随机线性二次最优控制在投资组合中的应用

提出一类随机线性二次最优控制问题,给出了一个新的随机黎卡提方程,若此方程有解,就可以得到系统的最优反馈控制;作为其应用,讨论了连续时间的均值-方差投资组合选择问题,其目标是投资组合的最终收益最大,风险最小,通过"嵌入"方法将其转化为随机线性二次最优控制问题,并在非自融资的条件下,得出最优证券组合;最后将其理论应用于实例分析.     

Stochastic maximum principle for mixed regular-singular control problems of forward-backward systems

This paper considers a stochastic optimal control problem of a forward-backward system with regular-singular controls where the set of regular controls is not necessarily convex and the regular control enters the diffusion coefficient. This control problem is difficult to solve with the classical method of spike variation. The authors use the approach of relaxed controls to establish maximum principle for this stochastic optimal control problem. Sufficient optimality conditions are also investigated.     

State Estimation of Discrete Stochastic Singular System

1　 IntroductionIn the area of singular systems control much work has been done recently. The basictheory of singular systems has been proposed. A lot of results are obtained fordeterministic singluar systems[1～ 2 ] . However,only a few results are from stochasticsingular systems[3～ 5] ,in which attention is paid to time-invariant stochastic singularsystems. The state estimation and observation problems of time-invariant stochasticsingular systems have been discussed in[3 ] .The state estim…     

常弹性方差模型下保险人的最优投资策略

假设风险资产价格服从常弹性方差(CEV)模型, 保险人面临的风险过程是带漂移的布朗运动. 投资过程与承保风险过程完全相关. 根据随机最优控制理论, 建立保险基金投资问题的HJB方程. 由于该方程是非线性偏微分方程, 不易求解, 因此采用Legendre变换将其转换成对偶问题进行研究. 最后针对特定参数值分别得到以CARA和CRRA效用函数为目标的保险人的最优投资策略, 这样的投资策略更符合金融市场的实际要求.     

满足圆形极点、方差指标与H∞性能约束的满意容错控制

对一类具有范数有界不确定性的线性连续系统,研究了当执行器发生故障时的静态输出反馈满意容错控制问题。在更实际、更一般的执行器连续增益故障模型下,所设计的控制器确保闭环系统同时满足期望的区域极点指标、稳态方差指标和H∞范数指标约束。同时分析了容错控制中三类期望指标的相容性,并给出了静态输出反馈满意容错控制器存在的充分条件和设计方法。仿真算例验证了方法的有效性,并且通过与不考虑故障的系统比较,进一步说明对系统进行满意容错控制的必要性。     

电子商务系统中长鞭效应的随机控制

在电子商务系统中,高效、集成的物流配送体系是经贸交易成功实施的重要保证。对一个具有单个配送中心和多个顾客,且只经营一种产品的物流配送体系的长鞭效应问题进行了研究。给出了电子商务下的物流配送模型,它是以库存作为状态变量,以订货量作为控制变量,以需求的不确定需求作为随机扰动的动态系统。并利用状态、控制变量和需求波动的偏差值对系统中的长鞭效应进行了定量化的描述。应用随机控制理论方法,设计了抑制长鞭效应的随机控制策略。最后,以一个石化企业的电子商务系统为对象,进行了抑制长鞭效应的随机控制仿真实验,结果表明电子商务系统的长鞭效应得到了抑制。     

Stabilization of Continuous-Time Systems Against Stochastic Network Uncertainties: Fundamental Variance Bounds

