Adaptive guidance systems for hypersonic reusable launch vehicles

This paper presents an adaptive guidance system approach applied to hypersonic reusable launch vehicles (RLV). After an effector failure, it is assumed that the inner-closed-loop system utilizes a reconfigurable control algorithm to recover nominal maneuvering capabilities to the extent possible. However, nominal performance will typically not be fully recovered for RLV, and the outer-loop guidance system must account for the degraded vehicle response. Two main approaches for the adaptive guidance system are presented. The first approach augments the existing production guidance system by adding adaptation capabilities. A case study shows that stability is maintained following a primary pitch effector failure. This is achieved by adapting gains in the guidance feedback loops. However, it is shown that the trajectory commands to the guidance loops must also be re-targeted in order to achieve a safe landing. The second approach employs an online optimal trajectory re-targeting algorithm. Here, the calculus of variations is used to generate a database of admissible neighboring extremals. This database is then encoded in an efficient manner to generate mappings between the current states and vehicle capabilities and the costates defining the admissible optimal trajectories. These mappings are interrogated online at regular intervals to obtain the optimal guidance commands. A proof-of-concept case study of this approach shows that the final landing conditions are achieved following a primary speed control effector failure.