具有噪声干扰的非线性时变系统多维泰勒网辨识和控制

针对含噪声不确定非线性时变系统,提出一种基于多维泰勒网(MTN)稳定的自适应控制方案,其中3个MTN分别被用来实现非线性滤波、系统辨识与自适应控制.首先, MTN滤波器(MTNF)用来消除测量噪声,以得到无随机干扰的模型输出.然后, MTN辨识器(MTNI)用来表示系统动态映射且比传统神经网络泛化能力更强.而后,MTN控制器(MTNC)用来实现系统精确跟踪控制,其中时变被控对象由MTNI辨识并将其动力学特性信息实时提供给MTNC使其"光滑"自适应.此外,利用改进的灵敏度计算方法来剪除MTNI和MTNC的冗余输入和冗余中间层回归项.最后,证明基于MTN的闭环系统稳定性,并给出最优学习率以期实现快速学习.仿真结果表明,该方法具有精确的辨识能力、良好的跟踪性能和较强的抗干扰能力,可实现含有不确定性、随机因素和时变特性的非线性系统自适应实时控制.     

一类非线性时滞输出反馈系统的自适应控制

针对一类参数化非线性时滞输出反馈系统,提出了一种无记忆自适应跟踪控制器的设计方案.采用时滞滤波器估计系统状态,用Domination处理非线性时滞项,应用Backstepping技术设计控制器和参数自适应律.放宽了对时滞项的要求.通过构建一个Lyapunov＿Krasoviskii泛函,证明了闭环系统的稳定性,实现了对目标轨线的渐近跟踪,保证了所有信号一致有界.实例仿真说明了该方案的可行性.     

基于多维泰勒网的多入多出非线性时滞系统辨识

针对多入多出 （multiple input multiple output,MIMO） 非线性时滞系统辨识的准确性与实时性问题,提出基于多维泰勒网（multi-dimensional taylor network,MTN） 的辨识方案。MTN作为辨识模型,综合利用权剪枝 （weight-elimination,WE） 算法和共轭梯度（conjugate gradient,CG）算法,即WE-CG算法作为MTN辨识模型的学习算法;WE算法可以有效精简MTN辨识模型结构,从而降低计算复杂度、提高模型实时性能。最后,引入一个数值仿真例子和一个工程实例来验证所提辨识方案的有效性,同时与传统的MTN辨识方案作对比,给出了两者的准确性与复杂度分析,突出所提辨识方案的准确性与实时性。实验结果表明,所提方案够准确地对MIMO非线性时滞系统进行辨识。同时,相比传统的MTN辨识方案,所提辨识方案结构更精简,具有更低的算法复杂度。     

基于有限时间的一类时滞非线性切换系统滑模控制

研究一类时滞非线性切换系统的有限时间滑模控制问题.针对所研究的系统模型,构造每个子系统对应的积分滑模面,基于滑模控制理论,设计带有状态时滞的滑模控制器使得每个子系统能在有限时间内到达相应的滑模面上,并对系统中存在的非线性项采用Lipschitz条件进行处理.根据多李亚普诺夫函数、平均驻留时间方法以及分割策略引理,给出滑模趋近段和滑模动态有限时间有界的充分条件,并通过对线性矩阵不等式的求解得到控制器增益.最后,通过一个数值仿真例子验证该设计方法的有效性.     

永磁同步电机调速系统的多维泰勒网逆控制

针对永磁同步电机(PMSM)的高性能控制问题,在充分考虑时变特性、不确定性以及测量噪声等随机因素的基础上,通过PMSM的逆系统将被控对象补偿成为具有线性传递关系的系统,提出一种基于改进自适应逆控制的控制方案.采用矢量控制的双闭环控制结构,将多维泰勒网逆控制方法引入速度环.首先,对PMSM数学模型的可逆性进行证明以解决非线性系统逆建模的存在性问题;然后,建立新颖的动态网络化控制器-----多维泰勒网(MTN),其具有结构简单、计算复杂度低的优点;最后,为了实现高精度的速度控制,将3个MTN分别作为实现系统建模的自适应模型辨识器、逆建模的自适应逆控制器和噪声干扰消除的非线性自适应滤波器,并将PMSM的动态响应控制和消除干扰的控制分为相对独立的过程进行,同时实现最优控制.仿真结果表明,所提出控制方案能够实现PMSM伺服系统精确的速度控制,具有良好的跟踪性能和较强的抗干扰能力.     

