Optimal guidance for hypersonic vehicle using analytical solutions and an intelligent reversal strategy

Abstract

This paper presents a novel three-dimensional guidance law to quickly generate optimal commands in the gliding phase. Owing to the complicated trajectory constraints and highly uncertain disturbances, the novel gliding profile is developed for the longitudinal plane. Subsequently, this longitudinal profile is utilized to obtain analytical equations of motion for performance index and path constraints prediction. Moreover, stringent terminal constraints, which comprise the altitude, the flight path angle, and the position of transition from gliding phase to terminal diving phase, are integrated into the longitudinal profile. The proposed analytical expressions simplify the onboard generation of optimal gliding trajectory as it does not involve complicated nonlinear programming problems. In addition, to regulate the terminal velocity onboard and simultaneously eliminate the heading error, two bank-reversal opportunities are used in the lateral profile and determined using a reinforcement learning algorithm. In the proposed guidance law, the terminal deviations are first predicted in real-time using the analytical expressions; and then the three-dimensional trajectory is regenerated by addressing a simple optimization problem and planning bank-reversal positions. The computation of the guidance law is facilitated using analytical solutions and a reinforcement learning method; therefore, the frequency of replanning facilitates the robustness, accuracy, and optimality. We evaluated the effectiveness of our method using some representative scenarios.