Rapid Trajectory Optimization For Flight Vehicles With Complex Constraints

This thesis aims to optimize the reentry trajectory for a new style of glide-reentry vehicles in the phase of scheme design. In the reentry process of such vehicles, the range of altitude and velocity is very large and the strict constraints in heating protection and overload, as well as the high sensitivity to the control variables for the hypersonic reentry trajectory bring the challenge of this research. Rapid trajectory optimization can handle emergency easily and therefore is of great research value. The main work of this thesis can be described as follows:At first, longitudinal point-mass dynamics of a flight vehicle are derived based on some assumptions. Then, lift and drag models are built according to the data of some Common Aero Vehicle. After that, we analyze the constraints of overload and heating rate which must be considered in the reentry process. At last, an optimal control problem is formulated to maximize the flight range based on a dimensionless model description.In the second place, we discuss the similarity and difference of several pseudospectral methods based on Legendre polynomials. According to the practical problem we are trying to solve, we focus on the principles of Legendre pseudospectral method and how to solve an optimal control problem using this method.Furthermore, the trajectory optimization problem we obtained before is transformed to a normalized nonlinear programming problem through the Legendre pseudospectral method. We use MATLAB as a powerful tool to solve the constrained nonlinear programming problem, which applies the Sequential Quadratic Programming when solving medium-scale problems. Simulations demonstrate the effectiveness of this algorithm. Some preliminary analysis is presented too.Finally, we optimize the multiple-stage reentry trajectory in order to increase the accuracy while not to increase unnecessary computational burden. The original trajectory is segmented at some points with specific energy. By comparing trajectories without segments, trajectories with two segments and trajectories with three segments, conclusion can be drawn that multiple-stage optimization can increase the flight range with high computation efficiency and precision.