柔性关节机器人动力学建模及控制

发布时间：2018-01-01 18:28

本文关键词：柔性关节机器人动力学建模及控制　出处：《南京理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：谐波减速器具有传动比大、承载能力大、体积小、高效率等优点,被广泛应用于关节型机器人的机械传动系统中。由于谐波减速器内部存在着柔性元件,在给机器人关节带来附加自由度的同时,也会带来柔性振动、不确定性和未知干扰等一系列问题。为使柔性关节机器人具备良好的控制性能,控制算法需要最大限度地抑制关节柔性振动,具有强的抗扰能力,并且对系统中的不确定因素具有鲁棒性。本文对柔性关节机器人的数学建模与控制方法开展研究。论文主要工作及创新点如下:1)具有柔性关节的四自由度机器人动力学建模。应用拉格朗日法建立了带有柔性关节的四自由度机器人的动力学模型,为柔性关节机器人控制方法研究奠定了基础。2)针对柔性关节机器人的定点控制任务,研究了分级滑模控制(Hierarchical Sliding Mode Control,HSMC)方法。在假设干扰有界和系统状态全反馈条件下,设计了一种分级滑模控制律,并应用Lyapunov理论证明闭环系统的稳定性。针对滑模控制中的抖振抑制问题,提出了两种基于分级滑模思想的改进滑模控制器。首先,通过引入幂函数替代符号函数削弱抖振的影响。进一步地,设计了基于双幂次趋近律的分级滑模控制律,在抑制抖振的同时,使得系统状态兼具较快的趋近速度和较好的动静态性能。仿真结果验证了所提控制算法的有效性。3)针对柔性关节机器人时变轨迹跟踪控制任务,研究了基于动态面的滑模控制方法。在假设干扰有界和系统状态全反馈条件下,设计了一种动态面滑模控制(Dynamic Surface Sliding Model Control,DSSMC)方法,简化了传统反演控制中对虚拟控制量的求导过程,应用Lyapunov理论证明闭环系统半全局一致有界。进一步地,提出了一种模糊动态面滑模控制(Fuzzy Dynamic Surface Sliding Model Control,FDSSMC)方法,采用模糊规则对滑模控制中的切换项增益进行自调整,在对干扰进行补偿的同时,有效减小了抖振。仿真验证了所提控制方法的有效性。4)针对柔性关节刚度系数变化、连杆转动惯量不确定以及柔性振动引起的内部扰动等问题,研究了一种柔性关节自抗扰控制器(Active Disturbance Rejection Controller,ADRC)的设计方法。在建立Quanser柔性关节实验系统动力学模型的基础上,将参数不确定及柔性振动等效为干扰,构造五阶线性扩张状态观测器对其进行估计,并设计了带有干扰补偿的自抗扰控制器。该控制器可调参数少、调试简单,不依赖于关节角速度反馈且鲁棒性好。实验结果验证了控制器的鲁棒性。
[Abstract]:The harmonic reducer with large transmission ratio, large bearing capacity, small volume, high efficiency and other advantages, the mechanical transmission system is widely used in the robot. The harmonic reducer exists inside the flexible element, the additional degrees of freedom at the same time to the robot joint, flexible vibration will also bring a series of problems, uncertainty and the unknown disturbance. The flexible joint robot has good control performance and control algorithms need to minimize joint flexible vibration, strong anti-interference ability, and is robust to the uncertainties in the system. This research focuses on the flexible joint robot modeling and control method. The main work and the innovation points are as follows: 1) the four degree of freedom robot dynamics modeling with flexible joints. The application of Lagrange method to establish the four DOF robot with flexible joints. Dynamic model,.2 provides a basis for research on control method of flexible joint robot) point control for flexible joint robot task, study the hierarchical sliding mode control (Hierarchical Sliding Mode Control, HSMC). Under the assumption of bounded disturbance and the system state feedback condition, design a hierarchical sliding mode control law and application Lyapunov theory to prove the stability of the closed-loop system. To suppress chattering in sliding mode control, presents two kinds of improved sliding mode controller based on the idea of hierarchical sliding mode. First of all, to weaken the buffeting by introducing the power function instead of sign function. Further, the design of hierarchical sliding mode dual power reaching law control law based on the suppress chattering at the same time, the state has fast reaching speed and good dynamic and static performance. The simulation results verified the validity of the control algorithm for.3) The trajectory tracking control of flexible joint robot task of time-varying sliding mode control method based on dynamic surface. Under the assumption of bounded disturbance and system state feedback condition, design a dynamic sliding mode control (Dynamic Surface Sliding Model Control, DSSMC) method, simplifies the derivation process of the virtual control of the traditional inversion control the application of Lyapunov theory, the closed-loop control system is proved to be semi globally uniformly bounded. Furthermore, the paper proposed a fuzzy sliding mode control (Fuzzy Dynamic dynamic surface Surface Sliding Model Control, FDSSMC) method, using the fuzzy rules of switching gain of sliding mode control of self adjustment in compensation to reduce the interference at the same time, the chattering..4 simulation verifies the effectiveness of the proposed control method) according to the changes of the joint stiffness coefficient, connecting the uncertainty of inertia and flexible vibration caused by the Disturbing problems, study a flexible joint ADRC (Active Disturbance Rejection Controller, ADRC) design method. Based on establishing the dynamic model of flexible joint Quanser experimental system, the uncertain parameters and vibration equivalent disturbance, construct five order linear extended state observer to estimate, and designed with disturbance compensation ADRC controller. The controller is less adjustable parameters, simple debugging, do not rely on feedback in the joint angle speed and robust. The experimental results verify the robustness of the controller.

【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】