【作者】 涂文桃；

【导师】 蔡国梁；

【作者基本信息】 江苏大学 ， 系统工程， 2006， 硕士

【摘要】 本文主要研究一类混沌系统的动力学分析及混沌控制问题，包含以下几个方面的内容：首先对国内外混沌研究的历史和现状进行了系统分析，并阐述了混沌控制的兴起、发展和未来的展望。其次，阐述了混沌的定义及其特性，以及本文采用的控制混沌的方法所涉及的理论基础，例如：Lyapunov稳定性原理、Lyapunov指数、Lyapunov函数等。第三，提出了一个新的混沌系统及其混沌吸引子，并分析了该混沌吸引子的动力学特性。第四，研究了混沌系统的自适应控制问题。阐述了Backstepping方法的基本原理，并将该方法应用于对参数未知的Liu混沌系统的控制。理论推导说明用Backstepping方法设计出的单控制器能使受控系统收敛到一有界点，而且设计的参数自适应律也能完成对系统未知参数的识别，Matlab数值仿真也证明了该方法的可行性和有效性。第五，研究了混沌系统Backstepping控制策略的改进问题。在指出了一种错误的Backstepping改进措施后，本文提出了一种新的改进方法。采用改进后的Backstepping控制方案，实现了不确定统一混沌系统的控制，并完成了对系统未知参数的辨识。最后通过数值仿真，详细研究了改进方案的参数变化对系统控制效果和参数辨识效果的影响。更多还原

【Abstract】 In this paper, the dynamic analysis and controlling of chaotic system are studied. The main content is depicted as follows:In the first part, the author analyzes systemically the history and international actualities of the research on chaos, and formulate the starting, developing and the future of this field.Secondly, the definition and the characteristic of chaos are presented. Then we formulate the theory of chaotic control which will be used in this paper, such as Lyapunov stability theory, Lyapunov exponent and Lyapunov function.Thirdly, a new chaotic system and its chaotic attractor are proposed. And the dynamic analysis of the system is following.Then, this paper investigates the method for controlling the uncertain Liu chaotic system with known parameters k and h via adaptive backstepping control. With the method, parameters identification and control can be achieved simultaneously and quickly with only one controller within finite steps. Numerical stimulations are provided to show the effectiveness and feasibility of this method.Finally, an improved adaptive backstepping design is proposed to control the uncertain unified chaotic system. At first, we prove that an improved adaptive backstepping design method with only one controller is unavailable. Then, we proposed another new backstepping design. With the improved method, parameters identification and control of the unified chaotic system can be achieved simultaneously with only one controller within finite steps. At last, numerical stimulations are provided to demonstrate the effect of the control and parameters identification with the improved method when the parameter is changing.更多还原