Regression techniques for aircraft parameter identification from noisy measurements in maneuvering flight

System identification for reconfigurable flight control is proposed. System identification using using linear regression where the highest derivatives are measured is analyzed. Knowledge of the stability and control derivatives is incorporated into the linear regression algorithm through ridge regression and mixed estimation methods. The example problem is a linearized fourth-order model of the F-16 pitch channel, and includes appropriate measurement (sensor) noise. The results show that even during periods of low excitation, the short period parameter estimates are improved by utilizing prior Phugoid information