The static system identification scheme for reconfigurable flight control is modified to encompass overactuated aircraft with nonlinear aerodynamics. The first step is to obtain a reliable estimate of the critical equivalent loop gain and stability derivative in the respective pitch, yaw, and roll control channels. Short data windows are used so that the system identification algorithm operates in “real-time” and reconfigurable control is feasible. Moreover, excitation enhancing control signals are not used in this step. The next step is to estimate the effector that caused the changes in loop gain. A de-correlating excitation is injected into the null space of the effecters through the control allocation. This allows the individual control derivatives to be estimated. In addition, the adverse effects of nonlinearities on the identification performance of the control and stability derivatives, are investigated. Results are presented for a tailless fighter aircraft simulation.
[1]
Meir Pachter,et al.
Universal linearization concept for extended Kalman filters
,
1993
.
[2]
Wright-Patterson Afb,et al.
Integration of On-line System Identification and Optimization-based Control Allocation*
,
1998
.
[3]
M. Pachter,et al.
On‐line system identification for aircraft with distributed control effectors
,
1999
.
[4]
Meir Pachter,et al.
System identification for adaptive and reconfigurable control
,
1995
.
[5]
Wright-Patterson Afb,et al.
Regularization Techniques for Real-Time Identification of Aircraft Parameters
,
1997
.