针对非线性动态逆控制鲁棒性差的缺点，综合最优控制和神经网络控制，提出一种自适应非线性控制策略。首先采用非线性动态逆进行基本控制律设计，然后对由于建模误差或舵面故障等因素导致的动态逆误差，利用神经网络进行在线补偿，根据最优控制理论得到神经网络权值的自适应律，并基于Lyapunov直接法证明该自适应律的稳定性和收敛性。针对某飞机的仿真结果表明，在存在较大逆误差的情况下，所设计的控制系统具有良好的鲁棒性。

In order to improve the robustness of the nonlinear dynamic inversion control lawa neurally augmented control scheme was designed via nonlinear dynamic inversion and optimal control theory.The basic control law was designed using the nonlinear dynamic inversionand a sigma-pi neural network was used to compensate for inversion errors and changes in aircraft dynamicsincluding actuator failures.The necessary condition of optimality was used to derive an adaptive law using the gradient method.The adaptive law was proved to be convergent based on the Lyapunov stability theory.Numerical simulation results for an aircraft performed at actuator failures demonstrate that the robustness of the control law is increased.