基于性能指标约束的一类输入死区非线性系统最优控制

针对一类考虑指定性能和带有输入死区约束的严格反馈非线性系统,本文提出了一种自适应模糊最优控制方法.采用模糊逻辑系统逼近系统的未知非线性函数及代价函数,利用backstepping方法及命令滤波技术,设计前馈控制器.针对仿射形式的误差系统,结合自适应动态规划技术,设计最优反馈控制器.采用指定性能控制方法,将系统跟踪误差约束在指定范围内.利用死区斜率信息解决具有死区输入的非线性系统的控制问题.基于Lyapunov稳定性理论,证明闭环系统内所有信号是一致最终有界的.最后仿真结果验证了本文方法的可行性和有效性.     

一类非线性动态系统基于强化学习的最优控制制

提出一类非线性不确定动态系统基于强化学习的最优控制方法. 该方法利用欧拉强化学习算法估计对象的未知非线性函数, 给出了强化学习中回报函数和策略函数迭代的在线学习规则. 通过采用向前欧拉差分迭代公式对学习过程中的时序误差进行离散化, 实现了对值函数的估计和控制策略的改进. 基于值函数的梯度值和时序误差指标值, 给出了该算法的步骤和误差估计定理. 小车爬山问题的仿真结果表明了所提出方法的有效性.

     

约束非线性系统多变量最优控制研究

近年来,非线性规划算法在最优控制领域中正受到越来越多的关注。该文深人研究并实现了一种新的非线性规划算法——FSQP算法,该算法具有所有迭代点均处于可行域之内、收敛速度较快的特点。提出了一种基于FSQP算法的约束非线性系统最优控制方法。然后,运用该方法解决了带有约束的复杂非线性系统的多变量时间最优控制问题,并通过计算机仿真表明了该控制算法的可行性和良好的控制效果。     

状态时滞系统的输入多采样最优控制

本文针对离散状态时滞系统，首先将其变形为无时滞形式，设计出最优控制器；然后运用离散提升技术对输入进行多采样，得到扩展的离散系统模型，再运用最优控制技术对扩展系统进行最优设计。最后对系统进行仿真，结果表明，该算法具有较好的控制效果。具有较好的稳定性。     

一类输入饱和非仿射非线性系统的非线性动态逆控制

输入饱和是实际系统中经常遇到的问题,很多已有的控制方法要求被控系统具有仿射结构.本文针对一类具有输入饱和的非仿射纯反馈非线性系统提出了一种基于奇异值摄动理论的非线性动态逆控制方法.首先构建一个快变子系统,在慢时间尺度下将非仿射非线性系统转换为具有仿射结构的线性系统,从而应用已有的控制算法实现控制目的.为了消除输入饱和带...     

基于强化学习的业务流程中的柔性约束研究

软件业务流程需要较高的灵活性和适应性。目前,众多的方法集中在流程建模阶段,通过一些方法提高流程的灵活性,却忽略了流程的运行阶段。由于外界环境的变化具有动态性,在建模阶段不易对其描述。由于外界环境的变化能够在运行阶段体现出来,分析流程的运行阶段是必要的。以流程中的约束为研究内容,通过对流程运行阶段的数据进行分析,提出一种基于强化学习的柔性约束模型,以提高流程的适应性。同时,将该算法应用于一类以用户为中心的复杂信息系统,实例分析表明算法是实用和有效的。     

具有不确定性的非线性切换系统的约束预测控制

针对一类具有不确定性和变量约束的非线性切换系统, 提出了一种基于Lyapunov函数的预测控制方法, 其中状态约束分为两种情况: 1)要求状态变量在所有时刻都满足约束(称为硬约束); 2)允许状态在某些时刻超出约束(称为软约束). 主要思想是: 对切换系统的每一个子系统, 在输入和状态均受约束的情况下, 设计基于Lyapunov函数的有界控制器和预测控制器, 在两者之间适当切换, 得到初始稳定区域的描述并使得子闭环系统保持稳定. 对整个切换系统, 设计适当的切换律以保证: 1)在切换时刻, 闭环系统的状态处在切入系统的稳定区域内; 2)切入模块的Lyapunov函数是非增的, 从而可保证稳定性. 在状态变量的约束是软约束时, 对每一子模块首先设计一个控制策略, 尽快将状态控制到初始稳定区域, 然后再利用稳定区域内的控制律使系统稳定.     

具有未知死区输入非线性系统的迭代学习控制

针对一类具有死区输入非线性系统,提出一种实现有限作业区间轨迹跟踪控制的神经网络迭代学习算法.基于Lyapunov-like方法设计学习控制器,回避了常规迭代学习控制中受控系统非线性特性需满足全局Lipschitz连续条件的要求.为处理输入死区,利用神经网络逼近这种强非线性特性;同时,通过对神经网络逼近误差界的估计并在控制器中设置补偿作用以消除其影响,从而提高系统的跟踪性能.     

