【作者】 张勇；

【导师】 汪成文；

【作者基本信息】 太原理工大学 ， 机械工程， 2024， 硕士

【摘要】 液压机械臂相较于电驱机械臂,具有承载能力强、功率密度高、抗环境干扰能力强等优点,但在运动控制方面却存在着诸多问题,主要是由于机械臂和电液伺服系统是高度非线性的,且存在未知干扰的影响,这使得伺服控制的难度大幅增加。本文旨在实现液压机械臂高精度运动控制的同时提高其抗干扰能力。首先,分析所研究的液压驱动机械臂的三维模型,并借助该模型详细介绍了液压机械臂的结构与工作原理,建立相应的DH坐标系,借此推导出每个关节处的旋转矩阵与原点坐标,并对每个关节质心处的雅克比矩阵进行计算,为后续动力学分析打下基础。其次,采用欧拉-拉格朗日方程这一经典的动力学建模方法,借助之前推导出的运动学方程,逐一推导出机械臂的惯性矩阵、重力矩阵以及离心力与科氏力矩阵。并结合液压执行器的动力学方程,推导出液压驱动机械臂的综合动力学模型。然后,通过对液压机械臂综合动力学模型的分析,可以清晰地了解到阀控液压作动单元压力控制的精确性对于机械臂运动控制至关重要。精确掌握液压执行器的压力将实现对机械臂驱动力的准确调节,从而实现对机械臂关节角度的精准跟踪控制。因此,设计了液压执行器压力控制策略进行研究,并通过压力测试设备进行了实验分析。在实验中发现电液伺服系统中存在许多不确定性,使用滑模控制容易引发抖振。所以引入扩张状态观测器来观测和补偿未知干扰,将估计的干扰值作为前馈信号补偿给滑模控制器,以减小切换项增益来抑制滑模抖振,将之命名为滑模抗干扰控制。利用Lapunov稳定性分析方法证明了该控制器的稳定性,通过仿真与实验,又验证了该控制方法的有效性。最后,综合考虑机械臂与液压执行器的动力学方程,针对液压机械臂中存在的高度非线性、未建模摩擦力、外部随机干扰等问题,设计了一种基于非线性干扰观测器的反演鲁棒控制方法。该非线性干扰观测器专门针对机械臂系统设计,特点是将非线性干扰观测器的设计问题表述为了线性矩阵不等式的求解。该观测器可以对机械臂中存在的关节摩擦、外部干扰力矩等进行准确估计,与反演鲁棒控制相结合,显著提高了控制性能。对所提控制方法的稳定性进行理论推导与分析,并通过AMESim、MATLAB/Simulink联合仿真验证其有效性。仿真研究表明,该控制方法能够准确观测与补偿未知干扰,控制效果以及抗干扰能力明显优于PID以及反演鲁棒控制。更多还原

【Abstract】 Compared to electrically driven manipulators,hydraulic manipulators offer advantages such as stronger load-bearing capacity,higher power density,and better resistance to environmental disturbances.However,they face various challenges in motion control due to the highly nonlinear nature of the manipulator and the electro-hydraulic servo system,coupled with the influence of unknown disturbances,the difficulty of servo control is significantly increased.This paper aims to achieve high-precision motion control of hydraulic manipulators while enhancing their anti-disturbance capability.Firstly,a fully hydraulic driven robotic arm was designed to meet the special and hazardous operating requirements such as underground mining in coal mines.Using Solidworks,a three-dimensional model of the hydraulic manipulator was created.This model detailed the structure and working principles of the hydraulic manipulator,and corresponding DH coordinate systems were established.Through this,the rotation matrices and origin coordinates at each joint were derived,and the Jacobian matrices at the centroid of each joint were calculated,laying the foundation for subsequent dynamic analysis.Secondly,employing the classical Euler-Lagrange equation as a dynamic modeling method and utilizing the previously derived kinematic equations,the inertial matrix,gravity matrix,and centrifugal and Coriolis force matrices of the robotic arm were derived one by one.Combined with the dynamic equations of the hydraulic actuators,the comprehensive dynamic model of the hydraulic-driven robotic arm was established.Subsequently,through analysis of the comprehensive dynamic model of the hydraulic manipulator,it becomes clear that the precision of pressure control in valve-controlled hydraulic actuators is crucial for motion control of hydraulic actuator pressure.Precise control of the hydraulic actuator pressure will achieve accurate adjustment of the driving force of the manipulator,thereby achieving precise tracking control of the joint angles.Therefore,a pressure control strategy for hydraulic actuators was designed and studied,and experimental analysis was conducted using a pressure test device.It was found in experiments that there are many uncertainties in the electro-hydraulic servo system,and using sliding mode control easily induces oscillations.Therefore,an extended state observer was introduced to observe and compensate for unknown disturbances.The estimated disturbance value was used as a feedforward signal to compensate the sliding mode controller,reducing the switching gain to suppress sliding mode oscillations,thus named as sliding mode anti-disturbance control.The stability of this controller has been demonstrated using the Lyapunov stability analysis method,and its effectiveness has been further validated through simulation and experimentation.Finally,considering the dynamic equations of the robotic arm and hydraulic actuators,a robust control method based on a nonlinear disturbance observer was designed to address the highly nonlinear nature,unmodeled friction,and external random disturbances present in hydraulic manipulators.This nonlinear disturbance observer is specifically designed for robotic arm systems and is characterized by formulating the design problem of the nonlinear disturbance observer as solving linear matrix inequalities.This observer can accurately estimate joint friction and external disturbance torques in the robotic arm system.Combined with backstepping robust control,it significantly improves control performance.The stability of the proposed control method was theoretically derived and analyzed,and its effectiveness was validated through joint simulation using AMESim and MATLAB/Simulink.Simulation studies showed that this control method can accurately observe and compensate for unknown disturbances,with control performance and anti-disturbance capability significantly superior to PID and backstepping robust control.更多还原