This research effort aims to develop software control algorithms that will correct for roll reversal before it happens. Roll reversal occurs when an aircraft is steered in one direction but rolls the opposite way due to aerodynamic conditions. The problem often compounds as a pilot attempts to correct for the motion by over-steering in the original direction, leading to uncontrollable roll. Unexpected yaw and subsequent roll reversal caused the loss of an unmanned hypersonic technology vehicle (HTV-2) overseen by DARPA in 2011. The team has employed novel predictive software within adaptive controller technology to detect conditions likely to result in aircraft roll reversal and then automate compensating maneuvers to avoid catastrophic loss.
Work to date: University of Michigan’s retrospective cost model refinement (RCMR) control algorithm has been integrated into a flight simulator and tested with prerecorded, open-source parameter data similar to that from the HTV-2 anomaly. The simulation returned control recommendations that would have prevented the incident, suggesting that RCMR has the potential for preventing future roll-reversal anomaly.
Looking ahead: Next steps involve upgrading the RCMR code to account for a 6-degree-of-simulation environment (forward/back, up/down, left/right, pitch, yaw, and roll) with eventual application in a flight test environment. Partners will include the University of Michigan, DARPA, and aerospace firms.
|Organizations Performing Work||Role||Type||Location|
|Dryden Flight Research Center (DFRC)||Lead Organization||NASA Center||Edwards, CA|
|Defense Advanced Research Projects Agency (DARPA)||U.S. Government|
|University of Michigan||Academic||Ann Arbor, MI|
A final report document may be available for this project. If you would like to request it, please contact us.