Abstract: In view of the attitude control problem of the reentry roll and yaw channels of the aerospace plane, an intelligent neural network adaptive compound control method is designed. Based on error feedback learning criterion, the neural network weights are updated online to compensate the attitude control instructions output by the total attitude control law. At the same time, for the strong coupling problem in reentry process, the coupling control coefficient is introduced to reduce the control interference between the roll and yaw channels. In addition, an adaptive chain control allocation law is proposed. The orthogonal optimization multi sine excitation is introduced into the control signal, and the aerodynamic parameters are identified online based on the recursive least squares method, and then the chain allocation strategy is updated in real time. Finally, the effectiveness and robustness of the neural network adaptive compound control method for the reentry roll and yaw channels of the aerospace plane are verified by the mathematical simulation.

Key words: Aerospace plane, Neural network compound control, Chain allocation, Aerodynamic identification, Recursive least squares