Abstract
A highly maneuverable and reusable spaceplane, which is remarkable for its ability to send payloads and data back to Earth in the volatile environment, is attached with high priority in responsive reconnaissance missions. In view of the fact that a classical nonlinear optimization algorithm is mostly restricted in discovering the optimum solution of a multi-stage trajectory-design problem, a two-level optimization algorithm is established to enhance the convergence of the solution. By means of dividing the entire trajectory into ascending, on-orbit, deorbit, and reentry segments, respective optimizations of the four segments above compose the sublevels of the optimization scheme and are associated elaborately one after another by transmitting the terminal value of a former segment to its following segment. A genetic algorithm, which is employed in the top level of this scheme, will incorporate the intermediate solution and state quantities from sublevels as the fitness function of the GA and generate the new optimized value for the first segment in each sublevel during each iteration. This two-level iteration will be operated under given constraints and will eventually acquire the optimum trajectory with a maximum target coverage time. Furthermore, a Monte Carlo simulation is conducted to observe the robustness of the optimized solution.
