Shen Zhixi, Tan Lian, Yu Shuangshuang, Song Yongduan
IEEE Trans Neural Netw Learn Syst. 2021 Dec;32(12):5610-5622. doi: 10.1109/TNNLS.2021.3071094. Epub 2021 Nov 30.
Most existing control methods for quadrotor unmanned aerial vehicles (UAVs) are based on the primary assumption that the center of gravity (CoG) is fixed and is in the same position as the centroid, which is not necessarily true with swing load as continuously making CoG vary with the swing angle and substantially complicating the dynamic model of UAV. This article presents an adaptive learning and fault-tolerant control scheme for quadrotor UAVs with varying CoG and unknown moment of inertia. First, we establish the dynamic model of quadrotor UAVs in the presence of time-varying CoG, input saturation, and actuator fault. Then, we design a fault-tolerant adaptive learning controller for the quadrotor UAVs and show that both linear and angular velocity tracking errors are ensured to converge to a residual set around zero in the presence of full-state constraints. Furthermore, all signals in the closed-loop system are uniformly ultimately bounded. Simulation studies also confirm the effectiveness of the proposed control method.
重心（CoG）是固定的，且与质心位于同一位置，但对于摆动负载而言，情况并非一定如此，因为摆动负载会使重心随摆动角度不断变化，从而极大地复杂化了无人机的动力学模型。本文提出了一种针对重心变化且转动惯量未知的四旋翼无人机的自适应学习与容错控制方案。首先，我们建立了存在时变重心、输入饱和及执行器故障情况下的四旋翼无人机动力学模型。然后，我们为四旋翼无人机设计了一种容错自适应学习控制器，并表明在存在全状态约束的情况下，线速度和角速度跟踪误差都能确保收敛到围绕零的一个残差集。此外，闭环系统中的所有信号都是一致最终有界的。仿真研究也证实了所提控制方法的有效性。
