双级执行器建模和控制方法研究

发布时间：2019-03-27 09:35

【摘要】：随着科学技术和人类社会的不断发展,传统的定位技术在精度上已经无法满足人们的要求,因此超精密定位系统的研究越来越受到国内外学者的重视。近年来,双级执行器系统的广泛应用极大推动了精密定位领域的发展,它能够将传统的单级执行器与基于智能材料的高精度执行器相结合,同时达到大行程和高精度的要求。本文的双级执行器由伺服电机和压电陶瓷执行器构成,本文重点对基于压电执行器的第二级执行器建立了数学模型并设计了控制器,通过设计基于空间滤波器的协同控制方案实现了两级平台的协调控制,主要内容有:(1)对第二级执行器建立数学模型引入动态迟滞算子建立迟滞动态模型。利用该算子将系统输入拓展到三维空间,通过此三维空间将迟滞的多值映射转换成单一映射。另外在标准极限学习机(ELM)的基础上加入了局部线性嵌入(LLE)算法,提出了LLE-ELM算法,利用LLE-ELM对上述单一映射进行逼近并建立迟滞动态模型,本文所提出的LLE-ELM迟滞模型具有结构简洁,精度高,可以在线调整的特点。(2)对第二级执行器设计控制器针对一类含有backlash-like迟滞的全状态受限的纯反馈非线性系统设计了一种自适应控制器。首先利用中值定理将纯反馈系统转换成严格反馈系统。其次通过引入障碍性李雅普诺夫函数(BLF)来解决全状态受限问题,并且利用RBF神经网络来近似逼近backlash-like迟滞中的类扰动项,最后利用反步控制法设计了自适应控制器。仿真结果表明了该方案的可行性。(3)对双级执行器系统设计协同控制方法设计了一种基于空间滤波器的协同控制方法。通过此滤波器将信号分离成长行程和短行程两部分,两级平台可以根据分离之后的信号独立地设计控制方案,第一级执行器采用近似时间最优控制完成大行程快速定位,第二级执行器采用多层滑模自适应控制完成微定位。本文提出的协同控制方法不但能大幅降低控制器设计的难度,而且能提高系统定位速度。实验结果说明了该方案的可行性。
[Abstract]:With the development of science and technology and human society, the traditional positioning technology has been unable to meet the demands of people in the precision, so the research of the ultra-precision positioning system is more and more important to the domestic and foreign scholars. In recent years, the wide application of the two-stage actuator system has greatly promoted the development of the precise positioning field, which can combine the traditional single-stage actuator with the high-precision actuator based on the intelligent material, and simultaneously achieve the requirements of large-stroke and high-precision. The two-stage actuator of this paper is composed of a servo motor and a piezoelectric ceramic actuator. In this paper, the mathematical model of the second-stage actuator based on the piezoelectric actuator is established and the controller is designed. The coordinated control of the two-stage platform is realized by the design of the cooperative control scheme based on the spatial filter. The main contents are as follows: (1) The mathematical model of the second-stage actuator is introduced to introduce the dynamic hysteresis operator to set up the hysteresis dynamic model. The operator is used to expand the system input into a three-dimensional space, and the multi-value mapping of the hysteresis is converted into a single map by the three-dimensional space. In addition, on the basis of the standard limit learning machine (ELM), a local linear embedding (LLE) algorithm is added, the LLE-ELM algorithm is proposed, the single mapping is approximated by the LLE-ELM, and a hysteresis dynamic model is established. The LLE-ELM hysteresis model proposed in this paper has the advantages of simple structure and high precision. And can be adjusted on line. (2) A self-adaptive controller is designed for a class of pure-feedback non-linear systems with back-to-like hysteresis. First, a median theorem is used to convert the pure feedback system into a strict feedback system. Secondly, the whole state-limited problem is solved by the introduction of the Aplastic Lyapunov function (BLF), and the class-disturbance term in the background-like hysteresis is approximated by the RBF neural network, and the adaptive controller is designed by using the reverse-step control method. The simulation results show the feasibility of the scheme. (3) A cooperative control method based on spatial filter is designed for the design of the two-stage actuator system. the two-stage platform can independently design the control scheme according to the signal after the separation, the first-stage actuator can complete the large-stroke rapid positioning by using the approximate time optimal control, And the second-stage actuator adopts the multi-layer sliding mode adaptive control to complete the micro-positioning. The cooperative control method proposed in this paper can not only greatly reduce the difficulty of the design of the controller, but also improve the positioning speed of the system. The experimental results show the feasibility of the scheme.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP273

【参考文献】