69 Update Codebase to Perform Propagation in Gym Environment using NN controllers#70
Open
schmidt1139 wants to merge 18 commits into66-separate-nns-for-throttle-and-directionfrom
Conversation
This is a clean up commit that reformats some variables into dictionary keys and so on after the refactoring of the other nn training scripts. I neglected to do this in the previous branch.
Possible use cases include situations where no loss function is defined, the function should be able to handle this without crashing.
These need to basically match the Hamiltonian ODEs but without the co-states dictating the action. The action is an input array.
These can now be called interchangeably (with the appropriate input coords) the SC object with update the other set whenever one is called.
- Removed unneeded central body radius from ODE - Added g0 in km/s^2 for unit consistency - Updated EOM function handle - Removed comment - Updated final SC state update to use cartesian method
I added an arbitrary thrust profile thats burning constantly with a smoothly changing attitude. The env reset method uses a fixed seed for reproducibility.
Each of these methods involves testing stepping in the environment for one step using 1) no maneuvers, 2) an arbitrary maneuver, and 3) a maneuver queried from a neural network. They both product log files with numerical results from the step/propagation and check against truth files. Meant to ensure no unintended behavior happens with env or propagation.
MartinAstro
approved these changes
Oct 30, 2025
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pull request mainly contains work to update my code base to match the expected control paradigm of the trained NN controllers. The gym env code was a little stale in this regard, hadn't really been touched since before NN training and Hamiltonian control work.