高精度定位系统迟滞非线性建模与控制方法研究

发布时间：2018-04-17 04:34

本文选题：迟滞 + 极限学习机　；参考：《浙江理工大学》2017年硕士论文

【摘要】：随着微纳米技术的快速发展,高精度定位系统也随之成为研究的热点。压电陶瓷、形状记忆合金等智能材料由于具有响应速度快、定位精度高、稳定性好等优点,被广泛应用于高精度定位领域中。然而这些智能材料存在固有的迟滞非线性特性,影响了系统的表现,造成系统的控制精度降低和稳定性能变差。为了消除迟滞对高精度定位系统造成的不良影响,使系统能更好地运行,需要对智能材料中的迟滞非线性进行建模,并对其设计相应的控制器进行有效控制。论文的主要研究工作如下:(1)建立了基于Duhem算子的极限学习机迟滞模型。首先,引入Duhem基本迟滞算子将迟滞的多值映射转化为一对一的映射关系,然后应用极限学习机逼近迟滞关系并进行训练辨识,最后通过对实际的压电执行器进行迟滞建模验证该建模方法的有效性,仿真结果表明该方法能够准确地建立迟滞模型。(2)设计了迟滞非线性系统预设性能自适应滑模反步控制器。针对一类含有迟滞的非线性系统,应用Backlash-like模型来描述系统中存在的迟滞特性。首先介绍了Backlash-like模型的数学表达式以及分析了该模型中的类扰动变量,利用径向基神经网络对该类扰动项进行近似逼近。然后引入预设性能函数对迟滞非线性系统进行误差变换,并结合滑模反步法以及Lyapunov稳定理论,为该系统设计了自适应滑模反步控制器。最后通过仿真进行了验证。(3)设计了输入受限迟滞非线性系统的自适应滑模控制器。考虑一类含有Backlash-like迟滞特性的输入受限非线性系统,首先利用径向基神经网络对迟滞模型中的类扰动项进行近似逼近,然后通过定义一个稳定自适应的辅助补偿系统,采用输入饱和误差动态放大的方法来实现控制饱和的补偿,最后结合Lyapunov函数为该系统完成自适应滑模控制器设计。该方法考虑了控制输入受限,符合实际工程情况,有效减少了智能材料中迟滞非线性对系统造成的不良影响,提高了系统的控制精度和稳定性能。仿真结果表明该控制方法有效。
[Abstract]:With the rapid development of micro-and nano-technology, high-precision positioning system has become a hot spot.Piezoelectric ceramics and shape memory alloy (SMA) have been widely used in the field of high precision positioning due to their high response speed, high positioning accuracy and good stability.However, these intelligent materials have inherent hysteresis nonlinear characteristics, which affect the performance of the system, resulting in the reduction of the control accuracy and the deterioration of the stability of the system.In order to eliminate the bad effect of hysteresis on high precision positioning system and make the system work better, it is necessary to model the hysteresis nonlinearity in intelligent materials and control the corresponding controller effectively.The main research work of this paper is as follows: (1) A hysteresis model of extreme learning machine based on Duhem operator is established.Firstly, the Duhem basic hysteresis operator is introduced to transform the multi-valued mapping of hysteresis into one-to-one mapping relationship, then the hysteresis relation is approximated by the extreme learning machine and the training identification is carried out.Finally, the effectiveness of the modeling method is verified by the hysteresis modeling of the practical piezoelectric actuator. The simulation results show that the proposed method can accurately establish the hysteresis model. 2) an adaptive sliding mode backstepping controller with preset performance for a hysteresis nonlinear system is designed.For a class of nonlinear systems with hysteresis, the Backlash-like model is applied to describe the hysteresis characteristics of the system.Firstly, the mathematical expression of Backlash-like model is introduced, and the similar perturbation variables in the model are analyzed. The radial basis function neural network is used to approximate the disturbance term.Then an adaptive sliding mode backstepping controller is designed for the hysteretic nonlinear system based on the sliding mode backstepping method and Lyapunov stability theory.Finally, an adaptive sliding mode controller for input constrained hysteretic nonlinear systems is designed and verified by simulation.In this paper, a class of input-constrained nonlinear systems with Backlash-like hysteresis characteristics is considered. First, the radial basis function neural network is used to approximate the perturbation term in the hysteresis model, and then a stable adaptive auxiliary compensation system is defined.The method of dynamic amplification of input saturation error is used to compensate the control saturation. Finally, the adaptive sliding mode controller is designed for the system with Lyapunov function.This method takes into account the limited control input and accords with the actual engineering conditions. It can effectively reduce the adverse effects of hysteresis nonlinearity on the system in intelligent materials and improve the control accuracy and stability performance of the system.Simulation results show that the control method is effective.
【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】