单侧机翼损伤飞机的神经网络自适应鲁棒非线性控制

针对飞机飞行中单侧机翼突然损伤问题,结合对损伤飞机的特性分析,提出基于神经网络自适应补偿的鲁棒非线性模型逆控制方法。利用未损伤飞机模型伪控制量中的单隐层神经网络自适应项和鲁棒项,并联合e modification自适应律对模型误差、外界扰动、神经网络近似误差进行补偿。除此之外,利用动态非线性阻尼技术对上述伪控制律进行扩展,从而适应损伤机体未建模舵动态。最后对上述算法进行严格的稳定性证明,并推导了逆过程的实现方法。仿真结果表明在单侧机翼突然损伤并伴随外部扰动和未建模舵动态下,该控制方法具有较强的稳定性和鲁棒性。

Neural network based robust adaptive nonlinear control for aircraft under one side of wing loss

Combined with the characteristics of wing damaged aircraft, a method of neuro adaptive compensation based robust nonlinear model inversion control is proposed for one side of wing damage suddenly in flight. The method employs one single hidden layer neural network (SHL NN) adaptive element and one robust element in pseudo control of the undamaged aircraft model with e modification adaptive laws to compensate model errors, external disturbances and NN approximation errors simultaneously. In addition, a dynamic nonlinear damping technique is employed to expand the pseudo control law above for robustifying the unmodelled actuator dynamics of the damaged plant. Finally, the strict stability proof is given and the realization of the inversion process is derived. The simulation results validate the strong stability and robustness of the control algorithm under wing damage accompanied with output noise and unmodelled actuator dynamics.