【作者】 李宏胜；

【导师】 周杏鹏；

【作者基本信息】 东南大学 ， 检测技术与自动化装置， 2005， 博士

【摘要】 轮廓运动控制是运动控制领域研究的重要课题之一,其广泛应用于精密机械加工与其它自动化运动控制系统中。随着相关自动化设备对轮廓运动精度要求的提高,研究先进的轮廓运动控制策略具有理论意义和重要的应用价值。本文针对精密机械加工控制系统跟随控制与轮廓控制两类控制策略进行了相关研究。通过对常规各轴独立跟随控制的研究,分析了因伺服系统有限带宽引起的半径误差和运动轴性能不匹配引起的椭圆误差,并对数控机床进给伺服系统及传动机构中影响轮廓运动精度几个主要因素进行仿真研究,给出了在不同因素影响下轮廓运动的典型误差分布曲线。为提高系统的跟随性能,本文还对零相差跟随控制算法与干扰观测器结合的控制方案及干扰观测滤波器的设计方法进行了讨论。跟随控制系统在轮廓控制意义上是开环的,事实上在类似机床数控系统等许多应用中,轮廓误差较跟随误差具有更为重要的意义。本文在轮廓误差估计算法的基础上,分析了交叉耦合轮廓闭环控制特性,提出并研究了采用在线模糊自整定参数交叉耦合的控制方案。工程上,许多的运动控制需要执行沿一个轨迹的重复运行,通常的控制算法并未考虑此类运动的重复特性。迭代学习控制通过对每次运动误差的学习来提高运动精度。本文给出了一类离散线性系统高阶迭代学习算法的最优解,并提出了交叉耦合迭代学习控制的思想。同时还给出了存在不确定及干扰非线性时变系统迭代学习控制收敛的充分条件,并对闭环迭代学习控制收敛条件和收敛速度进行了频域分析和讨论。最后,本文介绍了在典型轮廓控制系统——机床数控系统研究中取得的成果和所研制数控系统的技术特点。更多还原

【Abstract】 Contouring motion control is one of the important topics in motion control research, which has been widely applied in motion control tasks in manufacturing and other automation systems. As higher contour precision in modern automation systems is demanded, research on advanced contouring control strategies are very meaningful in theory and application.This dissertation focuses on two control methods, i.e., tracking control and contouring control. Based on mathematic modeling and distinguishing feature of each independent tracking control axis, two errors, that is, the circular arc radius error due to the finite bandwidth of the servo system and the elliptic error due to performance mismatch between motion axes, have been analyzed respectively. Several major factors affecting the contouring motion precision in CNC feeding servo systems have been studied via simulation. Typical error distribution curves have been thus obtained for contouring motion under the influence of different factors. To enhance the tracking performance, in this dissertation, the zero phase error tracking control (ZPETC) algorithm for discrete-time systems and the disturbance observer (DOB) have been investigated with a proposal to combine ZPETC and DOB for improved tracking performance. Moreover, the DOB filter design method is also discussed.If without any special treatment, any multi-axis tracking control contouring motion system is in nature an open-loop system in contour loop. In fact, in many applications such as CNC systems, contouring error is more important than tracking error. Therefore, this dissertation focused on contour error estimation algorithm and closed-loop contouring control based on the so-called“cross-coupling control”(CCC). A fuzzy logic approach has been proposed to handle the uncertainties in the system with an on-line fuzzy automatic parameter setting strategy in CCC.In many motion control tasks, it is usually required to follow a trajectory repeatedly. Conventional control algorithms do not take advantage of the repetitiveness. The basic idea of iterative learning control (ILC) is to construct a compensation signal based on the tracking error in each repetition so as to reduce the tracking error in the next repetition. In this dissertation, an optimal ILC algorithm for discrete time linear systems has been discussed and an integrated control strategy combined ILC algorithm and CCC was proposed. Also, this dissertation established a sufficient condition for ILC convergence of a class of uncertain nonlinear time varying systems with a discussion on the convergence speed. Furthermore, for a class of linear systems, frequency-domain convergence analysis has been provided for a closed-loop type ILC algorithm with a convergence speed comparison with the conventional Arimoto-type open-loop ILC algorithm.Finally, this dissertation introduces the research achievements in typical contouring control system–computerized numerical control (CNC) system for machine tool and technological characteristics of the developed CNC system.更多还原