一类脉冲切换系统的最优控制

研究一类脉冲依赖于状态的脉冲切换系统的最优控制问题.考虑了目标函数的两种情况:当目标函数光滑时,通过将跳跃瞬间转化为一个新的待优化参数,得到了该脉冲切换系统的必要最优性条件;当目标函数不光滑时,利用非光滑分析的知识,得到了广义微分形式的必要最优性条件.算例分析验证了所提出方法的有效性.     

一类脉冲切换系统的最优控制

研究一类脉冲依赖于状态的脉冲切换系统的最优控制问题. 考虑了目标函数的两种情况: 当目标函数光滑时, 通过将跳跃瞬间转化为一个新的待优化参数, 得到了该脉冲切换系统的必要最优性条件; 当目标函数不光滑时, 利用非光滑分析的知识, 得到了广义微分形式的必要最优性条件. 算例分析验证了所提出方法的有效性.     

Event-triggered-based integral reinforcement learning output feedback optimal control for partially unknown constrained-input nonlinear systems

In this paper, an adaptive output feedback event-triggered optimal control algorithm is proposed for partially unknown constrained-input continuous-time nonlinear systems. First, a neural network observer is constructed to estimate unmeasurable state. Next, an event-triggered condition is established, and only when the event-triggered condition is violated will the event be triggered and the state be sampled. Then, an event-triggered-based synchronous integral reinforcement learning (ET-SIRL) control algorithm with critic-actor neural networks (NNs) architecture is proposed to solve the event-triggered Hamilton–Jacobi–Bellman equation under the established event-triggered condition. The critic and actor NNs are used to approximate cost function and optimal event-triggered optimal control law, respectively. Meanwhile, the event-triggered-based closed-loop system state and all the neural network weight estimation errors are uniformly ultimately bounded proved by Lyapunov stability theory, and there is no Zeno behavior. Finally, two numerical examples are presented to show the effectiveness of the proposed ET-SIRL control algorithm.     

Optimized tracking control using reinforcement learning strategy for a class of nonlinear systems

This paper is to develop a simplified optimized tracking control using reinforcement learning (RL) strategy for a class of nonlinear systems. Since the nonlinear control gain function is considered in the system modeling, it is challenging to extend the existing RL-based optimal methods to the tracking control. The main reasons are that these methods' algorithm are very complex; meanwhile, they also require to meet some strict conditions. Different with these exiting RL-based optimal methods that derive the actor and critic training laws from the square of Bellman residual error, which is a complex function consisting of multiple nonlinear terms, the proposed optimized scheme derives the two RL training laws from negative gradient of a simple positive function, so that the algorithm can be significantly simplified. Moreover, the actor and critic in RL are constructed by employing neural network (NN) to approximate the solution of Hamilton–Jacobi–Bellman (HJB) equation. Finally, the feasibility of the proposed method is demonstrated in accordance with both Lyapunov stability theory and simulation example.     

Neural observer-based adaptive compensation control for nonlinear time-varying delays systems with input constraints

This paper describes a neural network state observer-based adaptive saturation compensation control for a class of time-varying delayed nonlinear systems with input constraints. An advantage of the presented study lies in that the state estimation problem for a class of uncertain systems with time-varying state delays and input saturation nonlinearities is handled by using the NNs learning process strategy, novel type Lyapunov-Krasovskii functional and the adaptive memoryless neural network observer. Furthermore, by utilizing the property of the function tan h²(?/?)/?, NNs compensation technique and backstepping method, an adaptive output feedback controller is constructed which not only efficiently avoids the problem of controller singularity and input saturation, but also can achieve the output tracking. And the proposed approach is obtained free of any restrictive assumptions on the delayed states and Lispchitz condition for the unknown nonlinear functions. The semiglobal uniform ultimate boundedness of all signals of the closed-loop systems and the convergence of tracking error to a small neighborhood are all rigorously proven based on the NN-basis function property, Lyapunov method and sliding model theory. Finally, two examples are simulated to confirm the effectiveness and applicability of the proposed approach.     

状态和输入受限的切换奇异布尔控制网络的最优控制

本文研究了状态和输入均受限的切换奇异布尔控制网络的最优控制问题.利用矩阵半张量积方法获得受限切换奇异布尔控制网络的等价代数形式.然后通过类似针变化得到了存在最优控制的必要条件,并且提出了一个算法设计切换序列和控制策略使收益函数最大化.最后给出例子验证所得结果的有效性.     

