【作者】 韩庆涛；

【导师】 郭庆鼎；

【作者基本信息】 沈阳工业大学 ， 电力电子与电力传动， 2003， 硕士

【摘要】 虚拟轴机床具有结构简单、刚度大、承载能力强、位置误差不积累等特点，在应用上与串联机床成互补关系，现已成为机床领域的研究热点，并在机构学、运动学和动力学研究上取得了大量成果，而对运动控制的研究非常少。然而，虚拟轴机床结构的简化所导致的控制困难已严重阻碍了其发展进程，成为亟待解决问题。本文以目前主要存在的滑块式和Stewart平台虚拟轴机床为研究对象，从寻求实用、便于实现的控制策略角度出发，对其运动控制进行了研究。 本文首先介绍了虚拟轴机床的发展概况、研究现状和应用前景，然后对其运动学、动力学、直线伺服系统模型进行分析，针对目前动力学模型过于复杂或完全从机构学角度建立不便于控制的问题，提出了简单、易于实现的控制思想，即将虚拟轴机床的动力学方程反解所确定的各杆受力(矩)视为电机的可变负载力，以杆长伸缩量或滑块位移为伺服系统输出，从此构成位置闭坏，实现轨迹跟踪控制。 在建立了模型的基础上，本文针对单滑块控制系统要受到其它滑块及负载扰动的影响，采用H_∞控制方法设计的H_∞控制器与传统IP位置控制器结合，使得系统具有IP位置控制器和H_∞控制的双重优势，有效地抑制了变化扰动对位置输出的影响，仿真结果表明该方法大大提高了系统的鲁棒性； 本文提出的另一种方案是将杆间的耦合作用和负载扰动看作电机参数变化，利用观测器进行观测，从而将耦合的机床动力学系统解耦成为6个独立的电机动力学系统。对单杆系统，位置和速度控制器由神经元网络实现，网络输出为产生理想加速度所需的电流，实现了加速度的跟踪控制，提高了动平台的轨迹跟踪性能。文中将神经网络与IP控制器结合，权的初始值可据其意义设定，大大加快了网络的收敛速度。 本文提出了基于DSP的虚拟轴机床控制系统的设计构想，并通过对无刷直流机软件设计为开发虚拟轴机床控制系统奠定了基础。 最后，对虚拟轴机床的控制策略及控制系统实现进行了展望。更多还原

【Abstract】 Virtual Axis Machine Tool (VAMT) is characterized by the simple structure, high rigidity, large load handling capacity, and non-accumulation of position error. As the supplement of Series Machine Tool (SMT), VAMT has become the focus of research in machine tool, and plentiful research achievements have been acquired in mechanics, kinematics, dynamics, but few in kinematic control. However, difficulty in control of VAMT, brought by the simplification of its mechanical structure, has baffled its development, and has been a matter of great urgency. In this dissertation, based on the 6 hexaglide and Stewart platform VAMT, the kinematics control strategies were researched, in order to find the practical and accepted solutions.In this dissertation, the development survey, research actualities and application prospect of VAMT are introduced at first, and the models of kinematics, dynamics and linear motor servo are analyzed then. Currently the dynamics model is complex or established mechanically, which restricts the study of control theories. Thus this dissertation present a simple and easy-realized control idea, namely, the fore and torque of every leg determined by the inverse-solution of mechanics are regarded as changeable load force (disturbance) applied in motor, the elongation of the leg or displacement of sliding is regarded as the output of linear servo system. Then the position close-loop is constituted and trajectory track control is realized.On the base of model, considering the disturbance from the other 5 legs and variating load, we design an H-inf controller integrated with IP controller, which makes the system with the advantage of both H-inf control and IP control, and restrains the bad effect from variating disturbance. The simulation shows that this stradgy enhances the system robustness.In another solution, the coupling and load disturbances are regarded as the parameter variations of motor model, and the variation is observed, thus, the coupled dynamics of VAMTis decoupled to 6 single motor systems. For single loop system, the position and speed controllers are realized by neural network, the output of network is determined by the electric current related with the desired acceleration, and the acceleration tracking control is realized indirectly, which improves the trajectory tracking performance of the mobile platform. The initial weights of the neural network can be given according to the material meaning, which expedites the network convergence.This dissertation presents the conceiving of DSP-based V AMT control system. Developing software for brushless direct current electrical motor will benefit the practical development of DSP-based VAMT control system.Finally, we forecast the future control strategies and control system from different aspects.更多还原