改进全局滑模控制及应用研究

发布时间：2018-01-14 01:17

本文关键词：改进全局滑模控制及应用研究　出处：《天津工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：滑模变结构控制作为现代控制理论中的重要分支之一,以其物理实现简单、对系统内部参数的不确定性以及外部干扰不敏感的优点在线性系统和非线性系统的控制中得到了广泛应用。传统滑模控制的响应可分为趋近模态和滑动模态,然而系统对于不确定性的鲁棒性只体现在滑动模态,也就是并不具有全程的鲁棒性。全局滑模控制方法通过引进全程时变因子使系统从一开始便落到滑模切换面上从而实现了被控系统的全程鲁棒性,但是在全局滑模切换面的演变过程中只有时间趋近于无穷时才能实现完全转变。本文针对全局滑模控制中所出现的动态滑模面不能在有限时间内演变为传统线性滑模面的缺点,提出了一种改进的全局滑模控制策略,所设计的动态滑模切换面中的非线性函数将普通全局切换面中的单项衰减函数替换为三项一阶可导的指数衰减函数的和,设计出的新函数能在指定时刻衰减为0,从而保证了有限时间内滑模切换面向传统线性切换面的转换,在保证系统全局鲁棒性的基础上有效减少了过渡时间,从而能加快系统的动态响应速度,缩短收敛时间。以n阶不确定连续系统为被控对象设计了改进的等效全局滑模控制律,并选取Lyapunov函数对其控制器的稳定性进行证明。为验证本文所提出的控制策略在实际中的应用,将所设计的改进全局滑模控制律应用到电力系统的混沌控制以及不确定性迟滞混沌神经元/网络与确定性迟滞混沌神经元/网络的同步控制中,并通过对比仿真实验分别验证了本文所提方法相较于普通全局滑模控制方法的优越性。在被控对象为连续系统的基础上进一步探索,将本文所提方法应用到离散时间系统中,重新构造了离散形式的衰减函数,并设计了基于指数离散趋近律的改进全局滑模控制律,通过仿真实验同样验证了其可行性。
[Abstract]:Sliding mode variable structure control is one of the important branches of modern control theory. The advantages of indeterminacy of internal parameters and insensitivity to external disturbances are widely used in the control of linear and nonlinear systems. The response of traditional sliding mode control can be divided into approaching mode and sliding mode. However, the robustness of the system to uncertainty is only reflected in the sliding mode. The global sliding mode control method can make the system fall to the sliding mode switching surface from the beginning by introducing the global time-varying factor to achieve the global robustness of the controlled system. However, in the evolution of the global sliding mode switching surface, only when the time approaches infinity, can complete transformation be realized. In this paper, the dynamic sliding mode surface in the global sliding mode control can not evolve into the traditional linear in finite time. The disadvantage of sliding mode surface. An improved global sliding mode control strategy is proposed in which the nonlinear function in the dynamic sliding mode switching surface is designed to replace the single attenuation function in the general global switching surface with the sum of the three first order differentiable exponential attenuation functions. The designed new function can attenuate to 0 at the specified time, thus ensuring the transition of sliding mode switching to the traditional linear switching surface in finite time, and effectively reducing the transition time on the basis of ensuring the global robustness of the system. Thus the dynamic response speed of the system can be accelerated and the convergence time can be shortened. An improved equivalent global sliding mode control law is designed for the n-order uncertain continuous system. The Lyapunov function is selected to prove the stability of the controller, and to verify the application of the proposed control strategy in practice. The improved global sliding mode control law is applied to chaos control of power system and synchronization control of uncertain delayed chaotic neuron / network and deterministic hysteresis chaotic neuron / network. The superiority of the proposed method compared with the conventional global sliding mode control method is verified by the comparative simulation experiments. Further exploration is made on the basis of the controlled object being a continuous system. The method proposed in this paper is applied to discrete time systems, and the attenuation function in discrete form is re-constructed, and an improved global sliding mode control law based on exponential discrete approach law is designed. The feasibility is also verified by simulation experiments.
【学位授予单位】：天津工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP273

【参考文献】