Structured Adaptive Model Inversion Controller for Mars Atmospheric Flight

Abstract

Thenewvision formannedMars exploration presents challenges that are critically different frompast experiences with robotic missions. Safe precision landing requirements formannedmissionsmake it necessary to fly a controlled entry trajectory, rather than themore robust ballistic entry trajectory traditionally used for robotic missions. Large variations and uncertainties in Mars atmospheric properties also pose a significant challenge. Model reference adaptive control has not been previously used for Mars entry control, but it is an attractive candidate. It has an adaptation mechanism that reduces tracking errors in the presence of uncertain parameters such as atmospheric density and vehicle properties such as aerodynamic coefficients and inertias. This paper develops and implements two different model reference adaptive control schemes for the Mars ellipsled vehicle and evaluates them with non-real-time simulation for a complete Mars entry trajectory-tracking scenario from entry interface to parachute deployment. Performance metrics are introduced, and the relative benefits and disadvantages of each controller are discussed. Each controller is evaluated in terms of trajectory-tracking performance while in the presence of uncertainties in the atmospheric density, uncertainties in the aerodynamic coefficients, and uncertainties in the vehicle inertias. Results presented in the paper demonstrate that the learning rates for both controllers can be tuned to obtain acceptable tracking performance during an entire entry trajectory. Which of the two model reference adaptive controllers is most suitable for a particularMars entry application depends upon the tradeoff between level