Numerical Optimization Study of a Constrained Hypersonic Reentry Vehicle

TL;DR

This paper presents a numerical optimization approach for the trajectory planning of hypersonic reentry vehicles, effectively handling complex path constraints to maximize crossrange with high accuracy.

Contribution

It introduces a novel method for solving state-path constrained optimal control problems, accurately identifying constraint activation points and enforcing necessary conditions.

Findings

Method effectively locates constraint activation/deactivation points.

Achieves highly accurate solutions with minimal user intervention.

Analyzes various problem formulations for hypersonic vehicles.

Abstract

The trajectory optimization of the atmospheric entry of a reusable launch vehicle is studied. The objective is to maximize the crossrange of the vehicle subject to two control-inequality path constraints, two state-inequality path constraints, and one mixed state-and-control inequality path constraint. In order to determine the complex switching structure in the activity of the path constraints, a recently developed method for solving state-path constrained optimal control problems is used. This recently developed method is designed to algorithmically locate the points of activation and deactivation in the path constraints and partition the domain of the independent variable into subdomains based on these activation and deactivation points. Additionally, in a domain where a state-inequality path constraint is found to be active, the method algorithmically determines and enforces the additional necessary conditions that apply on the constrained arc. A multiple-domain formulation of Legendre-Gauss-Radau direct collocation is then employed to transcribe the optimal control problem into a large sparse nonlinear programming problem. Two studies are performed which analyze a variety of problem formulations of the hypersonic reusable launch vehicle. Key features of the constrained trajectories are presented, and the method used is shown to obtain highly accurate solutions with minimal user intervention.