opportunity to gather experience in operating multi-spacecraft missions while completing the experiments from the nominal PRISMA timeline plus two additional experiments allocated exclusively for GSOC. The five month period of GSOC operations consists of eight timeslots in the nominal timeline according to Table 1, and at completion of GSOC's operations, PRISMA will be handed back to SSC for extended mission experiments.

Experiment Duration [days] Experimenter

Autonomous Formation Control 2 19 DLR

Autonomous Formation Flying Completion 8 SSC

High Performance Green Propellant 4 30 ECAPS

Autonomous Rendezvous Completion 8 SSC/GSOC

On Orbit Servicing (new) 14 GSOC/SSC

Autonomous Orbit Keeping 30 DLR

Formation Re-acquisition (new) 10 GSOC

Re-handover to SSC 0 GSOC/SSC

Table 1: Timeline overview during GSOC operations period

The common objective of the timeline in Table 1 is to approach Tango in a multitude of ways, especially based on visual information, without the ground control center in the control loop. While performing these Tango approaches, high resolution color photographs will be collected with the PR-camera for on-ground offline processing to aid future missions utilizing formation flying technologies demonstrated in the PRISMA mission. A visualization when taking a photograph with the PR-camera can be seen in Figure 1, where the real photograph is embedded to the right.

GSOChandoverBlogPicture

Figure 1: Animation of Mango acquiring Tango at 4 meters distance. To the right is the real photograph embedded from the PR camera

The mission operations concept at SSC has been adopted by GSOC to be able to deploy a parallel operations centre within the limited preparation time available. The operational concept consist of an operational environment based on SSC's command and control software RAMSES¹ and mission specific tools developed by SSC, which all have been delivered to GSOC. Some specific mission tools were not possible to transfer to GSOC whose functionality has either stayed at SSC or has been re-developed at GSOC. The validated procedure database which is part of the operational concept has also been transferred to GSOC.

Throughout the mission GSOC staff has been part of the PRISMA operations team at SSC to learn and observe the daily operations, which has provided SSC with welcome manpower and satellite operations experience.

During SSC operations only the Esrange ground station has been used, but GSOC will expand the ground station network to include and utilize at least the German ground station Weilheim. The possibility to include two other ground stations around the world is still under investigation. This is the first opportunity for the SSC developed ground system RAMSES¹ to be used at multiple ground stations within the same mission.

By the end of the nominal mission ending with the re-handover to SSC, it is expected that a significant amount of delta-V still is available which has enabled SSC to announce the opportunity to define additional experiments for anybody interested. The additional experiments will take place after the re-handover to SSC and interested parties can contact SSC at bjorn.jakobsson@ssc.se or bengt.larsson@ssc.se.

¹ Rocket And Multi Satellite EMCS² Software

² EGSE³ Mission Control System

³ Electrical Ground Support Equipment

⁴ German Space Operations Center

Written by

Thomas Karlsson

2011-09-02 / 12:42:35

AFC2 sets up a photo shoot for Tango

After 12 days of the AFC2 experiment, on March 27^th 2011, the on-going fly-around and inspection phase has reached the closest planned relative orbit. Referring to the plan, it is mentioned as geometry “I” and it is characterized by a minimum relative distance of 30m.

This scenario, together with the “Target pointing” attitude mode, sets up an appealing chance for performing some activities with the Digital Video System (DVS): a photo shoot of the Tango satellite was realized!

As part of the operational activities, a video sequence starting at 22:40:0.0 UTC was planned. It lasted 2850s and shooting pictures every 75s. At the starting epoch Mango was occupying the green spot of Fig. 1. Each dot on the nominal relative orbit identifies a scheduled picture. In particular, the ones marked in red correspond to the pictures shown in Figures 2, 4, 6, respectively named pictures A, B and C. Our strategy can be easily understood by focusing on the Radial/Cross-track view of Fig. 1, where Mango moves in clockwise direction. Picture-A was meant to show Tango at almost the minimum relative distance with no Earth in the background. Pictures-B and -C were scheduled some time later in order to catch the re-approach to Tango with the Earth entering the camera field of view.

Fig. 1 Nominal relative trajectory of Mango (blue line) w.r.t. Tango (origin).

Afterwards, the real photos where analyzed to improve our understanding of the obtained images. To this aim, they were compared with the corresponding simulated-images produced by a specific simulation tool that models the camera behavior. The relative orbit geometry is rebuilt replaying the orbit estimated on-board. Whereas the attitude is given by the nominal guidance computed starting from such orbital information.

This post-processing campaign mainly addressed two objectives: improving our insight of the real images and tuning our DVS planning tool. Both aspects can be fruitful in view of future operational campaigns, also involving images taken at a larger distance which may be of difficult interpretation.

Fig. 2 Tango at 22:45:00 UTC (Picture-A)

Fig. 3 Simulated image of Picture-A

Fig. 4 Tango at 23:17:30 UTC (Picture-B)

sim_30

Fig. 5 Simulated image of Picture-B

Fig. 6 Tango at 23:23:45 UTC (Picture-C)

sim_35

Fig. 7 Simulated image of Picture-C

The difference in the timings between real and simulated images is a consequence of having replayed the on-board estimated orbit. Such information is in fact available every 10 s with some drift with respect to the GPS alignment.

