Policy Iteration Adaptive Dynamic Programming Based Control for Hypersonic Flight Vehicles

A novel approximate optimal tracking control scheme using policy iteration adaptive dynamic programming is proposed for hypersonic flight vehicles. By analysing the uniqueness of the equilibrium point during stable level flight of a hypersonic flight vehicle after achieving tracking objective, the output tracking problem is transformed into a state regulation problem. The initial admissible control law is obtained based on feedback linearization and sliding mode control under the discrete quadratic form cost function. Policy iteration adaptive dynamic programming algorithm is used to design the approximate optimal control law and the neural network approximation method is applied to approach the cost function. Simulation results illustrate the effectiveness of the proposed scheme.