Traditional flying qualities metrics use first-order descriptors of the 'classical' aircraft modes such as the phugoid, short-period, Dutch roll, etc. These modes are often difficult to distinguish from one another in modern aircraft whose dynamics are usually coupled, leaving designers to develop lower order equivalent systems as approximations to the real aircraft's behavior. Flying qualities metrics typically include limits on the modal parameters of the lower order equivalent system-constraints that are conducive to classical control techniques and pole placement strategies. On the other hand, modern optimal or robust control techniques directly address Multiple Input Multiple Output (MIMO) systems, with coupled dynamics, and they function well with higher-order measures of system behavior like state covariances or energy norms. This work presents a new method of approximating the traditional flying qualities requirements for piloted aircraft through a flight condition dependent recasting of these requirements as upper bounds on the state variances. A linear matrix inequality feasibility problem is developed to compute a control law which stabilizes the system for the given flight condition, while satisfying both the state variance and actuator constraints. The method is demonstrated through a horizontal tail sizing study of a transport aircraft, which shows that a much smaller tail than is conventional can satisfactorily fulfill the flying qualities constraints.
展开▼
