ODTK Tracking System Design

You would like to use a simulator to verify the position uncertainty achievable by using different ground station network. This mission under consideration is a LEO radar for EO.

You have been provided an ephemeris file that covers the entire repeating ground track period (21 days) and a Tracking System object that is comprises a radar site (Kiruna) and two optical sites (Maspalomas and Perth). The ephemeris file has been created from STK by propagating the satellite's initial state with HPOP.

After the simulator run, a complete OD process can be performed in order to verify the performances of the tracking system.

Create a New Scenario

Let's start by creating a new scenario and defining it's properties.

  1. Create a new scenario () called TrackingSystemDesign.
  2. Double-click the Scenario to open the scenario's properties.
  3. Locate the Process field.
  4. Set the StartMode to Start Time.
  5. Set the Start Time to 31 Mar 2014 10:00:00.000 UTCG.
  6. Set the EndMode to End Time.
  7. Set the Stop Time to 21 Apr 2014 10:00:00.000 UTCG.

Set Up the Tracking System

  1. Select TrackingSystemDesign () in the Object Browser.
  2. Extend the File menu.
  3. Select the Import option.
  4. Select the Import Object option.
  5. Change the Files of Type to Tracking System Files (*.tso).
  6. Browse to the location of the tracking system database.
  7. Insert the GroundStationNetwork.tso tracking system (). (Typically, <install Directory>\AGI\ODTK 6\ODTK\UserData\TrackingStations)
  8. Click Open.

Model the Satellite

Let's take a moment and model the satellite in ODTK.

  1. Insert a new Satellite () named FeedSat. This is going to be the feeder satellite for generating dummy data using the simulator.
  2. Open the satellite's properties.
  3. Set the EstimateOrbit flag to false. You are using an external ephemeris file not ODTK, so set this option to False.
  4. Load the ephemeris file FeedSat.e from the install directory. (Typically, <install Directory>\AGI\ODTK 6\ODTK\UserData\Ephemeris)
  5. Set the CovarianceSource flag to none. You don't have covariance data at this stage.
  6. Click Apply.
  7. Leave the TrackingID to 1000. You can change this number later.

Since the ephemeris file under consideration has been generated using the default HPOP settings in STK, there is no need to change the force model settings in the ODTK (but in general this may be necessary to model the proper force model).

ODTK Simulator

ODTK includes a tracking data simulator that can produce a simulated measurement file for any type of measurement that can be processed in ODTK. The measurement file can be written in any of the measurement file formats already available in ODTK, and new formats can be added by writing a plugin.

The simulator allows you to add random deviations in the following ways:

Add a Simulator

In order to generate measurements using the finite maneuver that you just modeled, you need to define the specific measurement models that may be used for the estimation process.

  1. Insert a new Simulator () object and name it Simulator.
  2. Expand the SatelliteList.
  3. Click Add.
  4. Select FeedSat.
  5. Click OK.
  6. Ensure the Start and Stop times were inherited properly.
  7. Locate the ProcessControl field.
  8. Set the TimeStep to 0.5 min. This determines the frequency of measurements produced.
  9. Click Apply.

Set the Deviations

You can specify what deviations in the measurements. You can keep the same orbit, but add uncertainty, bias, noise, etc.

  1. Expand the ErrorModeling field.
  2. Set the DeviateOrbits flag to false.
  3. Set all other parameters to true, but keep NoDeviations set to false.
  4. Click Apply.

Generate a Simulator Schedule

In this scenario, you have a radar tracking station (KIR-1) and two telescopes (MSP-1 and PER-1). While the radar sensor can track the satellites in any lighting condition, the telescope needs to be in umbra (during the night) and the satellite to track needs to be illuminated by the Sun. ODTK has an HTML utility that automatically creates a valid observation schedule according to the tracking data type.

  1. Open the GenSimSchedule30.htm utility from the LaunchPad.
  2. Enable FeedSat as the Target.
  3. Select Radar for KIR-1.
  4. Set the Min Elev Angle to 10 for KIR-1.
  5. Select Optical Mission for MSP-1.
  6. Set the Min Elev Angle to 10 for MSP-1.
  7. Select Optical Mission for PER-1.
  8. Set the Min Elev Angle to 10 for PER-1.

Generate an STK Scenario

  1. Click the Setup STK Scenario to generate an STK scenario for access calculation.
  2. Click GO! to run the access calculation in STK. This action enables some access constraints (minimum elevation angle and Maximum Sun Elevation Angle for the optical sensors) before running STK.
  3. Locate the CustomTrackingIntervals field.
  4. Open the Schedule option.

