【摘要】 针对高超声速变体飞行器变后掠翼系统中存在关节耦合、气动载荷导致的左右机翼变形不同步问题，提出一种基于径向基函数（RBF）神经网络和干扰观测器的分布式协同变形控制方案。将变后掠翼系统分解成多个单输入单输出子系统，关节耦合、气动载荷等效成子系统的集总干扰，机翼的变形控制问题被改造成各驱动关节的分布式控制问题。为提高各关节间运动协同程度，设计出一种仅需交互部分关节长度信息的协同拓扑网络结构，据此构造协同误差后推导的协同变形控制律可同时消除跟踪误差和同步误差。考虑到变形翼后掠角参数未知，以关节长度作为RBF神经网络的输入对变形控制系统模型进行在线辨识。进一步，在协同控制律中引入干扰观测器的估计值以抵消集总干扰对控制系统造成的影响。仿真结果表明，所提出的分布式协同控制方法能够有效提高系统对变形指令的跟踪精度和双翼运动的同步程度。更多还原

【Abstract】 A distributed cooperative control scheme based on a radial basis function(RBF) neural network and disturbance observer is proposed to address the issue of asynchronous wing deformation in hypersonic morphing vehicles, which is caused by joint coupling and aerodynamic loading. The variable-sweep wing system is decomposed into multiple single-input single-output subsystems, where joint coupling and aerodynamic loading are modeled as lumped disturbances.Consequently, the morphing control problem is transformed into a distributed control problem for each driving joint. To enhance the coordination of joint movements, a cooperative topology network structure is designed, requiring only partial interactive joint position information. A cooperative morphing control law is derived by constructing cooperative errors to eliminate both tracking and synchronization errors. Given the unavailability of the morphing wing’s sweep angle parameters, joint positions are utilized as inputs to the RBF neural network for online identification of the morphing control system model. Additionally, the disturbance observer’s estimates are incorporated into the cooperative control law to mitigate the effects of lumped disturbances on the control system. Simulation results indicate that the proposed distributed cooperative control scheme effectively improves the tracking accuracy of morphing commands and the synchronization of bilateral wing movements.更多还原