一类非线性大系统分散自适应预设性能有限时间跟踪控制

研究一类非线性互联大系统的分散自适应预设性能有限时间跟踪控制问题.结合神经网络自适应技术、实际有限时间控制理论和预设性能控制方法,提出一种新的预设性能控制设计方法,以解决传统预设性能方法难以实现分散控制的问题.所设计的控制器能够保证大系统中各个子系统的跟踪误差被有限时间性能函数约束,在任意给定的停息时间内收敛到平衡点的一个给定的邻域内,且该闭环大系统的所有信号是实际有限时间稳定的.特别地,该停息时间与系统初始状态无关.两个仿真例子验证了所提出控制方法的有效性和优越性.     

带有输出约束的水面船舶实际有限时间控制

考虑带有输出约束的水面船舶系统,提出一种自适应神经网络航迹跟踪实际有限时间控制算法.基于反步法设计有限时间控制律,构造障碍李雅普诺夫函数处理输出约束问题,采用神经网络逼近船舶模型中的不确定信息.在控制算法递推过程中,通过设计一个关于跟踪误差的可微幂函数来避免控制器中的奇异问题.借助李雅普诺夫稳定性分析理论,证明了航迹跟踪误差在有限时间内收敛到有界的邻域内.最后,以一艘1:70的比例模型船作为仿真对象,来验证所提出的航迹跟踪实际有限时间控制算法的有效性.     

一种具有噪声干扰的MIMO非线性时变系统自适应控制

针对含噪声多输入多输出不确定非线性时变系统,提出一种基于多维泰勒网(MTN)的自适应控制方案,其中两个MTN分别用来实现优化控制和非线性滤波.首先,提出多维泰勒网控制器(MTNC)以实现实时跟踪控制.将滤波输出与期望值之间的闭环误差作为MTNC的输入,根据系统不确定因素引起的误差,基于稳定的学习率,设计线性再励的自适应变步长算法以快速更新MTNC权值.其次,提出多维泰勒网滤波器(MTNF)以消除测量噪声.由于定义了测量值与MTNF输出之间误差的Lyapunov函数,自适应MTN滤波系统兼具基于Lyapunov理论的自适应滤波(LAF)和MTN的特有性质.最后,通过在Lyapunov意义下选取适当的权值更新律,可使MTNF输出渐近地收敛到期望信号,并证明了滤波器的收敛性和稳定性.仿真结果验证了所提出方案的有效性.     

基于多维泰勒网的运载火箭鲁棒自适应控制

针对运载火箭主动段飞行过程中的参数时变及存在弹性振动、液体推进剂晃动和风干扰等不确定性和非线性问题,提出了一种基于多维泰勒网(MTN)的直接鲁棒自适应控制方案.基于鲁棒自适应控制技术,通过多维泰勒网可以逼近任一非线性函数的性质,实时逼近时变参数模型的理想控制输出.对比通过实验的方式获得某些特征时刻运行参数,再根据稳定裕...     

一类非线性时滞系统自适应控制研究

针对一类非线性时滞系统，本文提出一种自适应控制器的设计方案，采用backstepping和domination方法构建了一个无记忆自适应控制器。放松了对非线性时滞函数的要求(例如全局Lipschitz条件)，实现了对给定目标轨线的全局渐近跟踪，保证了闭环系统所有信号全局一致有界：基于Lyapunov—Krasoviskii泛函方法证明了闭环系统的稳定性。仿真结果说明了这种控制方法的可行性和优点。     

Adaptive neural output feedback finite-time command filtered backstepping control for nonlinear systems with full-state constraints