In all the simulated images, the dashed-black square defines the camera field of view. The red and green segments in the center of the plot respectively indicate the x and y direction of the camera-fixed frame. Thanks to this information one can visualize to where the bore-sight (aligned to z) points. The layout of the Tango spacecraft is simplified as a box whose darkest face stands for the solar panel. The body-frame is marked with dashed red-green-blue segments, respectively referring to x-y-z axis. This helps in completing the rough representation of the spacecraft configuration: the +y axis lies between the two tilted FFRF antennas, the +z axis coincides with the normal to panel. The last two segments dashed yellow and black mark the Sun and Nadir directions in Tango body-frame. The angle between Sun and +z directions is due to the particular setting of the attitude guidance which currently favors the Zenith bearing of the Sun/Zenith pointing configuration. As expected, in Fig. 5 and 7 it can be noted that the Earth appears in the field of view.

Written by

Gabriella Gaias

2011-03-31 / 17:29:51

Planning the fly-around and inspection phase during AFC2

On March 23^rd 2011, as a consequent step after the successful rendezvous phase, the AFC2 experiment slot has entered the 4-days long fly-around and inspection phase. An updated plan including this phase is shown below in Fig. 1 covering a total of 12 days of experiment time. The DLR’s GNC functions are asked to reduce the separation between Mango and Tango through the acquisition of formations E to I while keeping anti-parallel relative eccentricity and inclination vectors. A specific challenge of this operational phase is given by the usage of the so called “Target pointing” attitude mode (cf. Fig. 1, column ACS mode). In this attitude mode, Mango is pointing its Digital Video System (DVS) camera to the Tango spacecraft during the fly-around to take pictures. The 360 deg rotation of Mango causes a degradation of the GPS-based navigation due to the unfavorable pointing of the GPS antennas. Despite the reduction of relative navigation accuracy, the GNC subsystem has to demonstrate its robustness and maintain the control of the formation. As shown in the plan, every second day DVS sessions are planned to collect sequence of images of Tango at various separations. The last formation to be acquired and controlled during this phase (I) is characterized by a minimum separation of 30m between Mango and Tango.

Written by

Simone D'Amico

2011-03-30 / 16:56:48

AFC2 autonomous rendezvous phase complete

The AFC2 experiment slot is proceeding according to plan. The first 8 days of the experiment prescribed a total of 5 formation geometries to be acquired and maintained autonomously by the Mango spacecraft using the navigation and control functions developed by DLR. All reconfigurations have been smoothly handled by the spacecraft platform and closely monitored by GSOC and the DLR’s experiment team. Key results obtained in orbit are depicted in Figures 1 and 2 below. Fig. 1 shows the relative motion of Mango w.r.t. Tango during the reconfiguration from A to B (cf. AFC2 plan from previous post), whereas Fig. 2 shows similar results obtained during the reconfiguration from B to C. Both reconfigurations required the introduction of an along-track drift to reduce the mean along-track separation from 5km to 3.5km (A to B) and from 3.5km to 1.5km (B to C).

Fig. 1 Relative motion of Mango w.r.t. Tango projected onto the along-track/radial and cross-track/radial planes with origin on Tango. Flight data from reconfiguration A to B (cf. AFC2 plan from previous post) on March 18^th 2011.

For clarity the center of the instantaneous relative motion elliptic trajectory is superimposed on the figures below (cf. pink line in along-track/radial plane). The center coordinates are basically given by the relative semi-major axis (radial) and the relative mean longitude (along-track). The motion of the center gives insight into how the control scheme works. The along-track drift is first introduced and later removed through step-wise corrections of the relative semi-major axis. The relative eccentricity and inclination vectors are kept in a nominal anti-parallel configuration throughout the complete reconfiguration to minimize the collision risk at all times. In both occasions the reconfiguration is performed in ca. 3 orbital revolutions.

Fig. 2 Relative motion of Mango w.r.t. Tango projected onto the along-track/radial (top) and cross-track/radial (bottom) planes with origin on Tango. Flight data from reconfiguration B to C (cf. AFC2 plan from previous post) on March 19^th 2011.

The autonomous rendezvous phase of the AFC2 experiment slot has been completed on March 23^rd, 2011 when formation E has been finally acquired. Fig. 3 shows the relative motion of Mango w.r.t. to Tango which is characterized by a minimum separation of 200m (in radial direction) in this configuration. The obtained maximum control tracking errors are below 5m, 10m and 1m respectively in radial, along-track and cross-track directions.

Fig. 3 Relative motion of Mango w.r.t. Tango projected onto the along-track/radial (top) and cross-track/radial (bottom) planes with origin on Tango. Flight data from formation E (cf. AFC2 plan from previous post) on March 23^rd 2011 for a duration of 16 hours.

Written by

Simone D'Amico

2011-03-30 / 16:54:27

Smooth and promising start of the DLR’s AFC2 experiment slot

The final experiment slot (AFC2) of the Spaceborne Autonomous Formation Flying Experiment (SAFE) has officially started on Wednesday March 16th at 18:27:00 UTC time during the PRISMA orbit number 3948. This date marks a remarkable step in the PRISMA mission since AFC2 is the first experiment conducted under the supervision of GSOC as mission control center.

