Gain scheduling for H∞ control of propulsion controlled aircraft

A novel design algorithm is proposed for the pitch control of propulsion controlled aircraft (PCA) over a flight envelope. The PCA system is adopted for emergency flight control of airplane with control surface failure. The linear controllers are synthesized at various equilibrium operating points by an H_∞ method to solve the model matching problem. The radial basis networks (RBN) are employed to train the gain scheduling controller and flight parameters are directly selected as input vectors to the neural networks. Simulations results demonstrate that the scheduled H_∞ controller can stabilize the crippled airplane to the desired model and possesses the robustness against the engine delay, input disturbance and sensor noise