Handling Quality Oriented Fault Tolerant Longitudinal Flight Control Algorithm Design For A Jet Trainer Aircraft

Abstract

In the scope of the thesis an alternative longitudinal flight control algorithm, which satisfies predefined stability and performance requirements, in case of a specific fault is aimed for a jet trainer aircraft. Base flight control algorithm, which is being used currently, requires 〖CG〗_X position information to schedule controller parameters to achieve predefined stability and performance goals. In case of fault in 〖CG〗_X position information, where current 〖CG〗_X position information is not available or has low fidelity, flight control algorithm is automatically being fed with mid 〖CG〗_X position parameters. According to current 〖CG〗_X position, especially when current 〖CG〗_X position closes to edge of 〖CG〗_X range, there are significant stability and performance degradations in longitudinal flight control characteristics of the aircraft. An alternative longitudinal flight control algorithm which does not require 〖CG〗_X position information to sustain nominal stability and performance is the topic of the thesis. Six of degree of freedom nonlinear aircraft model is constructed with sub-blocks which represents different systems, to represent jet trainer aircraft model completely. This nonlinear model is trimmed around a specific design point and linear time invariant models are generated with small perturbations around this design point, to be used in flight control algorithm design processes. Flight control algorithm design requirements in terms of stability and performance are explained in detail to guide design process and asses the final controllers. Detailed design process of base controller and two alternate controllers are explained and applied for a specific design point. Base controller and the first alternative controllers are based on same architecture, PI with feedforward elements, but with different design methodologies. Base controller parameters are calculated via pole – zero assignment. Parameter space approach methodology for SAS and structured H_∞ synthesis methodology for CAS design are followed in the first alternative controller . The second alternative controller is based on explicit model following architecture with disturbance rejection capability. SAS feedback gains are calculated via parameter space approach methodology and structured H_∞ synthesis is used to design disturbance rejection compensator. Base controller and two alternative controllers are assessed in terms of predefined stability and performance requirements. The second alternative controller, explicit model following with disturbance rejection, has been found as the best one. It satisfies all stability and performance requirements even tough in case of fault in 〖CG〗_X position information. It has been observed that exact reference model match has been achieved with the second alternative controller with significant disturbance ejection capability. Stability has been ensured with parameter space approach which gives a 2D feedback gains basis which guarantees predefined stability requirements in Nyquist diagram. On the other hand, the first alternative controller has showed that even with a different design technique the architecture, PI with feedforward elements, is weak in terms of disturbance rejection. The main idea in the second alternative controller architecture is that; changes in internal aircraft dynamics and uncertainties can be assumed as disturbance and exact nominal performance can be reached with a sufficient rejection of those disturbances.