Enhancement of the Computation Speed for the Satellite Conjunctions Screening by Combining Analytical Method and OpenMP

Autor: Thanathip Limna, Phasawee Saingyen, Keerati Puttasuwan, Sittiporn Channumsin, Suwat Sreesawet
Rok vydání: 2021
Předmět:
Zdroj: 2021 9th International Electrical Engineering Congress (iEECON).
DOI: 10.1109/ieecon51072.2021.9440363
Popis: In the current space era, the demand of satellites in various organizations is incredibly growing over the past several years because advance technology leads to the significant reduction of spacecraft manufacture and launching budgets. With the hypervelocity of space objects, it is not only risky to damage or destroy active satellites but also generates a huge number of new space objects. This scenario is known as "Kessler Syndrome". Therefore, Geo-Informatics and Space Technology Development Agency (GISTDA) has developed a space traffic management system called "ZIRCON" to monitor the possible conjunction events of Thailand Earth Observation Satellite (THEOS) that is owned and operated by GISTDA. ZIRCON is designed to automatically screen and provides conjunction analysis (Time Closest Approach (TCA), miss distance, and collision probability) to support operators to effectively decide an avoidance maneuver planning. The process consumption of previous ZIRCON is more than 10 hours for screening over 20,000 objects for 7 days in advance. As a result, the system can screen for one times/day that is insufficient to monitor and support other satellites. This paper presents a technical enhancement of the computational speed by developing a screening algorithm and combining parallel computing techniques. The developed algorithm utilizes the conjunction screening technique that combines analytic (Simplified General Perturbation: SGP4) and numerical propagation to enhance the computation speed and maintain the accuracy of propagation. Finally, this paper applies the parallel computation technique by using OpenMP. The development enables to distribute the computation of over 20,000 objects into available multiple processor cores on the single computer machine for parallel process. The results show that the technique can significantly reduce computation time from 10 hours to 1.30 hour.
Databáze: OpenAIRE