基于自适应神经网络的四旋翼无人机轨迹跟踪控制

Trajectory Tracking Control of Quadrotor UAV Based on Adaptive Neural Network

四旋翼无人机在飞行过程中易受到难测量的非线性多源干扰,进而影响系统的稳定性和控制精度。针对此问题,提出了一种基于自适应神经网络的鲁棒轨迹跟踪控制策略。首先,将四旋翼无人机的跟踪问题转化为对位置子系统和姿态子系统的期望指令跟踪控制问题,针对2个子系统设计基于干扰估计器的控制策略,使系统快速逼近期望轨迹;其次,针对干扰估计器存在估计误差问题,在位置子系统引入了自适应神经网络补偿,实现非线性多源干扰的高精度补偿;最后,通过仿真与实验验证所提控制策略的性能。实验结果表明,所提控制策略在快时变干扰下仍然能够实现期望轨迹跟踪控制,风扰环境下悬停和圆形轨迹跟踪的位置控制精度分别达到0.02m和0.1m,姿态控制精度达到0.1°。

Quadrotor unmanned aerial vehicle(UAV)is vulnerable to nonlinear multi-source disturbance which is difficult to measure during flight,thus affects the stability and control accuracy of the system.To solve this problem,a robust trajectory tracking control strategy based on adaptive neural network is proposed.Firstly,the tracking problem of the quadrotor UAV is transformed into the desired command tracking control problem of the position subsystem and the attitude subsystem.Secondly,in response to the estimation error problem of the disturbance estimator,an adaptive neural network compensation is introduced in the position subsystem to achieve high-precision compensation for nonlinear multi-source disturbance.Finally,the performance of the proposed control strategy is verified by simulation and experiments.The experimental results show that the proposed control strategy can still achieve the desired trajectory tracking control under the condition of fast time-varying disturbances.The position control accuracy of hover and circular trajectory tracking under wind disturbances reaches 0.02 m and 0.1 m respectively,and the attitude control accuracy reaches 0.1°.