Interplanetary Low Thrust Example

With the release of STK 11.5, Astrogator improved support for modeling low-thrust engines. There is a training tutorial on how to model a low-thrust satellite maneuver from LEO to GEO.  This article describes how to model low-thrust interplanetary missions, using an Earth-to-Mars scenario.

In the attached scenario, the transfer satellite was originally produced with a differential corrector targeting coefficients for a time-varying attitude profile in the finite maneuver. However, AGI found that this problem was simple enough to just use rough numbers for the coefficients for the time-varying profile. Some thought it was necessary to identify initial mass (fuel, dry) settings as well a maneuver duration that would all be consistent with the engine model and provide sufficient control authority to reach the goal. This was alleviated somewhat as this particular example is adapted from a textbook example. It gave us a decent idea of what Isp and thrust the engine should have as well as an initial guess for the maneuver duration.

For a more practical use case, you’d typically have some notion of what thrust/Isp parameters are available on your spacecraft as well as a notion of the duration of your thrust leg. With this, you might start with the thrust aligned with a particular direction throughout the maneuver, perhaps the velocity direction. Then you could use the resulting finite maneuver to seed the guess for the optimal finite maneuver.

Detailed description of how to build this scenario

The scenario was built following the general process of modeling the low-thrust maneuver as laid out in the Astrogator: Spiral to GEO using an Optimal Finite Maneuver tutorial. It was modified for Earth to Mars instead of LEO to GEO.

The first seven sections of this tutorial are about getting your first initial orbit guess. This is where you would set up your parameters for Earth-Mars, your engine, etc. It isn't getting you the final orbit, just close. The transfer satellite was produced with a differential corrector targeting coefficients for a time-varying attitude profile in the finite maneuver. The creators used basic numbers in the targeting sequence and let the differential corrector get the targeted orbit.

Before running the differential corrector, the initial state and engine models (the ISP, thrust, etc.) were designed so that they were consistent with a low-thrust model as well as the maneuver duration. 

If you haven't used the differential corrector before, you can learn the basics from the Astrogator: Hohmann Transfer (Target Sequence) tutorial. 

Sections 8-13 of the tutorial go through using the Optimal Finite Maneuver to get to the final desired orbit. For the solver, the creators used 120 nodes. You can go through an iterative process, slowing increasing the nodes to find a smooth solution for your problem. 

Then you need to set your boundaries (initial and final) for where the satellites should start and end. You also set the path boundaries for the thrust spherical angles (when not using unit vectors as controls in the Algorithm Options) on the trajectory.

Once you run the optimizer, it will produce a log file of data used in the solution. The log file for this initial run is included in the scenario folder, called Transfer_Maneuver_optMnLog.txt. 

If you follow this tutorial, look at the attached STK scenario, and refer to the Transfer_Maneuver_optMnLog.txt, you have all the information you need to rebuild this example.