Notice that there is a row for each ground station and the inclusion intervals have been filled out by the GenSimSchedule30 utility.

Run the Simulator

  1. Bring ODTK to the front.
  2. Run () the Simulator.
  3. Generate a Measurement Times by Type graph to check the simulated tracking data.

Create an Access Reports

  1. Use the newly created STK scenario to generate an access report.
  2. Verify the access times are similiar between ODTK and STK. The image below is just an example.

Insert the OD Satellite

  1. Insert a new satellite (). This is the object of our OD process. We assume here that the measurements data created by the simulator.
  2. Name the new satellite ODSat.

Change the Initial State

  1. Change the Initial State using the Initial State Tool from the Launchpad.
  2. Select ODSat as the Select Satellite option.
  3. Select the same ephemeris file (FeedSat.e) used to run the simulator.
  4. Press GO to get the initial state directly in the ODSat properties.

Change the Satellite ID

The measurements have been created with a satellite tracking ID equal to 1000. You need to be sure that this ID is just assigned to the ODSat object.

  1. Open FeedSat's () properties.
  2. Set the TrackingID to 1. You are disassociating FeedSat with the measurements.
  3. Click Apply.
  4. Open ODSat's () properties.
  5. Set the TrackingID to 1000.
  6. Run the Preview Measurements tool to check that ODSat is in the tracking data, as well as the three ground stations.

Data Filtering (Three Stations)

It's now time to perform the OD process by filtering the tracking measurements.

  1. Add a new filter () called Filter.
  2. Add ODSat to the SatelliteList.
  3. Locate the ProcessControl field.
  4. Ensure the Start Time is set to 31 Mar 2014 10:00:00.000 UTCG.
  5. Set the LastMeasurement as the Stop Mode.

Enable Smoother Data Creation

  1. Locate the Output section of Filter's properties.
  2. Enable the generation of smoother data.
  3. Run () the Filter.

Add a Smoother Object

  1. Add a new smoother () object.
  2. Use the *.filrun file from the filter as input.
  3. Run () the smoother and create the following reports:

The results are pretty good, with a maximum position uncertainty less than 90 meters in the in-track direction.

Data Thinning (Radar Only)

You'd like to investigate if removing one or more ground stations from the ground segment and still match the requirements. You can start by just keeping Kiruna (radar) as tracking station.

  1. Open the Filter () properties.
  2. Set the CustomDataEditing enabled to True.
  3. Double-click the schedule to edit it.
  4. Click Add.
  5. Set the Action to Thin. You want to thin the data to two minutes to filter out smaller passes.
  6. Modify the TrackerList parameter in order to accept only the tracking data coming from Kiruna station.
  7. Run the filter and smoother.
  8. Create a new Smoother Position Uncertainty report.

Of course, in this case you get a worse result with the result of the uncertainty level up to 150m in the in-track direction.

Data Filtering

The results from the filter and the smoother run look good for a situation where you have all three tracking stations generating data. The results are good also good at any time there is a satellite pass that meets the lighting and az/el constraint. In reality, the measurements you get are often much more sparse than this. And you might not be able to have both radar and optical data because of budget constraints, etc. The rest of this tutorial will walk you through how to run trades where you limit the data and determine if you can still achieve the uncertainty required by the mission.

Radar Only Filtering

You'd like to investigate if removing one or more ground station from the ground segment and still match the requirements. You can start by keeping Kiruna (radar) as the tracking system.

  1. Open the Filter () properties.
  2. Modify the TrackerList parameter in order to accept only the tracking data coming from the Kiruna station.
  3. Run () the filter and smoother.
  4. Create a new Smoother Position Uncertainty report.

Data Filtering (Telescopes Only)

Let's now keep the two optical stations (Maspalomas and Perth) and discard (Kiruna).

  1. Open the Filter properties.
  2. Modify the TrackerList parameter in order to accept only the tracking data coming from Maspalomas and Perth stations.
  3. Run the filter and smoother.
  4. Create a new Smoother Position Uncertainty report.

It is interesting to note that until the tracking data is available with short gaps (and this happens for almost the entire time span), the maximum uncertainty is pretty similiar to the radar case. When the number of tracking opportunities decreases (this being strictly related to the geometry of the system), the uncertainty goes up to 250 meters.

ODTK 6.5