论文部分内容阅读
为解决反步法在虚拟控制求导时存在计算膨胀的缺陷问题,提出了一种自适应反步控制算法。采用CMAC神经网络在线学习系统不确定性以及各阶虚拟控制量的导数信息,从而避免了系统阶次较高时引起的计算膨胀问题,并且该控制算法在实际控制输入前加入低通滤波器,避免了不连续输入可能引起的抖振问题。Lyapunov稳定性分析表明,该控制方法具有渐近稳定性。将所提出的控制算法应用于先进战斗机飞行控制系统,并进行了仿真验证,结果表明所设计的控制器具有较强的鲁棒性。
In order to solve the defect that the backstepping method has the calculation expansion when the virtual control derives, an adaptive backstepping control algorithm is proposed. The CMAC neural network is used to learn the system uncertainty and the derivative information of each level of virtual control, so as to avoid the computational expansion problem caused by higher system order. Moreover, the control algorithm adds a low-pass filter before the actual control input, Avoids chattering problems that can be caused by discrete inputs. Lyapunov stability analysis shows that the control method has asymptotic stability. The proposed control algorithm is applied to advanced fighter flight control system, and the simulation results show that the designed controller has strong robustness.