OPTIMAL REFERENCE ORBIT TRACKING AROUND ASTEROIDS VIA PARTICLE SWARM OPTIMIZATION AND INVERSE DYNAMICS TECHNIQUE

Optimal proximity maneuvers are essential to obtain accurate space missions. Indeed, they are extremely important in perturbed environments, such as around asteroids, where the gravity field is very irregular because of the irregular shapes. For this same reason, gravity field estimation is also very challenging. This paper focuses on optimal control of proximity maneuvers obtained thanks to the combination of Particle Swarm Optimization, inverse dynamics technique and B-spline curves, used to approximate the trajectory. Furthermore, an Extreme Learning Machine-based algorithm is employed to estimate the gravity vector from the position vector in the asteroid body-fixed frame. This choice allows to achieve a good accuracy while decreasing at the same time the computational time that other precise methods, like the polyhedron model, would require. Finally, the proposed framework is applied to reference orbit tracking around asteroid 951 Gaspra.