Application of neural network technology to fly-by-light control systems,

摘要

Abstract: McDonnell Douglas Aerospace (MDA) is developing a Neural Network based Intelligent Flight Control System that utilizes fly-by-light data transmission combined with high capacity flight processors to implement control advancements and fault monitoring processes. In this control system design, high speed data transfer and electromagnetic interference protection obtained through fiber optic technology is linked with Neural Network based flight control hardware processors that are programmed with damage adaptive control capability, thus providing maximum survivability for fighter aircraft. This system also provides enhanced component fault diagnostics that can identify subsystem failures during flight, thus providing reduced life cycle cost through efficient maintenance action and less downtime of the aircraft. The Intelligent Flight Control products apply to fighter and transport aircraft. This program is Task 2C of the ARPA Fly-by-Light Advanced Systems Hardware (FLASH) Technology Reinvestment Program. The principal partner with MDA for the Task 2C Intelligent Flight Control development is Martin Marietta Control Systems. The program will mature the system hardware and software for laboratory demonstrations of component fault diagnostics and highly adaptive flight control performance.!0