Dynamic Inversion Flight Control Law Design for Fixed-Wing Aircraft

Abstract

This thesis describes the design, implementation and flight testing of flight control laws based on Incremental Nonlinear Dynamic Inversion (INDI). The method compares commanded and measured accelerations to compute increments on the current control deflections. This results in highly robust control solutions with respect to model uncertainties as well as changes in aircraft dynamic characteristics of failure cases during flight. At the same time, the complexity of the algorithms is similar to classical ones. The key for practical implementation is in ensuring synchronization between angular acceleration and control deflection measurements or estimates. The underlying theory and practical design methods of INDI are very well understood, but implementation and testing has remained limited to sub-scale UAVs. The main contributions of this thesis are: 1) the design and validation of manual attitude control functions for a Cessna Citation II experimental aircraft, covering control structure design, application of INDI, design optimization, robustness analyses, software implementation, ground and flight testing; 2) a novel method based on the complimentary filtering technique to obtain more accurate angular acceleration estimates from angular rate measurements; 3) mathematical proof that the inversion error of INDI due to neglecting the so-called system dynamics increment increases with the combined actuator, sensor and sampling delay. The flight tests were highly successful and marked the first successful demonstration of INDI on a CS-25 certified aircraft. The flight test results proved that INDI clearly outperforms "classical" NDI and provided valuable lessons-learnt for future applications.