Neural-Network-Based Dynamic Flight Envelop Protection Within Integrated Aircraft Control System
Rong Xie, Huanhuan He, Xiaobao Meng, Xueling Fan
Abstract
A sparse recurrent fuzzy neural network (SRFNN)-based dynamic flight envelope protection (FEP) method is proposed and integrated into an aircraft control system against possible failures. As a baseline of the aircraft control system, a fault-tolerant flight controller is used to compensate the uncertainties caused by failures, such as damaged wing. As an augmentation, the SRFNN-based dynamic FEP is employed to prevent damaged aircraft from breaking through the safe flight envelope bounds, which are updated adaptively with the help of the safety-critical information obtained by the SRFNN. Additionally, the performance of the fault-tolerant flight controller can be improved after the proposed dynamic FEP is integrated into the aircraft control system. The integrated aircraft control system not only can improve the performance of aircraft with some failures but also can prevent the damaged aircraft from getting out of control. Simulations are conducted to test and verify the effectiveness of the fault-tolerant flight controller and the SRFNN-based dynamic FEP. Simulation results show the efficacy of the proposed integrated aircraft control system.