Machine learning and evolutionary techniques in interplanetary trajectory design
This work addresses trajectory planning for interplanetary missions, but it is incremental as it extends prior findings from landing and quadcopter scenarios.
The paper tackled the problem of representing optimal guidance profiles for interplanetary missions using deep neural networks, with results demonstrating applicability to an Earth-Mars orbital transfer case.
After providing a brief historical overview on the synergies between artificial intelligence research, in the areas of evolutionary computations and machine learning, and the optimal design of interplanetary trajectories, we propose and study the use of deep artificial neural networks to represent, on-board, the optimal guidance profile of an interplanetary mission. The results, limited to the chosen test case of an Earth-Mars orbital transfer, extend the findings made previously for landing scenarios and quadcopter dynamics, opening a new research area in interplanetary trajectory planning.