纯反馈切换非线性系统的自适应神经控制器设计

发布时间：2018-02-20 23:02

本文关键词： 切换系统纯反馈非线性系统 Backstepping技术神经网络控制　出处：《渤海大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来,切换非线性系统因其复杂的属性及其在实际工程中广泛的应用,使得该类系统一直都是控制领域中一个热门的研究课题。虽然吸引了不同研究背景人员的兴趣和关注,但切换非线性系统仍有许多难题需要研究解决。本文针对几类切换纯反馈非线性系统,结合Backstepping技术和自适应神经网络控制技术,研究了相应控制器的设计方法以及闭环系统的稳定性和收敛性问题。主要工作如下:针对一类带有反冲滞后的纯反馈切换随机非线性系统,首先构造一个放射变量避免使用中值定理来拆分系统,并通过加入一个额外的低通滤波器处理项数爆炸问题。然后,结合自适应Backstepping技术和神经网络技术,提出自适应神经网络控制方法,来保证闭环系统的稳定性和收敛性。仿真实例验证了控制方法的有效性。针对一类带有时滞和量化输入的纯反馈切换非线性系统,首先应用函数分离技术进行变量分离,通过构造适当的Lyapunov-Krasoviskii泛函处理系统中的时滞项。然后,结合Backstepping技术、双曲正切函数、自适应神经网络控制技术及动态面控制技术,提出自适应神经网络控制方法,保证闭环系统的稳定性和收敛性。仿真实例说明了控制法则的有效性。纯反馈切换非线性系统的控制问题研究还处在起步阶段,本文初步研究了几类纯反馈切换非线性系统的自适应控制设计问题。目前,还有许多相关控制设计问题亟待解决,例如此类系统的输出反馈控制、约束控制以及互联纯反馈系统的控制设计等。
[Abstract]:In recent years, switched nonlinear systems have been a hot research topic in the control field because of their complex properties and their wide application in practical engineering, although they have attracted the interest and attention of people from different research backgrounds. However, there are still many problems to be solved for switched nonlinear systems. In this paper, Backstepping technology and adaptive neural network control technology are combined for several classes of switched pure feedback nonlinear systems. The design method of the corresponding controller and the stability and convergence of the closed-loop system are studied. The main work is as follows: for a class of pure feedback switched stochastic nonlinear systems with recoil delay, First, a radiative variable is constructed to avoid the use of the mean value theorem to split the system, and an additional low-pass filter is added to deal with the item number explosion problem. Then, combining adaptive Backstepping technology with neural network technology, An adaptive neural network control method is proposed to ensure the stability and convergence of the closed-loop system. The effectiveness of the control method is verified by a simulation example. For a class of pure feedback switched nonlinear systems with time delay and quantization input, First, the variable separation is carried out by using the function separation technique, and the delay term in the system is processed by constructing the appropriate Lyapunov-Krasoviskii functional. Then, combining the Backstepping technique, the hyperbolic tangent function, the adaptive neural network control technology and the dynamic surface control technology, the method is introduced. An adaptive neural network control method is proposed to ensure the stability and convergence of the closed-loop system. Simulation examples show the effectiveness of the control law. The control problem of pure feedback switched nonlinear systems is still in its infancy. In this paper, the adaptive control design problems for several classes of pure feedback switched nonlinear systems are preliminarily studied. At present, there are still many related control design problems to be solved, such as the output feedback control of such systems. Constrained control and control design of interconnected pure feedback system.
【学位授予单位】：渤海大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O231

【参考文献】