Methods and tools for preliminary low thrust mission analysis

The aim of the present thesis is to develop new methods that are useful for a space mission analyst to design low thrust trajectories in the preliminary phases of a mission study, where the focus is more on exploring various concepts than on obtaining one optimal transfer. The tools cover three main axes: generating low thrust trajectories from scratch, improving existing low thrust trajectories and exploring large search spaces related to multiple gravity assist transfers. Stress is put on the computational efficiency of the tools. Transfer arcs are generated with shaped based approaches, which have the advantage of having the ability to reproduce close to optimal transfers satisfying time of flight constraints and varied boundary constraints without the need for propagation. This thesis presents a general framework for the development of shape-based approaches to low-thrust trajectory design. A novel shaping method, based on a three-dimensional description of the trajectory in spherical coordinates, is developed within this general framework. Both the exponential sinusoid and the inverse polynomial shaping are demonstrated to be particular two-dimensional cases of the spherical one. The pseudo-equinoctial shaping is revisited within the new framework, and the nonosculating nature of the pseudo-equinoctial elements is analysed. A two-step approach is introduced to solve the time of flight constraint, related to the design of low-thrust arcs with boundary constraints for both spherical and pseudo-equinoctial shaping. The solutions derived from the shaping approach are improved with a feedback linear-quadratic controller and compared against a direct collocation method based on finite elements in time. Theoretical results are given on the validity of the method and a theorem is derived on the criteria of optimality of the results. The shaping approaches and the combination of shaping and linear-quadratic controller are tested on four case studies: a mission to Mars, a mission to asteroid 1989ML, to comet Tempel-1 and to Neptune. The design of low thrust multiple gravity assist trajectories is tackled by an incremental pruning approach. The incremental pruning of reduced search spaces is performed for decoupled pairs of transfer legs, after which regions of the total search space are identified where all acceptable pairs can be linked together. The gravity assists are not powered therefore the trajectory is purely low thrust and the transfer arcs are modelled by shaping functions and improved with the linear quadratic controller. Such an approach can reduce the computational burden of finding a global optimum. Numerical examples are presented for LTMGA transfers from Earth to asteroid Apollo and to Jupiter.EThOS - Electronic Theses Online ServiceGBUnited Kingdo