Refining the Shuttle training aircraft controller

NASA uses the Shuttle training aircraft (STA) to train the Shuttle pilot astronauts. Although as compared to the Shuttle, STA is a small aircraft, it does simulate, with some model following errors, the aerodynamics of the Shuttle. In this paper, we show how these differences can be reduced significantly by using a new pitch predict controller and a new adaptive fuzzy controller. In particular, we show how the collaborative work between NASA Ames and the Shuttle Training Branch of Johnson Space Center has improved the accuracy of STA model following from its current level of 71% to 91% which is equivalent to a 69% error reduction.