We propose to investigate some important open problems in direct adaptive
control of systems with actuator, sensor and dynamics failures, with
applications to flight control systems. The existing results on direct
adaptive control of systems with failures are for systems with known
and linear dynamics and some uncertain failures. Those results have
serious deficiencies when the systems under control have nonlinearities,
uncertain dynamics, or time-varying parameters which are common in
flight control systems.
The first topic we propose to address is the design of
direct adaptive control schemes for linear
systems with unknown parameters as well as uncertain actuator, sensor
or dynamics failures. We plan to employ a model reference approach combined
with actuator, sensor or dynamics gain tuning. A key task is to solve the
parametrization issue in the presence of two sets of parameters appearing
bilinearly: one for the dynamic system and one for the actuators or
sensors. Another key issue is adaptive compensation for systems uncertainties
caused by actuator or sensor failures, with or without disturbance matching
conditions.
The second topic is the design of direct adaptive control schemes for linear
systems with unknown and time-varying parameters as well as
uncertain actuator, sensor or dynamics failures. Of primary interest is
control of systems with piecewise-linear (jumping) or rapidly varying
parameters, which can be the models of flight control systems at different
operation conditions. Control of systems with other parameter variations
relevant to flight control systems will also be investigated.
The third topic is the design of direct adaptive control schemes for nonlinear
flight systems with known or unknown parameters as well as
uncertain actuator, sensor or dynamics failures. Nonlinear control design
tools such as feedback linearization and backstepping will be employed.
Both state feedback and output feedback
control schemes will be developed.
Multi-stage actuator, sensor or dynamics failures will also be investigated.
The proposal describes a project for the research on direct adaptive
control of systems with actuator, sensor and dynamics failures, with
applications to aircraft flight control systems. This research has been
supported by the NASA since June 1999, under grant NCC-1342. Before our
research the existing results on direct adaptive control of systems with
failures were for systems with known and linear dynamics and some
uncertain failures. Those results have serious deficiencies when the systems
under control have nonlinearities, uncertain dynamics, or time-varying
parameters which are common in flight control systems.
Supported by the NASA grant NCC-1342, we first developed direct adaptive
state feedback state tracking control schemes for linear time-invariant
plants with unknown parameters and unknown actuator failures
characterized by the patterns that
some of the plant inputs are stuck at some fixed values (or around them)
which cannot be influenced by control action. Three different failure models
are considered: constants, parametrizable variations, and unparametrizable
variations. Conditions and controller structures for achieving state model
matching in the presence of actuator failures are derived. Adaptive laws are
designed for updating the controller parameters when both the plant
parameters and actuator failure parameters are unknown. Closed-loop stability
and asymptotic state tracking are ensured. Simulation results show that
desired system performance is achieved with the developed adaptive actuator
failure compensation control designs. We also developed direct adaptive
actuator failure compensation schemes with state feedback for output tracking
or with output feedback for output tracking, by deriving design conditions,
designing controller structures and adaptive laws, and
analyzing system performance. We applied our adaptive control schemes to a
linearized Boeing 737 dynamic model and studied the control system
performance under different uncertainties with different control schemes.
We are currently working on the design of direct adaptive control schemes
for linear systems with unknown parameters as well as uncertain actuator,
sensor or dynamics failures. We will develop adaptive control
schemes for nonminimum phase systems and multi-input multi-output
systems with actuator failures. We will study more applications to aircraft
flight control systems. We will investigate the robustness of adaptive
actuator failure compensation control schemes with respect to system
parametric, structural and environmental uncertainties. We will formulate and
address the sensor or dynamics failure compensation problems. We will extend
our adaptive designs to linear time-varying systems with
piecewise-constant (jumping) or rapidly varying parameters, which can be the
models of flight control systems at different operation conditions.
As the main effort of this new project, we will develop direct
adaptive control techniques for nonlinear dynamic systems with
uncertain actuator, sensor or dynamics failures. We will model nonlinear
aircraft flight dynamics, design adaptive state feedback controllers for
minimum phase nonlinear systems with actuator failures, design adaptive
output feedback controllers for minimum phase nonlinear systems with actuator
failures, design adaptive state feedback controllers for non-minimum phase
nonlinear systems with actuator failures, design adaptive output feedback
controllers for non-minimum phase nonlinear systems with actuator failures,
design adaptive control schemes for systems with sensor failures, design
adaptive control schemes for systems with dynamics failures, and implement
adaptive control schemes by experiments on test aircraft.
The overall Phase I objective will be to demonstrate the feasibility of delaying flow separation for a Boeing 747 aircraft model using synthetic jet actuators. This will entail development of a complete, closed-loop simulation that includes aircraft dynamics, actuator models, and control system. The specific technical objectives of the Phase I program are as follows:
This research is to develop novel adaptive control approaches and
design techniques which are capable of effectively handling large
structural and parametric uncertainties caused by system failures and
damages, and to apply them to solve open aircraft flight control problems.
To develop a systematic reconfigurable adaptive control theory for
aircraft control applications, research activities will be carried out
in six synergic areas: flight system modeling for reconfigurable
control; critical performance metrics for reconfigurable adaptive
control; adaptive compensation of uncertain actuator failures;
adaptive aircraft flight control in the presence of system damages;
adaptive compensation of sensor uncertainties and failures; and
adaptive control of multivariable nonlinear systems.
Research tasks will be fulfilled using new system and control
methods: structural uncertainty modeling; metric parameter adaptation;
controller structural expansion; augmented system parametrization;
feedback based compensation; characterization and estimation of system
infinity zero structures; and adaptation of input-output interactions.
Expected theoretical advances and technical deliverables are:
new benchmark aircraft dynamic models taking into account system
failures and damages; analytical stability margins relevant to
adaptive control systems; new characterizations of multivariable
nonlinear systems (infinity zero structures, redundancy, and failure
compensability); new adaptive control schemes for multivariable
nonlinear systems with desired stability and tracking performance
and novel applications to aircraft flight control; new adaptive
controller structures and tuning algorithms suitable for compensation
of actuator failures, dynamics failures, damages, sensor failures, or
multiple or mixed failures; stability and robustness analysis of
adaptive failure compensation control systems; unified reconfigurable
adaptive control theory of relevance to aircraft flight control;
complete design, analysis and evaluation of aircraft flight
control systems with adaptive compensation of rudder, stabilizer,
engine, aileron or elevator failures, and wing or fuselage damages;
and systematic guidelines for designing control systems with
guaranteed stability and tracking performance in the presence of
system parameter and failure uncertainties.