In this paper, an adaptive neural finite-time control method via barrier Lyapunov function, command filtered backstepping, and output feedback is proposed to solve the tracking problem of uncertain high-order nonlinear systems with full-state constraints and input saturation. By utilizing the neural network (NN) to approximate unknown nonlinear functions, the finite-time command filters are used to filtering the virtual control signals and get the intermediate control signals in a finite time in the backstepping process. Because there are errors between the output of finite-time command filters and the virtual control signals, the error compensation signals are added to eliminate the influence of filtering errors. Based on the proposed control scheme, the states of the system can be constrained in the predetermined region, all signals in the system are bounded in finite time, and the tracking error can converge to the desired region in finite time. At last, a simulation example is given to show the effectiveness of the proposed control method.     

Global finite-time control for a class of switched nonlinear systems with different powers via output feedback

This paper is concerned with the problem of global finite-time control for switched nonlinear systems whose nonlinear terms satisfy homogenous growth conditions. At first, we design homogenous output feedback controllers for nominal switched nonlinear systems by adding a power integrator method. Then, we employ the homogeneous domination approach to scale homogeneous observers and controllers to render switched nonlinear systems with lower-triangular homogenous growth condition globally finite-time stable. Finally, the proposed control method can be extended to switched nonlinear systems satisfying with upper-triangular homogeneous growth condition. Two examples are provided to demonstrate the effectiveness of the proposed control scheme.     

Decentralised adaptive control of switched interconnected high-order nonlinear systems with unknown control direction and time-delay

This paper studies the dynamic state feedback control problem for a class of interconnected large-scale switched high-order nonlinear systems with unknown control direction and time-varying time-delay. The adaptive laws are designed to estimate the bounds of switched parameters under arbitrary switching for subsystems. The Nussbaum function is used to deal with the unknown control direction problem. By combining the backstepping and homogeneous domination technique, the decentralised adaptive control strategies are developed and the resulting closed-loop system is asymptotically stable. Finally, a simulation example is given and the results show the effectiveness of the proposed control design method.     

Observer-based adaptive control of uncertain time-delay switched systems with stuck actuator faults

This paper concerns the observer-based adaptive control problem of uncertain time-delay switched systems with stuck actuator faults. Under the case where the original controller cannot stabilize the faulty system, multiple adaptive controllers are designed and a suitable switching logic is incorporated to ensure the closed-loop system stability and state tracking. New delay-independent sufficient conditions for asymptotic stability are obtained in terms of linear matrix inequalities based on piecewise Lyapunov stability theory. On the other hand, adaptive laws for on-line updating of some of the controller parameters are also designed to compensate the effect of stuck failures. Finally, simulation results for reference [1] model show that the design is feasible and efficient.     

DDI-based finite-time stability analysis for nonlinear switched systems with time-varying delays

This paper investigates the finite-time stability (FTS) analysis problem for switched systems with both nonlinear perturbation and time-varying delays. For the system to be finite-time stable, a sufficient condition is proposed based on some delay differential inequalities (DDIs), rather than the Lyapunov-like functions which are commonly used in the FTS analysis of switched systems. Compared with the Lyapunov-like function method, the FTS conditions based on the DDI method are easier for checking and do not require FTS of each subsystem. Two examples are given to illustrate the effectiveness of the developed theory.     

Distributed fuzzy adaptive event-triggered finite-time consensus tracking control for uncertain nonlinear multi-agent systems with asymmetric output constraint

This article investigates the consensus problem for uncertain nonlinear multi-agent systems (MASs) with asymmetric output constraint. Different from BLF-based constraint consensus tracking control, a novel approach based on nonlinear state-dependent function is proposed to solve the asymmetric output constraint, which need not convert output constraint into tracking error bound. First-order sliding mode differentiator is incorporated into each step of backstepping control design to reduce computation burden. Further, in combination of proposed event-triggered mechanism based on time-varying threshold, a distributed fuzzy adaptive event-triggered finite-time consensus method is developed. It can ensure that the consensus tracking error tends to a small neighbor in a finite time and the asymmetric output constraint of each subsystem is not violated. Two simulations are given to demonstrate the effectiveness of control method.