神经网络在线估计器的悬挂载荷无人机控制  

Robust Control of Suspended Load UAV Based on Hybrid Automata Model

作　　者：李妍缘[1] LI Yanyuan(Zhengzhou Electric Power College,He’nan Zhengzhou 450002,China)

机构地区：[1]郑州电力高等专科学校,河南郑州450002

出　　处：《机械设计与制造》2025年第3期159-166,共8页Machinery Design & Manufacture

基　　金：2021年度河南省教育厅高等学校重点科研项目(21B46001)。

摘　　要：为了能够实现高精度无监督跟踪控制,且保证良好的鲁棒性,提出了一种基于神经网络在线估计器的悬挂载荷无人机控制。首先描述了具有悬挂载荷的无人机模型,从而能够有效描述这种质量变化的非线性复杂系统。然后设计了一种基于反馈线性化控制器的混合控制,该控制器与人工神经网络相结合,作为未知质量的在线估计器。最后仿真结果证明提出控制器提供了良好的跟踪性能,并且无需任何有关负载的信息,此外,不需要根据有效载荷调整控制器的参数。In order to achieve high-precision unsupervised tracking control and ensure good robustness,a suspended load unmanned aerial vehicle control based on neural network online estimator was proposed.Firstly,a drone model with suspended loads was described,which could effectively describe the nonlinear complex system with varying mass.Then a hybrid control based on feedback linearization controller was designed,which was combined with an artificial neural network as an online estimator of unknown quality.The final simulation results demonstrate that the proposed controller provides good tracking performance and does not require any information about the load.In addition,there is no need to adjust the controller parameters based on the payload.

关 键 词：无监督 跟踪控制 无人机 混合自动机 

分 类 号：TH16[机械工程—机械制造及自动化] TP273[自动化与计算机技术—检测技术与自动化装置] V279[自动化与计算机技术—控制科学与工程]