MICHAEL PHELAN / LONDON
"Self-learning" software could enable damaged aircraft to be flown to a safe landing
NASA has completed evaluation flights of a revolutionary "self-learning" flight control system that could enable damaged aircraft to be flown to a safe, controlled landing.
The Intelligent Flight Control System (IFCS) was tested on a modified NASA Boeing F-15B at NASA Dryden, California. It focuses on developing neural network software to compensate for the altered response of a damaged aircraft.
In its final form the IFCS would compare real-time operational data from the damaged aircraft and its systems with a database of how it would normally operate, and automatically adjust the flight controls to compensate for damaged or inoperative control surfaces and systems.
The IFCS project team evaluated in flight a passive online parameter identification (PID) algorithm - or software code - and an online learning dynamic cell structure (DCS) neural network algorithm. The PID and DCS algorithms were tested for their ability to identify aircraft stability and control characteristics, and map and retain this data as a function of flight condition.
"This work marks a significant step toward learning, thinking, aircraft that will be safer, more autonomous, and more reliable than ever before," says John Carter, Dryden's IFCS project manager.
The PID algorithm is an online function that determines the actual stability and control characteristics of the aircraft as it flies. When results from the PID algorithm differ from the "pre-trained neural network" (PTNN), an update to the system is required. The DCS tracks the differences between the PTNN and PID and provides a map of updates to improve stability and control. It has a long-term memory - critical for IFCS use - and the network can be enlarged by adding nodes.
IFCS software evaluations performed using the F-15B included handling qualities and envelope boundary manoeuvres, and control surface excitations for real-time PID.
Source: Flight International
