LQ终端控制器设计的生成函数方法及其应用

利用哈密顿系统生成函数的性质求解LQ终端控制问题,并给出了相应的数值方法.针对现有文献中此类问题的最优控制律在终端时刻存在无穷大增益的情况,利用第二类生成函数的性质求解哈密顿系统两端边值问题并构造了无终端奇异性的时变最优控制律.然后根据哈密顿系统状态的正则变换性质导出了求解生成函数系数矩阵微分方程和计算时变控制律的矩阵递推格式.最后用所提出的方法研究了以能量均衡消耗为约束条件的卫星编队重构问题,设计了符合要求的闭环控制系统并给出了数值仿真结果.     

一类双线性随机系统M量测不定LQ控制

研究了带有乘性噪声和受扰动观测的离散时间随机系统不定线性二次(Linear quadratic, LQ) 最优输出反馈控制问题. 对此类问题而言,二次成本函数的加权矩阵不定号,并且最优控制具有对偶效果.为在最优性和计算复杂度间 进行折衷,本文采用了一种M量测反馈控制设计方法.基于动态规划方法,将未来的测量结合到当前控制 计算当中的M量测反馈控制可以通过倒向求解一类与原系统维数相同的广义差分Riccati方程(Generalized difference Riccati equation, GDRE)得到.仿真结果 表明本文提出的算法与目前普遍采用的确定等价性方法相比具有优越性.     

线性—二次型最优控制问题的Chebyshev—Legendre拟谱方法

介绍了一种求解线性—二次型最优控制问题的拟谱方法.使用Legendre展开式逼近控制和状态函数,采用Chebyshev-Gauss-Lobatto（CGL）点作为插值点,对原问题进行离散,从而将最初的最优控制问题化归为一个与之等价的二次规划（QP）问题,对应QP问题的未知量分别为状态和控制函数的Legendre展开式系数.通过求解QP问题得到原问题的数值解.整个离散过程使用快速Legendre变换（FLT）以及相关的一些技巧,能方便计算出函数在各个CGL点上的函数值.数值实验结果表明用该方法解决这类最优控制问题的有效性和高精度.     

基于动态规划的约束优化问题多参数规划求解方法及应用

结合动态规划和单步多参数二次规划, 提出一种新的约束优化控制问题多参数规划求解方法. 一方面能得到约束线性二次优化控制问题最优控制序列与状态之间的显式函数关系, 减少多参数规划问题求解的工作量; 另一方面能够同时求解得到状态反馈最优控制律. 应用本文提出的多参数二次规划求解方法, 建立无限时间约束优化问题状态反馈显式最优控制律. 针对电梯机械系统振动控制模型做了数值仿真计算.     

乘性随机离散系统的最优控制

基于对系统随机不确定因素的分析,文中定义了一种新型随机离散系统--乘性随机 离散系统,并研究该类系统的线性二次型(LQ)最优控制问题.首先给出了该类系统的有限时间 和无限时间LQ最优控制律,并着重分析、证明了无限时间LQ最优控制问题的Riccati方程的 正定矩阵解的存在性及相应数值求解算法与收敛性,以及闭环系统的稳定性等问题.仿真结果 表明了该方法的有效性.     

连续时间线性等式约束LQ控制的混合能消元算法

将连续时间LQ控制问题的微分方程离散化后,建立了连续时间线性等式约束LQ控制 问题的混合能消元算法,可有效求解约束条件下的Riccati方程.文中给出了相应的算例.     

控制时滞系统基于观测器的最优扰动抑制

研究在持续外界扰动作用下含有控制时滞线性系统的最优扰动抑制问题. 首先利用模型转换将控制时滞系统转化为无时滞系统. 然后证明最优控制律的存在唯一性, 并通过求解Riccati方程和Sylvester方程设计含前馈补偿器和控制记忆项的最优控制律, 其中的前馈控制项和控制记忆项分别补偿了扰动和控制时滞对系统的影响. 通过构造扰动状态观测器, 解决了前馈补偿器的物理不可实现问题. 仿真实例验证了所设计的最优控制律的有效性.     

