Near-space vehicle trajectory optimization based on control vector parameterization

For online trajectory optimization of the near-space vehicle in computational guidance, a trajectory optimization method based on actual dynamic equations is proposed via control vector parameterization. Firstly, a trajectory optimization problem with initial and terminal state constraints is developed by discretizing the nonlinear dynamic equations using the flipped-Radau pseudospectral approach. Secondly, control vector parameterization reduces complexity and boosts computational efficiency by converting the optimization problem for control quantities at each discrete point into an optimization problem for multiple parameter vectors. Thirdly, to deal with parameter amplitude constraints, a trajectory optimization method based on convex programming techniques is proposed via control vector parameterization. Finally, given that the midcourse guidance mission is insensitive to terminal flight time, the optimization problem is transformed from being time-dependent to downrange-dependent, thereby further reducing the computational time. Simulation results demonstrate that the proposed method satisfies the high-efficiency and high-precision requirements for trajectory optimization in computational guidance.