Multiple Gravity Assist Interplanetary Trajectories PDF Download

Interplanetary travel is a difficult task due to the high fuel mass required to reach other planets. Minimizing the cost of the manoeuvres (and, in turn, the fuel mass) is the objective preliminary mission design. During this phase, a large number of potential solutions must be evaluated quickly in search of feasible trajectories. This means computationally fast but simple models are preferred over accurate but slow models. Additionally, the process demands for an automatic execution due to the vast amount of solutions that must be evaluated. One of the major improvements regarding space travel was the discovery of the gravity assist, where a spacecraft uses the gravitational pull of a flyby planet to change its velocity with respect to the Sun. This allows reducing the amount of fuel mass, which in turn increases the science payload available for the mission. This thesis deals with the optimization of interplanetary trajectories with gravity assist. From an engineering approach, the thesis aims at producing an automatic optimizer of interplanetary trajectories with gravity assist manoeuvres aimed to preliminary mission design applications. From a scientific approach, the thesis aims at identifying key issues in the literature that allow for improvement and presenting novel implementations. Finally, the thesis has a strong educational component: the code and tools are specially focused towards an easy understanding and analysis of the underlying methods rather than producing a computationally efficient code. The result from this work is an automatic optimizer of multi gravity-assist interplanetary trajectories. The tool is fully modular and works with a double-loop approach: an outer loop obtains feasible sequences of planets using the Tisserand graph and an inner loop finds the best trajectory for each sequence using a hybrid heuristic optimizer and a patched conics method. Five key issues have been investigated and improved upon during the thesis: we provide an improved solution method for the Kepler equation, we have conducted an extensive bibliographic research of Lambert's problem and analyzed the representative methods to select the best for our application, we have recovered and improved the Lambert's problem method by Simó , we present two different models for the patched conics method, we have developed an automatic method to traverse the Tisserand graph and finally we have implemented several heuristic optimization methods and coupled them with an islands model. The resulting tool has already proved to work in operational mission design scenarios. However, it lends itself to many improvements and upgrades, in particular increasing the level of automation, improving the physical model and the patched conics method robustness, improving the visualization capabilities during the optimization stage and translating the code into compiled language to increase the computational performance with complex missions and intensive simulations.