Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM68427
Volume:
10 |
10 |
Oct 2026
Pages:
1-4
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
Department Of Computer Technology
Priyadarshini College Of Engineering, Nagpur
Department Of Computer Technology
Priyadarshini College Of Engineering, Nagpur
Department Of Computer Technology
Priyadarshini College Of Engineering, Nagpur
Department Of Computer Technology
Priyadarshini College Of Engineering, Nagpur
Department Of Computer Technology
Priyadarshini College Of Engineering, Nagpur
Department Of Computer Technology
Priyadarshini College Of Engineering, Nagpur
Low Earth Orbit (LEO) is the most congested region of the near-Earth environment, with tens of thousands of catalogued objects and far more untracked fragments sharing a shell only a few hundred kilometres thick. Sustained operation in this regime demands conjunction assessment that is both physically faithful and computationally tractable. This paper presents a fully functional Python-based system that ingests NORAD Two-Line Element (TLE) sets, propagates them through a vectorized SGP4 implementation, screens all pairwise conjunctions in two stages, classifies encounter risk with injected thousands of trackable fragments into the 700–900 km band, and large broadband constellations have increased conjunction alert volume by an order of magnitude. Conjunction assessment reduces, in principle, to a simple question: for every pair of objects, what is the minimum separation over a look-ahead window? The naive formulation is brute force. For N objects propagated across T time steps, evaluating every pair at every step costs a supervised learning model, and renders the resulting C = N(N−1)/2 · T ∈ O(N²T) (1) geometry as an interactive three-dimensional scene in a web browser. Propagation is delegated to a batched sgp4_array call over a shared split Julian-date grid, yielding TEME position and velocity histories for the whole catalogue in one vectorized O(N·T) pass. Conjunction detection uses a coarse 60-second sweep to shortlist candidate pairs, followed by a 2-second refinement inside a ±300-second window to recover the Time of Closest Approach (TCA) and miss distance, reducing fine-grid work from O(NZT) to O(KT′) with K ≪ NZ. A Random Forest of 150 trees maps a six-dimensional orbital-geometry feature vector onto three risk classes, yielding a soft decision boundary in place of a single distance cutoff. In a 24-object, 24-hour demonstration the system detected three CRITICAL and three WARNING conjunctions, with the classifier reaching 99.83% held-out accuracy.
KEYWORDS
This paper presented an end-to-end conjunction assessment system built entirely from open-source components. Its central contribution is the demonstration that a two-stage temporal decomposition — coarse screening to bound the search, fine refinement to resolve the answer — reduces distance-evaluation cost by roughly thirtyfold at demonstration scale while preserving second-level TCA accuracy, making meaningful conjunction analysis achievable on commodity hardware. The vectorized propagation layer, the multi-feature learned risk boundary and the browser-native 3D visualization form a coherent pipeline from raw TLE text to actionable, inspectable alerts. The system’s limitations are documented candidly, and the paths toward operational credibility — continuously refreshed elements via the authenticated Space-Track API, Monte Carlo collision probability in place of deterministic distance, a delta-V recommendation engine, and validated labels for the learned boundary — are clearly identified as future work.
[1] D. A. Vallado, P. Crawford, R. Hujsak, and T. S. Kelso, “Revisiting Spacetrack Report #3,” in Proc. AIAA/AAS Astrodynamics Specialist Conf., Keystone, CO, USA, 2006, AIAA 2006-6753. [2] F. R. Hoots and R. L. Roehrich, “Models for Propagation of NORAD Element Sets,” Spacetrack Report No. 3, Aerospace Defense Command, Peterson AFB, CO, USA, Dec. 1980. [3] D. J. Kessler and B. G. Cour-Palais, “Collision Frequency of Artificial Satellites: The Creation of a Debris Belt,” J. Geophys. Res., vol. 83, no. A6, pp. 2637–2646, Jun. 1978. [4] ESA Space Debris Office, “ESA’s Annual Space Environment Report,” European Space Agency, Darmstadt, Germany, 2025. [5] T. S. Kelso, “CelesTrak.” [Online]. Available: https://celestrak.org [6] S. Alfano, “A Numerical Implementation of Spherical Object Collision Probability,” J. Astronaut. Sci., vol. 53, no. 1, pp. 103–109, 2005. [7] J. L. Foster and H. S. Estes, “A Parametric Analysis of Orbital Debris Collision Probability and Maneuver Rate for Space Vehicles,” NASA JSC-25898, Aug. 1992. [8] L. Breiman, “Random Forests,” Mach. Learn., vol. 45, no. 1, pp. 5–32, 2001. [9] F. Pedregosa et al., “Scikit-learn: Machine Learning in Python,” J. Mach. Learn. Res., vol. 12, pp. 2825–2830, 2011. [10] B. Rhodes, “sgp4: Track Earth Satellites Given TLE Data,” Python Package Index. [Online]. Available: https://pypi.org/project/sgp4/
Publication Date
10/28/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
1
Total views since publication
PDF Downloads
4
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...