节点文献

大跨度柔索驱动并联机器人关键问题分析及模型实验研究

Analysis of Key Problems and Model Experimental Research on a Long-span Wire Driven Parallel Robot

【作者】 汤奥斐

【导师】 仇原鹰;

【作者基本信息】 西安电子科技大学 , 机械制造及其自动化, 2007, 博士

【摘要】 采用索原长求解算法构建了大射电望远镜(Large Radio Telescope-LT)舱索系统的运动学正、逆解模型,实现了基于变形索长调整的动平台开环运动控制,为LT500m原型的精确控制奠定了理论基础。完成了LT5m模型的零位标定、舱体的静态及运动误差测定等实验,为LT改进构型的模型试验奠定了试验基础。确定了大跨度柔索驱动并联机器人(Wire Driven Parallel Robot-WDPR)的可达工作空间,提出了其刚度性能的判定准则,完成了LT50m舱索系统刚度性能的判定。导出了大跨度WDPR刚度的解析公式,发现和总结了索塔高度和索张力变化对结构刚度的影响规律,从而为其刚度改善、振动抑制、控制带宽的确定奠定了基础。而且,鉴于大跨度WDPR的动平台绕自身轴的扭转刚度弱、易晃动的特性,通过实验和仿真等方法确定了机构的优选改进方案。优化设计了两种改进构型(动平台增加盛水容器和动平台附加被动索系)的结构参数,验证了两种构型抑制悬索的虚牵和风致振动的有效性,为LT原型的构型选择和设计奠定了理论基础。设计了基于静态标定法的WDPR运动学参数标定方案,并结合LT5m模型试验的仿真结果验证了静态标定算法的收敛性,显示了几何参数对LT5m动平台位姿误差的影响起主要作用。为了实现在动态变化环境中WDPR系统的参数标定,提出了基于神经元网络(Artificial Neural Networks-ANN)的柔性标定法。柔性标定法的仿真分析则证明了柔性标定法的有效性,从而为大跨度WDPR系统提供了新的标定方法。构建了大跨度WDPR系统的逆动力学模型,并通过数值仿真验证了模型的合理性,提高了WDPR的运动控制的精度。建立了LT悬索张力和系统基频的相似型经验公式,导出了LT畸变模型对原型的畸变系数,实现了对LT500m原型的悬索张力和系统基频的预测。应用模糊数学理论提出了WDPR的相似度概念,确定了LT畸变模型与LT500m原型间的悬索张力和系统基频的相似度,验证了相似度数学描述的合理性,提供了LT模型相对原型畸变程度的衡量指标。

【Abstract】 Based on the solution of the unstressed length of the cable, the direct and inverse kinematics models for the cable-cabin system of a Large Radio Telescope (LT) are solved. Thus the unlooped control of movement of the cabin is realized by adjusting the transformative lengths of the cables, which lays the theoretic foundation for the fine control of the LT500m prototype. The label of the home pose, the measurements of static error and dynamic error of the cabin and so on experiments are finished, which is the experimental groundwork for the model experiments of the amendatory structure for LT.The reachable workspace of a long-span Wire Driven Parallel Robot (WDPR) and its estimation rule of the stiffness characteristics are advanced. The workspace and the stiffness capability for LT50m are determined respectively. The analytic expression of the stiffness matrix of WDPR is deduced. The variation law of stiffness with tower heights and cable tensions is founded, and the conclusions will be used as good reference for the stiffness amelioration, vibration control, and control bandwidth determination. What’s more, in view of the weak torsional stiffness and movability of the cabin for WDPR, the better amelioration structure is determined by experiments and simulation.The structure parameters of two amelioration structures, cabin with add-on vessel and cabin with inactive cables, are optimized. That the two constructions can control pseudo-drag cable and wind induced vibration is verified. The results will be used for the structure type and design for LT500m prototype.The approach of kinematics parameter demarcation for WDPR based on static demarcation is designed. The convergence of the static demarcation is validated by simulation on LT5m model experiments, showing that the geometry parameters have a dominating effect on the position and orientation of the moving platform for LT5m model. To realize dynamic parameter demarcation for WDPR, flexible demarcation based on Artificial Neural Networks (ANN) is brought forward. The simulation of flexible demarcation shows its validity so as to provide a new demarcation for LT500m prototype system.The inverse dynamic model for a long-span WDPR is founded, and the rationality of the dynamic model is verified by numerical analysis. The higher precision of the control for WDPR is obtained.The similar empirical formulas of the cable tension and the structural natural frequency for LT are established. The corresponding forecasted coefficients of aberrant model for LT versus LT prototype are deduced. The predictions of the cable tension and the structural natural frequency for LT500m are realized. Similar degree for WDPR system is defined on the fuzzy mathematical method, and the similar degrees of the cable tensions and the natural frequencies between LT model and LT prototype are confirmed by simulation. The rationalization of the similar degree definition is proved. Similar degree applies the scale guideline to aberrance grade between LT model and LT prototype.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络