网络诱导时延下遥操作系统的暂稳态性能约束控制研究

发布时间：2018-01-15 22:33

  本文关键词：网络诱导时延下遥操作系统的暂稳态性能约束控制研究　出处：《燕山大学》2016年博士论文　论文类型：学位论文

  更多相关文章： 非线性遥操作系统 网络诱导时延 暂稳态性能约束 有限时间控制 预定性能控制 终端滑模 障碍Lyapunov方程

【摘要】：遥操作系统作为能最大限度发挥人类和机械系统各自优势的远程操作系统,目前已被广泛应用于核事故救援、空间探测、海底作业和远程医疗等领域。然而,遥操作系统在实际应用中,一方面主从端之间信息交互不可避免地受到通信带宽的限制;另一方面主从系统模型不仅具有强烈的非线性特性,且易受到系统模型及参数不确定和外界工作环境扰动的影响。此外,日益复杂的工作任务以及不断扩大的应用范围,对遥操作系统控制性能提出了越来越高的要求。如何在复杂网络环境和恶劣工作环境下,构造新的控制策略使系统满足预设的暂稳态性能约束,从而满足系统对快速性、精确性、鲁棒性等多性能控制指标的要求,是本论文拟解决的主要问题。本文主要完善了非光滑控制以及预定性能控制理论,针对网络诱导时变时延下的遥操作系统首次提出了不依赖于时延精确信息的有限时间及预定性能控制策略。论文的主要研究内容如下:首先,针对网络诱导定常以及时变时延下遥操作系统的有限时间零误差稳态性能约束控制问题展开了深入的研究,通过设计新颖的非光滑控制策略,保证主从同步误差在有限时间内趋于零点。定常时延下,系统性地研究了遥操作系统在状态反馈以及输出反馈下基于终端滑模的有限时间控制策略的设计问题;构造了两种新的非奇异快速终端滑模面,解决了常规终端滑模控制中存在的奇异值问题;给出了新的快速有限时间速度观测器设计方法,改善了常规速度观测器观测速度慢且能力有限的问题。时变时延下,提出了新的快速有限时间参数自适应估计方法,实现了对系统不确定参数的有限时间精确估计;完成了不依赖于时延精确信息的有限时间控制策略的设计,进一步建立了控制器参数、时延变化率上界与系统稳定性的关系表达式,降低了系统的保守性。其次,针对网络诱导定常以及时变时延下遥操作系统的暂态以及有界误差稳态性能约束控制问题,提出了新的预定性能控制方案。定常时延下,基于反步递推策略设计了新的预定性能控制方法,保证遥操作系统渐近收敛的同时满足实际应用对系统超调量、收敛速度、收敛精度等暂稳态性能的要求;构造了新的自适应补偿方案,实现了执行器饱和下遥操作系统渐近收敛的同时满足预定的暂稳态性能要求。时变时延下,成功引入多维小增益定理,解决了时变时延下网络化遥操作系统不依赖于时延精确信息的预定性能控制器设计问题,给出了系统性能方程参数、控制器参数、时延变化率上界与系统稳定性的关系,为系统性能方程参数和控制器参数的选取提供了理论依据。最后,针对网络诱导定常时延下遥操作系统的暂态及有限时间零误差稳态性能约束控制问题,设计了终端滑模控制结合干扰观测器估计的新的有限时间控制策略。引入新的同步误差性能约束函数,构造了新的系统性能转换方程变量,设计了新的非奇异快速积分终端滑模面;提出了基于终端滑模思想的快速有限时间干扰观测器,其精确的估计能力有效抑制了滑模高频抖振现象的产生;基于障碍Lyapunov方程,构造了新的有限时间控制策略,实现主从机器人同步误差有限时间收敛至零点的同时,满足实际应用对系统超调量、收敛精度、收敛速度等多性能控制指标的要求。仿真结果验证了本论文所设计控制策略的合理性和有效性,对比仿真结果验证了所设计控制策略的优越性,实验结果验证了所设计控制策略的实用性。
[Abstract]:The remote operating system as a remote operating system to maximize the advantages of human and mechanical systems, has been widely used in nuclear accident rescue, space exploration, submarine operations and remote medical fields. However, the remote operating system in the practical application, on the one hand from the main communication bandwidth of information interaction between the inevitable end on the other hand; master-slave system model not only has strong nonlinear characteristics, and are vulnerable to the system model and parameter uncertainty and external disturbance environment. In addition, the increasingly complex tasks and expanding the scope of application, the control performance is put forward higher request. How to remote operating system in complex network the environment and poor working environment, the construction of a new control strategy to make the system meet the steady-state and transient performance constraint default, so as to meet the system of fast, accurate, The robustness of performance index of control requirements, is the main problem to be solved. This paper mainly improves the non smooth control and predetermined performance control theory for the network induced time-varying delay under the remote operating system is put forward for the first time limit time and the predetermined performance control strategy does not depend on the precise time delay information. The main research the contents of this paper are as follows: firstly, according to the finite time constant and time-varying network induced time delay. The operating system of zero error steady performance constrained control problem is studied deeply, by the design of a novel non smooth control strategy, ensure the master-slave synchronization error in finite time tends to zero. Constant time, systematic study on the remote operating system in the state feedback and output feedback is designed based on the finite time terminal sliding mode control strategy; the construction of two new nonsingular fast The terminal sliding surface, solves the problems of conventional terminal sliding mode control of singular value problem; give the design method of finite time fast speed observer, improved the conventional speed observer of slow speed and limited ability. When the variable time delay estimation method is proposed, finite time fast new adaptive parameter, to realize finite time the uncertain system parameters estimation; the control strategy design of finite time delay does not depend on the accurate information, to further establish the parameters of the controller, delay the rate of change of expression between the upper and the system stability, reduce the conservatism of the system. Secondly, according to the transient teleoperation system of network induced steady and time-varying time delay and bounded error of steady-state performance constrained control problem, put forward a new control scheme. The performance of a predetermined constant time, backstepping strategy based on The design of the control method of predetermined performance of the new remote operating system, to ensure the asymptotic convergence while satisfying the practical application to the system overshoot, convergence speed, convergence precision and transient performance requirements; to construct an adaptive new compensation scheme, realize the actuator saturation and the asymptotic convergence of the teleoperation system satisfies a predetermined temporary steady state performance requirements. Time delay, the successful introduction of multidimensional small gain theorem, solve the predetermined performance controller design problem of time-varying delay under networked remote operating system does not depend on the precise time delay information of the controller parameters is given equation parameters, system performance, relationship between the upper bound of the rate of delay variation and system stability, provide a theoretical basis for for the selection of system performance equation parameters and controller parameters. Finally, the network induced transient teleoperation system with fixed and finite time constant delay zero error steady performance The constrained control problem, design the terminal sliding mode control based on finite time disturbance observer to estimate the new control strategy is introduced. The synchronization error performance of new constraint function, constructing a system performance conversion equation of the variable, design a new nonsingular fast terminal sliding mode surface integral; put forward the finite time fast terminal sliding mode observer based on interference of thought the accurate estimation of ability, can effectively restrain the chattering of the sliding mode; obstacle based on Lyapunov model and construct new finite time control strategy, realize the master-slave robot finite time synchronization error converges to zero at the same time, to meet the practical application of the system overshoot, convergence precision, convergence speed and other performance index of control requirements. Simulation results verify the rationality and validity of the control strategy design of this paper, simulation results show that the designed control strategy The experimental results verify the practicability of the designed control strategy.

【学位授予单位】：燕山大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TP242;TP872
，

本文编号：1430393