具有未建模动态的纯反馈系统自适应动态面控制研究

发布时间：2018-01-06 19:36

本文关键词：具有未建模动态的纯反馈系统自适应动态面控制研究　出处：《扬州大学》2017年硕士论文　论文类型：学位论文

【摘要】：实际的非线性系统往往存在系统函数未知、控制增益未知、未建模动态等多种不确定性,其中,未建模动态可能会破坏系统的稳定性条件,易造成系统不稳定,因而包含未建模动态的不确定非线性系统的自适应控制得到了人们的广泛研究.而具有纯反馈形式的非线性系统由于其系统状态与系统函数不可分离,因此相对于严格反馈形式,对这类系统的研究更为复杂.近年来,关于非线性控制系统的设计,在满足稳定性要求的前提下,进一步提出了误差约束或动态性能指标,要求跟踪误差满足常值约束或以一定的收敛率收敛到预设的小残差集内,也就是输出约束及预设性能控制.将这种约束施加到严格反馈或纯反馈系统的可量测状态上,则形成状态约束问题.对此,目前常用的解决方法是引入模型预估控制、误差转换、障碍Lyapunov函数等.针对这些自适应控制热点问题,本文利用径向基函数神经网络的逼近能力,将Lyapunov稳定性理论、自适应控制理论、动态面控制、分散控制技术有机结合,对几类带有未建模动态和状态或输出约束的非线性系统,分别提出了几种自适应控制方案,主要成果如下:第一,针对一类具有状态和输入未建模动态且控制增益符号未知的纯反馈非线性系统,利用非线性变换、改进的动态面控制方法和Nussbaum函数性质,提出两种自适应动态面控制方案.利用正则化信号来约束输入未建模动态,从而有效地抑制其产生的扰动.通过引入动态信号,有效地处理了由状态未建模动态引起的动态不确定性.通过在总的Lyapunov函数中加入非负正则化信号,并利用动态面控制证明的特点,有效地处理了控制信号的有界性.通过理论分析证明了闭环控制系统是半全局一致终结有界的.通过倒立摆模型及三阶系统的数值仿真验证了所提方案的有效性.第二,针对一类具有未建模动态、输出及状态约束的纯反馈非线性系统,通过非线性变换将纯反馈非线性系统转化为形式上的严格反馈非线性系统.利用改进的动态面控制方法,提出了一种新的自适应控制方案.利用动态信号处理未建模动态,在第一步及递推的每一步引入障碍Lyapunov函数(BLF),设计相应的虚拟控制及自适应控制律并保证闭环系统有界,从而实现输出及状态约束.数值仿真验证了所提方案的有效性.第三,针对一类具有状态未建模动态和输出约束的非线性纯反馈耦合大系统,通过利用改进的动态面控制方法,提出了一种分散自适应动态面控制方法.在递推设计的第一步引入了积分型障碍Lyapunov函数(iBLF),利用Young's不等式及分离定理,设计虚拟控制器.利用稳定性分析中引入的紧集及动态面控制中稳定性分析的特点,有效地处理了耦合作用项,证明了闭环系统内的信号是半全局一致终极有界且输出信号满足输出约束要求.数值仿真验证了所提方案的有效性.
[Abstract]:A practical nonlinear system often has the system function is unknown, unknown control gain, unmodeled dynamics and other uncertainties, the stability conditions of unmodeled dynamics may damage the system, causing system instability, which contains dynamic adaptive control of uncertain nonlinear systems have been widely studied in people. With pure feedback nonlinear system due to the system state and system function can not be separated, so relative to the strict feedback form, study on this kind of system is more complex. In recent years, the design of nonlinear control system, in order to meet the stability requirements, further put forward the error constraint or dynamic performance requirements the tracking error satisfies constant constraints or with a certain convergence rate of convergence to the preset small residual set, which is the output constraints and performance control. This will be about presupposition Beam is applied to the feedback or pure feedback systems can be measured on the state strictly, forming a state constraint problem. In this regard, the common approach is the introduction of model predictive control, error conversion, barrier Lyapunov function. To solve these problems by using the adaptive control of hot spots, approximation ability of RBF neural network, the stability of Lyapunov in theory, adaptive control theory, dynamic surface control, decentralized control techniques are combined for several classes of nonlinear systems with unmodeled dynamics and state or output constraints, respectively, puts forward several adaptive control schemes, the main results are as follows: first, the pure feedback for a class of nonlinear systems with state and input unmodeled dynamics and control gain sign unknown, nonlinear transform, dynamic surface modified method and Nussbaum function control, puts forward two kinds of adaptive dynamic surface control scheme. To constrain the input unmodeled dynamics of the regularized signals, so as to effectively suppress the disturbance. By introducing dynamic signals, effectively deal with the dynamic state caused by unmodeled dynamics uncertainty. By adding a non negative regularization function of the total signal in Lyapunov, and using the dynamic surface control proved effective, to deal with the control signals are bounded. It is proved that the closed-loop system is semi globally uniformly ultimately bounded. Numerical simulation of pendulum model and three order system by inverted verify the effectiveness of the proposed scheme. In second, for a class of unmodeled dynamics, nonlinear systems in pure feedback and output state constraints, through the nonlinear transform of pure feedback nonlinear system is transformed into strict feedback nonlinear systems form. Using the improved dynamic surface control method, this paper presents a new adaptive control The use of dynamic signal processing scheme. The unmodeled dynamics in the first step and recursive every step of introducing Lyapunov function disorder (BLF), the design of virtual control and adaptive control law and ensure the boundedness of the closed loop system, so as to realize the output and state constraints. Simulation results verify the effectiveness of the proposed scheme third. For a class of nonlinear state, unmodeled dynamics and output constraints in pure feedback coupling system, by using the improved dynamic surface control method, we propose a new adaptive dynamic surface control method. The Lyapunov integral type barrier function is introduced in the first step of recursive design (iBLF), by using the Young's inequality and the separation theorem design, virtual controller. Using the characteristics of compact set and stability analysis of dynamic surface control introduced in the stability analysis, can effectively deal with the coupling effect, proved the closed-loop system in signal system It is a semi global and uniform ultimate bounded and output signal satisfies the requirement of output constraints. Numerical simulation shows the effectiveness of the proposed scheme.

【学位授予单位】：扬州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP273

【参考文献】