Optimization of interplanetary transfers with flybys and deep-space maneuvers

Interplanetary scientific missions pose several challenges for trajectory analysis because a direct transfer is often not possible due to the high delta-v. On the other hand, flybys and deep-space maneuvers allow a reduction of this delta-v at the cost of increased complexity in the trajectory. A well-known example of such an ambitious mission is the Cassini-Huygens robotic spacecraft. The study here presented focuses on the optimization of high-thrust interplanetary transfers with multiple flybys and deep-space maneuvers. A direct optimization method including black boxes is used to minimize the total delta-v in the presence of mission constraints. The complexity of these trajectories poses several challenges to the optimization process, due to the overall non-convexity and the significant number of variables and constraints involved. As finding the global minimum represents a very challenging task, an optimization process consisting of several successive steps is developed. In the first optimization phase, the two-body-patched-conics model is used (See e.g., [1]). This allows to use a local optimizer (SQP) recursively to find many solutions in a reasonable time, in the perspective of obtaining the global optimum. For particularly complex problems, with longer flyby sequences and many constraints, an evolutionary algorithm is used at first to identify good guess solutions and to tighten the bounds of variables. An ad-hoc heuristic approach has been developed for the second phase to explore alternative solutions more efficiently compared to a multipurpose “multi-start” algorithm, allowing to reduce the number of starting points. Eventually, the major results are presented to demonstrate the validity of this method in comparison with previous missions. Furthermore, a method has been developed to solve the Boundary Value Problem in the case of perturbed dynamics, with the aim of extending the approach to more detailed analyses in the future. The method, implemented in the Python environment, allows the addition of constraints with flexibility, guiding the user from problem definition to the optimal trajectory. [1] M. R. Sentinella and L. Casalino. “Hybrid Evolutionary Algorithm for the Optimization of Interplanetary Trajectories”. In: Journal of Spacecraft and Rockets 46.2 (2009), pp. 365–372.