基于Qball-X4的四旋翼飞行器自适应控制  

Adaptive Control for Quadrotor UAVs Based on Qball-X4

作　　者：张叶[1] 钟诚 王云涵 ZHANG Ye;ZHONG Cheng;WANG Yunhan(School of Automation Engineering,Northeast Electric Power University,Jilin 132012,China;Tangshan Power Supply Company,State Grid Jibei Electiric Power Co.,Ltd.,Tangshan 063000,China)

机构地区：[1]东北电力大学自动化工程学院,吉林吉林132012 [2]国网冀北电力有限公司唐山供电公司,河北唐山063000

出　　处：《控制工程》2024年第8期1512-1521,共10页Control Engineering of China

基　　金：吉林省科技发展计划项目(20210509053RQ)。

摘　　要：为实现四旋翼飞行器的精确轨迹跟踪控制,提出了一种基于神经网络逼近器的复合式自适应动态面量化控制方案。该方案主要贡献:①实现了自适应神经网络的量化控制,有效地缓解了量化器引起的强非线性,提高了四旋翼飞行器实时控制系统控制精度;②将自适应神经网络动态面控制技术应用于欠驱动四旋翼飞行器的控制系统中,克服了传统反推方法中的“微分爆炸”问题,并获得了跟踪性能的L∞范数。最后,在Qball-X4四旋翼飞行器运动控制平台上进行了实验,验证了所提控制方案的有效性。To achieve precise trajectory tracking control of a quadrotor UAV,a composite adaptive dynamic surface quantized control scheme based on neural network approximations is proposed.The main contributions are as follows:①the adaptive neural network quantized control is realized,which effectively alleviates the strong nonlinearity caused by the quantizer,which means that in the real-time control system of quadrotor,the control accuracy can be improved.②The adaptive neural network dynamic surface control technology is applied to the control system of under actuated quadrotor,which overcomes the problem of“explosion of complexity”in the traditional backstepping method,and obtains the L∞-norm of tracking performance.Finally,experiments on the motion control platform of Qball-x4 quadrotor verify the effectiveness of the proposed control scheme.

关 键 词：四旋翼飞行器 自适应量化控制 动态面控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]