Abstract: By designing a speed and attitude control system based on adaptive neural network compensation，this paper presents a solution of solving the cruise tracking control problem for the stratospheric airship. An adaptive RBFNN (Radial Basis Function Neural Network ) is used to compensate modeling errors, which come from the approximate model applied to a regular linear controller design. The network weight adaptation rule, derived from Lyapunov stability analysis, guarantees that the adapted weight errors and the tracking errors are bounded. The controller design exploits only a little information of model parameters. Simulation results demonstrate the excellent performance and robustness of the controller, even if environmental winds with unknown information exist.

Key words: stratospheric airship, cruise tracking, neural network, adaptive control