一种非线性离散动态投入产出系统的最优控制

阐述了基于动态投入产出模型的最优控制理论,并对当前国内外的研究成果进行了对比研究。在分析其优缺点的同时,针对非线性离散动态投入产出系统的特点,提出了一种动态投入产出系统最优控制的逐次逼近方法。此方法首先将系统的最优控制问题转化为非线性两点边值问题族,然后通过构造线性两点边值问题族,将非线性两点边值问题转化为非齐次线性两点边值问题族;得到的最优控制律由精确控制项和非线性补偿项两部分组成,精确控制项可以通过求解Riccati方程求出其精确解,非线性补偿项由逐次逼近法求解一族线性伴随向量方程的解序列求得;最优控制律的最终目标是在规划期内使实际产出尽可能地与理想产出接近。实验仿真测试表明,采用逐次逼近法获得了非线性离散动态投入产出的最优系统控制,从而为最优控制问题的有效解决提供了参考和借鉴。     

基于两级算法的对偶控制

考虑具有未知参数的随机系统的最优控制问题.采用包含新息方差指数项的损失函数优化系统的性能.新的损失函数由两部分组成:第一部分反映了对输出的调节作用;第二部分反映了辩识系统中未知参数应尽可能多地收集系统信息的需求.提出了一种两级优化算法.该算法首先把不可分问题转化为可分的两目标优化问题,再从两目标优化的非劣解集中挑出原问题的最优解.该控制律易于实施且具有对偶特点.仿真结果表明本文所得的控制律的有效性.     

拟人智能控制及鲁棒LQ控制在倒立摆基准问题中的应用

分别应用拟人智能控制策略解决倒立摆标称系统的控制和鲁棒LQ方法解决其鲁棒控制问题.拟人智能控制模仿人解决问题的归约思路,从物理角度出发分析被控系统并设计定性控制律.利用遗传算法良好的全局搜索收敛特点,对定性控制律中的参数进行优化搜索.当模型只存在结构化型不确定性且不确定性有界时,可通过求解一个Riccati方程来设计鲁棒LQ控制器.仿真结果表明给定的控制指标均得到满足,且控制律算法简单,实现比较方便.     

Robust stabilization of stochastic systems based on the LQ controller

The robust exponential stability in mean square for a class of linear stochastic uncertain control systems is dealt with. For the uncertain stochastic systems, we have designed an optimal controller which guarantees the exponential stability of the system. Actually, we employed Lyapunov fimction approach and the stochastic algebraic Riccati equation (SARE) to have shown the robusmess of the linear quadratic(LQ) optimal control law.And the algebraic criteria for the exponential stability on the linear stochastic uncertain closed-loop systems are given.     

Robust stabilization of stochastic systems based on the LQ cont roller

The robust exponential stability in mean square for a class of linear stochastic uncertain control systems is dealt with. For the uncertain stochastic systems ,we have designed an optimal controller which guarantees the exponential stability of the system. Actually ,we employed Lyapunov function approach and the stochastic algebraic Riccati equation (SARE) to have shown the robustness of the linear quadratic (LQ) optimal control law. And the algebraic criteria for the exponential stability on the linear stochastic uncertain closed- loop systems are given.     

Nonlinear optimal control as quantum mechanical eigenvalue problems

A novel approach for approximating the nonlinear optimal feedback control of a system with a terminal cost is proposed. To lessen the difficulty due to nonlinearity, we try to treat the system in a framework of linear theories. For this, we assume a quantum mechanical linear wave associated with the system. Since the control system is constrained by state equations, we handle the system according to quantum mechanics of constrained dynamics. A Hamiltonian is represented as a linear operator acting on a function that describes behavior of waves. Subsequently, nonlinear feedback is calculated without any time integration in the backward direction. Using eigenvalues and eigenfunctions of the linear Hamiltonian operator, an optimal feedback law is given as a combination of analytic functions of time and state variables. We take as an example a system described by two scalar variables for state and control input. Simulation studies on the system by the eigenvalue analysis show that the proposed method reduces calculation time to nearly a tenth that of a numerical calculation of a Hamilton-Jacobi equation by a finite difference method.     