As part of the hectic activities which led to the handover of mission operations from SSC to GSOC on March 14th and 15th, the joint GSOC/SSC team delivered a healthy PRISMA formation at the desired initial conditions for a proper start of AFC2. Fig. 1 below shows the relative orbital elements describing the motion of Mango w.r.t. Tango during the first hours of the AFC2 experiment slot.

Fig. 1 Relative orbital elements estimated on-board on March 17th during the first hours of the AFC2 experiment slot (open-loop).

A comparison of the initial conditions with the desired values presented in the previous post (cf. Fig. 2 from previous post) shows a 25 m error on the x-components of the relative eccentricity and inclination vectors and a drift of ca. 30 m/rev in along-track direction. Such errors were considered acceptable during the so-called open-loop phase of AFC2 (i.e., first day of AFC2) and have been promptly compensated by the DLR’s GNC subsystem after the switch of mode from open-loop to closed-loop which took place on March 17th at 18:27:00 UTC, exactly 24 hours after the start of the experiment.

The behavior of the GNC software after the switch to closed-loop can be appreciated from Fig. 2 which depicts the relative orbital elements during the first formation acquisition phase. The comparison between actual (estimated on-board) and nominal (desired from ground) parameters shows a smooth and accurate reconfiguration phase which lasts a total of 1 orbit for the relative inclination vector and 3 orbits for the relative eccentricity vector and the mean along-track separation.

Fig. 2 Relative orbital elements estimated on-board on March 17^th after mode switch from open-loop to closed-loop.

A better understanding of the relative motion of Mango w.r.t. Tango during this phase is given by Fig. 3 which shows the projection of the relative position in the along-track/radial and cross-track radial planes. Tango occupies the origin of the axes, while Mango describes the relative trajectory depicted in black. The desired relative position is depicted in red and corresponds to the desired relative orbital elements of formation A (cf. previous post). The center of the relative ellipse is plotted in pink. The accumulated along-track offset of ca. 500 m is smoothly removed by the GNC system through the increase of the semi-major axis of Mango (ca. 20 m initially). The initially tilted shape of the relative ellipse in the cross-track/radial plane is properly corrected during the reconfiguration phase.

Fig. 3 Relative position of Mango w.r.t . Tango in along-track/radial and cross-track/radial directions.

Written by

Simone D'Amico

2011-03-18 / 19:55:44

Planning the DLR’s AFC2 experiment slot under GSOC supervision

The DLR’s contributions to PRISMA are motivated by the opportunity to demonstrate, as an experiment, advanced formation flying in a satellite testbed. In a general language, the Spaceborne Autonomous Formation Flying Experiment (SAFE) is described by the following opportunity statement, first formulated in 2005 during the Phase A of PRISMA

SAFE_OpportunityStatement

Fig. 1 Opportunity Statement for Spaceborne Autonomous Formation Flying Experiment (SAFE).

SAFE is intended to demonstrate autonomous, precise and robust formation control on a routine basis. This is made possible by an advanced guidance, navigation and control (GNC) subsystem developed by DLR and integrated by SSC into the spacecraft platform. Operationally, the SAFE demonstration has been split into two experiment slots named AFC1 and AFC2. As extensively reported in the mission blog, the AFC1 experiment slot has been successfully conducted in the time frame between September 20th and October 6th 2010 during the early PRISMA mission phase. As a conclusive step of the SAFE demonstration AFC2 is intended to demonstrate

Autonomous formation acquisition, keeping and reconfiguration on a broad spectrum of constellation geometries.
Fly-around and inspection of Tango vehicle to allow the collection of images using the Digital Video System (DVS) on-board Mango.
Proper behavior and performance of the latest DLR’s flight software uploaded to the spacecraft to solve minor navigation anomalies encountered throughout the mission (lessons learned).

As compared to the AFC1 experiment slot, AFC2 will push the GNC subsystem to the limits, demonstrating autonomous proximity operations capabilities at distances down to 20 m with a rotating Mango always pointing at the Tango spacecraft during the fly around. Furthermore the AFC2 experiment slot is the first experiment conducted in the PRISMA mission phase operated by the German Space Operations Center (GSOC).

The total duration of AFC2 is 19 days with official starting date on Wednesday 16-March. The validated plan of the AFC2 experiment slot for the first 8 days is shown below in Fig. 2. The GNC subsystem is asked to acquire, maintain and reconfigure the indicated formations in full autonomy. The desired relative motion of Mango w.r.t Tango is depicted in the orbital frame centered on Tango and aligned with the radial, along-track and cross-track directions. A pure anti-parallel relative eccentricity/inclination vector separation is applied to guarantee the safety of the formation during the early rendezvous phases and to guarantee good visibility conditions of Tango for the first DVS images to be taken on days 6 and 8.

Fig. 2 Validated plan for the first 8 days of AFC2 experiment slot.

Written by

Simone D'Amico

2011-03-18 / 18:54:57