Knowledge discovery in conceptual design of space planes using multi-objective optimization and data mining

In this paper, multi-objective, multidisciplinary design optimization is conducted on the first stage vehicle of a horizontal takeoff and landing two-stage-to-orbit (TSTO) space plane with rocket-based combined cycle (RBCC) engines. A design optimization framework considering the interdependence among analysis disciplines (i.e., flight trajectories, propulsion, vehicle geometries, and aerodynamics) is constructed, and the vehicle design and the trajectory prior to the separation of the second stage vehicle are optimized simultaneously aiming to maximize the final velocity, minimize the gross mass of the first stage vehicle, and minimize the takeoff velocity. By solving this multi-objective optimization problem using a newly developed method, a set of pareto optimal solutions with a good spread is obtained. In addition to the discussions on some representative solutions, knowledge discovery in all the obtained solutions is performed via data mining techniques.