tle_conjunctions

Kerry N. Wood ([email protected])

A Numpy-vectorized version of the Healy Algorithm

Healy's ephemerides-based, first-order conjunction detection algorithm using SGP4 (thanks Brandon Rhodes) and Numpy.

The algorithm does a first-order approximation of miss-distance at each ephemeris step. So, runtime is sensitive to step-size. For LEO orbits ~5 minutes is sufficient. For GEO ~15 is usually sufficient. Use the minimum of the two.

This implementation takes two catalogs, a start time, end time, and step. Note that it does no conjunction "refining" when a miss is identified.

This implementation also naively generates ephemeris sets: it will generate time-aligned ephemeris for all objects over the specified window. IOW split runs into shorter time periods so that you don't have huge ephemeris sets.

Both ephemeris generation and conjunction detection is multi-core

To-do

• allow for externally generated ephemerides
• refine conjunctions; when a miss-distance threshold is identified, refine by propagating within the window

Original paper:

@article{healy1995close,
	title={Close conjunction detection on parallel computer},
	author={Healy, Liam M},
	journal={Journal of Guidance, Control, and Dynamics},
	volume={18},
	number={4},
	pages={824--829},
	year={1995}
}