【作者】 魏立新；

【导师】 王洪瑞； 宋维公；

【作者基本信息】 燕山大学 ， 控制理论与控制工程， 2003， 硕士

【摘要】 本文以燕山大学自行研制的液压驱动六自由度并联机器人实验室样机为研究对象。首先介绍了六自由度并联机器人的结构以及六自由度平台的发展概况、应用前景。目前有关并联机器人动力学方面的研究还不充分，所建立的动力学模型过于复杂，不便于控制理论和控制工程学者进行研究。为此，本文利用拉格朗日方法对并联机器人的动力学模型进行了详细推导，建立起了六自由度并联机器人的拉格朗日动力学模型。该模型不仅全面地表征了六自由度并联机器人的动力学特性，而且形式简单、通用，同串联机器人的动力学模型结构完全相同，使得已经发展成熟的串联机器人控制算法应用于并联机器人成为可能；在建立了动力学模型的基础上，从实时性要求的角度出发，提出利用多处理器进行并行运算，这样采用并行处理方法可以大大地降低运算时间，从而为投入实际应用打下坚实的基础；由于机器人系统参数的不确定性导致在系统坐标变换过程中出现误差进而影响系统的控制效果，本文利用估计的雅可比矩阵，提出一种PID控制器，仿真结果证明该控制方法有效；目前机器人控制器绝大部分要求全状态反馈，但是其中的速度反馈难以实现，针对这种情况，利用观测器观测的速度代替实际的速度信号，所得到的控制效果比较理想；另外随着系统日益复杂，不可能对系统实现精确地建模，而模糊控制集中了人类专家的知识，无需系统模型，自适应控制能够对系统参数的不确定性产生较好的控制效果，将自适应控制和模糊控制结合形成自适应模糊控制达到了很好的控制效果，并证明了系统的渐进稳定性，同时将其引入到6-DOF并联机器人的运动控制中，仿真结果验证了方法的有效性。更多还原

【Abstract】 This paper mainly studies the experimental prototype of hydraulically actuated 6-DOF parallel robot that developed by Yanshan university. Firstly, the structure, present situation of 6-DOF parallel robot and its application prospect are introduced. Now the research on dynamics of parallel robot is not sufficient and its dynamics model is too complex to study for control theory and control engineering scholars. To this end, using Lagrange method, this paper deduced the dynamics model of parallel robot in detail and set up a Lagrange dynamics model of 6-DOF parallel robot with simple and universal form, which shows the dynamics characteristics fully. Moreover, the same structure as serial robot’s makes the maturely serial robot control methods can be adopted to parallel robot. From the view of real time request, a multi-processor method to do parallel operation that can reduce operation time greatly is presented, thus it lays a good foundation for practice. It also will impact on the control effect for error during the system coordinate conversion resulted in the uncertainty of robot system parameter. Using estimated Jacobian matrix, a kind of PID controller is proposed in this paper and simulation result shows the expected control effect. The most robot controller require all-state feedback while the velocity feedback is hard to realize, point to this, we replaced the actual velocity signal with observed velocity to achieve the expect control effect. In addition, it’s impossible to attain an accurate modeling in view of increasingly complex system. Fuzzy control do not require system model and adaptive control do well with system uncertainty, so it will bring better control effect and testify the asymptotic stability as a result of adaptive control combined with fuzzy control, at the same time, it can introduce to the motion control of 6-DOF parallel robot. Simulation result validates its effectiveness.更多还原