Robust control for a class of uncertain nonlinear systems with input quantization

In this paper, we propose a robust tracking control scheme for a class of uncertain strict‐feedback nonlinear systems. In these systems, the control signal is quantized by a class of sector‐bounded quantizers including the well‐known hysteresis quantizer and logarithmic quantizer. Compared with the existing results in input‐quantized control, the proposed scheme can control systems with non‐global Lipschitz nonlinearities and unmatched uncertainties caused by model uncertainties and external disturbances. It is shown that the designed robust controller ensures global boundedness of all the signals in the closed‐loop system and enables the tracking error to converge toward a residual, which can be made arbitrarily small. Copyright © 2015 John Wiley & Sons, Ltd.     

Robust adaptive control for a class of semi‐strict feedback systems with state and input constraints

This paper proposes a dynamic surface control (DSC)–based robust adaptive control scheme for a class of semi‐strict feedback systems with full‐state and input constraints. In the control scheme, a constraint transformation method is employed to prevent the transgression of the full‐state constraints. Specifically, the state constraints are firstly represented as the surface error constraints, then, an error transformation is introduced to convert the constrained surface errors into new equivalent variables without constraints. By ensuring the boundedness of the transformed variables, the violation of the state constraints can be prevented. Moreover, in order to obtain magnitude limited virtual control signal for the recursive design, the saturations are incorporated into the control law. The auxiliary design systems are constructed to analyze the effects of the introduced saturations and the input constraints. Rigorous theoretical analysis demonstrates that the proposed control law can guarantee all the closed‐loop signals are uniformly ultimately bounded, the tracking error converges to a small neighborhood of origin, and the full‐state constraints are not violated. Compared with the existing results, the key advantages of the proposed control scheme include: (i) the utilization of the constraint transformation can handle both time‐varying symmetric and asymmetric state constraints and static ones in a unified framework; (ii) the incorporation of the saturations permits the removal of a feasibility analysis step and avoids solving the constrained optimization problem; and (iii) the “explosion of complexity” in traditional backstepping design is avoided by using the DSC technique. Simulations are finally given to confirm the effectiveness of the proposed approach.     

Parameter governors for discrete-time nonlinear systems with pointwise-in-time state and control constraints

Parameter governors are add-on control schemes that adjust parameters (such as gains or offsets) in the nominal control laws to avoid violation of pointwise-in-time state and control constraints and to improve the overall system transient performance via the receding horizon minimization of a cost functional. As compared to more general model predictive controllers, parameter governors tend to be more conservative but the computational effort needed to implement them on-line can be relatively modest because the few parameters to be optimized remain constant over the prediction horizon. In this paper, we discuss the properties of several classes of parameter governors which have a common property in that the governed parameters do not shift the steady-state equilibrium of the states on which the incremental cost function explicitly depends on. This property facilitates the application of meaningful cost functionals. An example, together with simulation results, is reported to provide additional insights into the operation of the proposed parameter governor schemes.     

Optimal tracking control of nonlinear partially-unknown constrained-input systems using integral reinforcement learning

In this paper, a new formulation for the optimal tracking control problem (OTCP) of continuous-time nonlinear systems is presented. This formulation extends the integral reinforcement learning (IRL) technique, a method for solving optimal regulation problems, to learn the solution to the OTCP. Unlike existing solutions to the OTCP, the proposed method does not need to have or to identify knowledge of the system drift dynamics, and it also takes into account the input constraints a priori. An augmented system composed of the error system dynamics and the command generator dynamics is used to introduce a new nonquadratic discounted performance function for the OTCP. This encodes the input constrains into the optimization problem. A tracking Hamilton–Jacobi–Bellman (HJB) equation associated with this nonquadratic performance function is derived which gives the optimal control solution. An online IRL algorithm is presented to learn the solution to the tracking HJB equation without knowing the system drift dynamics. Convergence to a near-optimal control solution and stability of the whole system are shown under a persistence of excitation condition. Simulation examples are provided to show the effectiveness of the proposed method.     

Two-loop reinforcement learning algorithm for finite-horizon optimal control of continuous-time affine nonlinear systems

This article proposes three novel time-varying policy iteration algorithms for finite-horizon optimal control problem of continuous-time affine nonlinear systems. We first propose a model-based time-varying policy iteration algorithm. The method considers time-varying solutions to the Hamiltonian–Jacobi–Bellman equation for finite-horizon optimal control. Based on this algorithm, value function approximation is applied to the Bellman equation by establishing neural networks with time-varying weights. A novel update law for time-varying weights is put forward based on the idea of iterative learning control, which obtains optimal solutions more efficiently compared to previous works. Considering that system models may be unknown in real applications, we propose a partially model-free time-varying policy iteration algorithm that applies integral reinforcement learning to acquiring the time-varying value function. Moreover, analysis of convergence, stability, and optimality is provided for every algorithm. Finally, simulations for different cases are given to verify the convenience and effectiveness of the proposed algorithms.