Abstract—This contribution discusses a novel many-objective optimization algorithm that combines an ant colony optimization based decomposition approach with a massive parallelization framework. A rigorous numerical analysis on the impact of the two varying key factors of the here considered parallelization approach is presented. Those factors are the number of co-evaluated solution candidates within an individual ant colony algorithm and the number of individual ant colony algorithms itself. Aim of the presented method is to solve a many-objective application corresponding to the interplanetary space trajectory of the Cassini probe, launched by NASA in 1997. The provided numerical results indicate that comprehensive mission analysis via a many-objective approach is possible and that the presented approach is highly suitable for massive parallelization.
Index Terms—Many-objective optimization, ant colony optimization, space flight trajectory, parallelization.
Martin Schlueter, Chit Hong Yam, Takeshi Watanabe, and Akira Oyama are with the Japanese Aerospace Exploration Agency (JAXA), Institute of Space and Astronautical Science (ISAS) 3-1-1 Yoshinodai, Chuo-ku, Sagamihara, Kanagawa 252-5210, Japan (e-mail: email@example.com, firstname.lastname@example.org, email@example.com, firstname.lastname@example.org).
Cite: Martin Schlueter, Chit Hong Yam, Takeshi Watanabe, and Akira Oyama, "Parallelization Impact on Many-Objective Optimization for Space Trajectory Design," International Journal of Machine Learning and Computing vol.6, no. 1, pp. 9-14, 2016.