Determination of optimal feedback terminal controllers for general boundary conditions using generating functions

Given a nonlinear system and a performance index to be minimized, we present a general approach to expressing the finite time optimal feedback control law applicable to different types of boundary conditions. Starting from the necessary conditions for optimality represented by a Hamiltonian system, we solve the Hamilton-Jacobi equation for a generating function for a specific canonical transformation. This enables us to obtain the optimal feedback control for fundamentally different sets of boundary conditions only using a series of algebraic manipulations and partial differentiations. Furthermore, the proposed approach reveals an insight that the optimal cost functions for a given dynamical system can be decomposed into a single generating function that is only a function of the dynamics plus a term representing the boundary conditions. This result is formalized as a theorem. The whole procedure provides an advantage over methods rooted in dynamic programming, which require one to solve the Hamilton-Jacobi-Bellman equation repetitively for each type of boundary condition. The cost of this favorable versatility is doubling the dimension of the partial differential equation to be solved.     

Solvability of indefinite stochastic Riccati equations and linear quadratic optimal control problems

A new approach to study the indefinite stochastic linear quadratic (LQ) optimal control problems, which we called the “equivalent cost functional method”, is introduced by Yu (2013) in the setup of Hamiltonian system. On the other hand, another important issue along this research direction, is the possible state feedback representation of optimal control and the solvability of associated indefinite stochastic Riccati equations. As the response, this paper continues to develop the equivalent cost functional method by extending it to the Riccati equation setup. Our analysis is featured by its introduction of some equivalent cost functionals which enable us to have the bridge between the indefinite and positive-definite stochastic LQ problems. With such bridge, some solvability relation between the indefinite and positive-definite Riccati equations is further characterized. It is remarkable the solvability of the former is rather complicated than the latter, hence our relation provides some alternative but useful viewpoint. Consequently, the corresponding indefinite linear quadratic problem is discussed for which the unique optimal control is derived in terms of state feedback via the solution of the Riccati equation. In addition, some example is studied using our theoretical results.     

Linear‐Quadratic Optimal Control Problem for Partially Observed Forward‐Backward Stochastic Differential Equations of Mean‐Field Type

This paper is concerned with the linear‐quadratic optimal control problem for partially observed forward‐backward stochastic differential equations (FBSDEs) of mean‐field type. Based on the classical spike variational method, backward separation approach as well as filtering technique, we first derive the necessary and sufficient conditions of the optimal control problem with the non‐convex domain. Nextly, by means of the decoupling technique, we obtain two Riccati equations, which are uniquely solvable under certain conditions. Also, the optimal cost functional is represented by the solutions of the Riccati equations for the special case.     

Employing the algebraic Riccati equation for a parametrization of the solutions of the finite-horizon LQ problem: the discrete-time case

In this paper, a new methodology is developed for the closed-form solution of a generalized version of the finite-horizon linear-quadratic regulator problem for LTI discrete-time systems. The problem considered herein encompasses the classical version of the LQ problem with assigned initial state and weighted terminal state, as well as the so-called fixed-end point version, in which both the initial and the terminal states are sharply assigned. The present approach is based on a parametrization of all the solutions of the extended symplectic system. In this way, closed-form expressions for the optimal state trajectory and control law may be determined in terms of the boundary conditions. By taking advantage of standard software routines for the solution of the algebraic Riccati and Stein equations, our results lead to a simple and computationally attractive approach for the solution of the considered optimal control problem without the need of iterating the Riccati difference equation.     

迭代学习控制在线性时变系统终端控制问题中的应用研究

An iterative learning control algorithm based on shifted Legendre orthogonal polynomials is proposed to address the terminal control problem of linear time-varying systems. First, the method parameterizes a linear time-varying system by using shifted Legendre polynomials approximation. Then, an approximated model for the linear time-varying system is deduced by employing the orthogonality relations and boundary values of shifted Legendre polynomials. Based on the model, the shifted Legendre polynomials coefficients of control function are iteratively adjusted by an optimal iterative learning law derived. The algorithm presented can avoid solving the state transfer matrix of linear time-varying systems. Simulation results illustrate the effectiveness of the proposed method.