机器人
機器人
궤기인
ROBOT
2015年
2期
152-160
,共9页
宋大雷%孟祥冬%齐俊桐%韩建达
宋大雷%孟祥鼕%齊俊桐%韓建達
송대뢰%맹상동%제준동%한건체
旋翼无人机%飞行机械臂%动力学建模%预测控制%飞行仿真
鏇翼無人機%飛行機械臂%動力學建模%預測控製%飛行倣真
선익무인궤%비행궤계비%동역학건모%예측공제%비행방진
RUAV (rotorcraft unmanned aerial vehicle)%aerial manipulator%dynamics modeling%predictive control%flight simulation
旋翼飞行机械臂(rotorcraft aerial manipulator,RAM)系统是安装在飞行机器人上的可操作型机械臂,悬停模式下执行准确的空中操作时旋翼无人机与所加机械臂之间存在相对扰动,通过分离机械臂与飞行机器人进行动力学建模并不能有效消除这种扰动。本文基于对相互扰动力学作用的分析建立整体动力学模型,并在悬停飞行模式下将其简化为线性控制参考模型。进而对旋翼系统控制延时所引起的动力学扰动进行补偿,同时设计预测控制器来消除末端执行器的位置和姿态误差。最后,在存在内部和外部扰动的情况下,设定销钉插入操作任务进行控制方法的对比仿真。末端执行器位姿偏差的仿真结果表明了模型结构与控制方法的有效性。
鏇翼飛行機械臂(rotorcraft aerial manipulator,RAM)繫統是安裝在飛行機器人上的可操作型機械臂,懸停模式下執行準確的空中操作時鏇翼無人機與所加機械臂之間存在相對擾動,通過分離機械臂與飛行機器人進行動力學建模併不能有效消除這種擾動。本文基于對相互擾動力學作用的分析建立整體動力學模型,併在懸停飛行模式下將其簡化為線性控製參攷模型。進而對鏇翼繫統控製延時所引起的動力學擾動進行補償,同時設計預測控製器來消除末耑執行器的位置和姿態誤差。最後,在存在內部和外部擾動的情況下,設定銷釘插入操作任務進行控製方法的對比倣真。末耑執行器位姿偏差的倣真結果錶明瞭模型結構與控製方法的有效性。
선익비행궤계비(rotorcraft aerial manipulator,RAM)계통시안장재비행궤기인상적가조작형궤계비,현정모식하집행준학적공중조작시선익무인궤여소가궤계비지간존재상대우동,통과분리궤계비여비행궤기인진행동역학건모병불능유효소제저충우동。본문기우대상호우동역학작용적분석건립정체동역학모형,병재현정비행모식하장기간화위선성공제삼고모형。진이대선익계통공제연시소인기적동역학우동진행보상,동시설계예측공제기래소제말단집행기적위치화자태오차。최후,재존재내부화외부우동적정황하,설정소정삽입조작임무진행공제방법적대비방진。말단집행기위자편차적방진결과표명료모형결구여공제방법적유효성。
Rotorcraft aerial manipulator (RAM) system is an aerial robot with manipulators. When performing precise operation in hovering mode, there exists relative disturbance between the rotorcraft aerial vehicle and the manipulator, which cannot be eliminated through establishing dynamic models of the manipulator and the rotorcraft separately. In this research, the overall dynamics model is firstly developed based on dynamic disturbance of the both components, which is simplified as a linear control reference model in hovering mode. The dynamics disturbance caused by rotor system’s control delay is compensated, and a predictive controller is designed to eliminate the errors of position and attitude of the end-effector. At last, control strategies are compared in simulative peg-in-hole tasks in cases of external and internal disturbances. The effectiveness of the proposed model and control method is verified by the simulation results of end-effector pose error.