Agile autonomous guidance using spatial value functions

This paper describes an autonomous guidance system based on receding horizon (RH) optimization. The system is integrated around a spatial, state-dependent cost-to-go (SVF) function that is computed as an approximation to the value function associated with the optimal trajectory planning problem. The function captures the critical interaction between the vehicle dynamics and environment, thereby resulting in tighter coupling between planning and control. The consistency achieved between the RH optimization and the SVF enables a more rigorous implementation of the RH framework to autonomous vehicle guidance. The paper describes the overall approach along flight experimental results obtained in an Interactive Guidance and Control Laboratory.