Design of reliable and reconfigurable auto-landing flight controller
Date of Issue2005
School of Electrical and Electronic Engineering
Part I of this report presents the methodology of reliable autolanding control design for aircraft. The nonlinear aircraft model is first linearized at various flight conditions under various fault conditions that covers the entire flight regime intended and all fault scenarios. Then a linear aircraft model with polytopic uncertainties in system matrices is used as the design model. The H2 performance measure is adopted for the design. A design methodology is proposed and applied to design a reliable autolanding control system. Simulation using a high fidelity nonlinear aircraft flight simulation system is used to validate the designed autolanding control system. The designed reliable autolanding control system is able to tolerate single aileron stuck fault of upto ±20° and single elevator stuck fault of upto about 9°, subject to various vertical and lateral wind disturbances and wind shear. The designed system is also able to tolerate upto 15% loss of control surface area at aileron during coordinated turn. Part II of this report presents a neural-aided controller that enhances the fault tolerant capabilities of a high performance fighter aircraft during the landing phase when subjected to severe winds and failures such as stuck control surfaces. The controller architecture uses a neural controller aiding an existing conventional controller using a feedback error learning mechanism. The neural controller employs a dynamic Radial Basis Function neural network called Extended Minimal Resource Allocating Network (EMRAN), which uses only on-line learning and does not need prior training. The information about actuator failures is not available to the controller for use in reconfiguration. It is also assumed that the aircraft control system does not use angle of attack and sideslip for purposes of feedback. The conventional controller is designed using a classical design approach to achieve the desired autonomous landing profile with tight touchdown dispersions called herein as the pillbox. This design is carried out for no failure conditions but with the aircraft being subjected to winds. The failure scenarios considered in this study are: (i) Single faults of either aileron or elevator ATTENTION: The Singapore Copyright Act applies to the use of this document. Nanyang Technological University Library iii stuck at certain deflections and (ii) Double fault cases where both the aileron and elevator are stuck at different deflections. Simulation studies indicate that the designed conventional controller has only a limited failure handling ability. However, neural controller augmentation considerably improves the ability to handle large faults and meet the strict touchdown dispersion requirements, thus enlarging the fault-tolerance envelope. The performance of these controllers is also compared to the Nonlinear Dynamic Inversion (NDI) controller and a high gain version of the baseline controller. A separately designed fault tolerant controller using Reliable H2 approach is also used as the baseline and it is shown that its performance is also improved by neural network augmentation. Finally parameter selection of the EMRAN learning algorithm using Genetic Algorithm based optimization is presented.
DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Control engineering