Nice recovery! Pilots get help saving uncontrollable aircraft

A pilot who loses control of an aircraft ordinarily has a couple of seconds to recover before system failure. But new technology may buy pilots extra time, saving lives in the process.

Within a few seconds of a pilot losing control, the plane leaves the normal flight envelope, a boundary that depicts for a specific aircraft the limits of speed, altitude and acceleration that the aircraft cannot safely exceed. Once the plane has left the flight envelope, it is nearly impossible for the pilot to recover the aircraft.

A novel adaptive flight controller, developed by University of Illinois researchers, enables the plane to stay within the flight envelope for a few additional seconds, giving the pilot precious time to regain control.

“The main feature that distinguishes our controllers from others is its predictable performance in the presence of unpredictable uncertainties, like turbulence and gust,” said Naira Hoyakimyan, a professor of mechanical science and engineering and member of Illinois’ Coordinated Science Laboratory. “Because it uses fast adaptation, it can be applied to various aircraft platforms without major redesign or retuning.”

In flights conducted by NASA on its Airstar testbed in June 2010, pilots confirmed the veracity of the L1 adaptive flight controller during post-stall scenarios. Using L1 control law, developed by Hoyakimyan and former postdoctoral student Chengyu Cao (now of the University of Connecticut), pilots were able to maintain a nose-high altitude for three to four seconds with bank angles not exceeding 20 degrees. Without the controller, the pilot could not hold the aircraft, which rolled and sliced, exceeding a 45-degree bank.

Irene Gregory, a senior research engineer at NASA Langley Research Center, believes L1 adaptive control theory is a breakthrough in the field because it is the first control system to allow for fast adaptation with guaranteed robustness and performance.

“It can predict transient behavior and adjust to it almost instantly,” Gregory said. “Moreover, L1 is able to adapt to large variations in system parameters and unmodeled dynamics, making it able to work on most of the aircraft in case of failures.”

In the past, adaptive control algorithms could only guarantee stability in steady-state operations and were unable to adapt quickly to unexpected circumstances. The novelty of L1 is that it uses fast adaptation instead of gain-scheduling, an expensive and time-consuming process that requires designing a different linear control for every piece of a flight trajectory. The process must be tailored to each aircraft for each flight route.

So far, NASA has only conducted test flights of subscale aircraft, but the L1 controller could also be used on commercial or military planes in the future. Additional tests will take place in September.

“The goal is to make flying safer on all aircraft,” Hoyakimyan said. “This is a big leap forward in that process.”
________________

Editor's note: FlightGlobal's blog by John Croft has links to actual videos of the flight tests posted by NASA.

Contact: Naira Hovakimyan, Department of Mechanical Science and Engineering, 217/244-1672.

Writer: Kim Gudeman, Coordinated Science Laboratory.

If you have any questions about the College of Engineering, or other story ideas, contact Rick Kubetz, Engineering Communications Office, 217/244-7716, editor.