aleksandarpetrov / traj-opt-paper

Optimizing multi-rendezvous spacecraft trajectories: ∆V matrices and sequence selection

Code for the experiments.

Installation

This code uses ephemeris data and Lambert targeting from pyKEP so pyKEP needs to be installed. Also required are: pygmo pathos and pyYAML:

pip3 install pygmo pathos pyYAML pykep

A Docker container with everything neccessary installed is also provided at aleksandarpetrov/traj-opt.

Getting Started

Running an experiment consists of:

1. Generating ∆V matrices with the batch_Lambert_sampling.py script, e.g.:

python3 batch_Lambert_sampling.py --dt 80 --njobs 8 --output data/DV_20obj_dt80_raw --objects 20

2. Applying wait-adjustment to the generated matrices with the wait_adjusting.py script, e.g.:

python3 wait_adjusting.py --njobs 8 --input data/DV_20obj_dt80_raw --output data/DV_20obj_dt80_wa

3. Computing sequences of 5 objects via direct concatenation with the main.py script, e.g.:

python3 main.py --njobs 8 --input data/DV_20obj_dt80_wa --output data/DV_20obj_dt80_dced --dvtol 10000 --objects 20 --results data/20obj_dt80_results.yml --inmemory true

All this can be run in the provided Docker container by using the provided run_all.sh as a template. Note that you'd need to have Docker installed to be able to use it.

Authors

• Aleksandar Petrov

License

Contact the authors