This paper studies the stabilizability and stabilization of continuous-time systems in the presence of stochastic multiplicative uncertainties. The authors consider multi-input, multi-output(MIMO) linear time-invariant systems subject to multiple static, structured stochastic uncertainties,and seek to derive fundamental conditions to ensure that a system can be stabilized under a mean-square criterion. In the stochastic control framework, this problem can be considered as one of optimal control under state-or input-dependent random noises, while in the networked control setting, a problem of networked feedback stabilization over lossy communication channels. The authors adopt a mean-square small gain analysis approach, and obtain necessary and sufficient conditions for a system to be meansquare stabilizable via output feedback. For single-input, single-output(SISO) systems, the condition provides an analytical bound, demonstrating explicitly how plant unstable poles, nonminimum phase zeros, and time delay may impose a limit on the uncertainty variance required for mean-square stabilization. For MIMO minimum phase systems with possible delays, the condition amounts to solving a generalized eigenvalue problem, readily solvable using linear matrix inequality optimization techniques.     

参数未知随机系统的最小方差对偶自适应控制

针对参数未知随机系统的自适应控制问题,研究了最小方差对偶自适应控制。分析了由于未知参数的不确定性,使得运用动态规划原理求解最优控制律时存在着困难,转而寻求次优控制律;给出了差分方程转换成状态空间模型的方法,将未知参数在一个确定的模型集中取值,运用动态规划原理得到各模型的控制律,通过各模型后验概率加权获得次优对偶控制律。给出了算例,以验证此算法的有效性,表明所得到的控制律既有调节作用,又有学习作用。     

FORWARD-BACKWARD STOCHASTIC DIFFERENTIAL EQUATIONS, LINEAR QUADRATIC STOCHASTIC OPTIMAL CONTROL AND NONZERO SUM DIFFERENTIAL GAMES

In this paper, we use the solutions of forward-backward stochastic differential equations to get the explicit form of the optimal control for linear quadratic stochastic optimal control problem and the open-loop Nash equilibrium point for nonzero sum differential games problem. We also discuss the solvability of the generalized Riccati equation system and give the linear feedback regulator for the optimal control problem using the solution of this kind of Riccati equation system.     

基于限价订单簿的最优高频做市商策略研究

建立具有成交风险和存货风险的价差过程模型,在引入存货惩罚函数的同时将策略的目标确定为效用最大化.将策略求解的过程看成是随机最优控制问题,并通过动态规划求解,离散模型框架下采用有限差分的方法对每个时间点不同存货及市场价差下的下单策略进行求解.该策略满足了模型定义之初对于成交强度,市场价差及存货量对下单行为影响的假设,而策略的实证及可靠性检验进一步表明了该策略具有较为稳定的收益.     

Robust H∞ control for uncertain stochastic saturating systems with time delays

The robust H∞ control problem for uncertain stochastic time-delay systems containing nonlinear actuators is considered. The uncertainties in the systems are assumed to satisfy specific match condition. The time delays exitst in state as well as control input. The new stochastic robust stabilization criterion and a sufficient condition for the existence of stochastic robust stabilizing control law are derived. The delay-independent memoryless robust H∞ controllers are constructed to stabilize the given systems in terms of a group of linear matrix inequalities. A numerical simulation example is presented to show that the proposed approach is valid.     

CEV模型下基于双曲绝对风险厌恶效用的最优投资策略

研究双曲绝对风险厌恶(HARA)型投资者在常弹性方差(CEV)模型下面临完全可对冲随机资金流时的最优动态资产配置问题.随机资金流可以视作一个外生负债,假定其服从带漂移的布朗运动.根据随机控制理论建立该问题的哈密顿-雅克比-贝尔曼(HJB)方程,通过猜测值函数的代数形式,将其化简为两个抛物型偏微分方程并分别求得显式解,从而得到最优投资策略.结果表明该非自融资组合的最优动态配置问题等价于初始财富为所有未来随机净资金流在风险中性测度下累积期望现值与初始稟赋之和的自融资组合的最优动态配置问题.投资策略由短视投资策略,动态对冲策略,静态对冲策略三部分组成.当对模型中参数取特殊值时,策略简化为已有文献的相应结果.最后分析了参数变化对于由随机资金流引起的额外投资需求的影响.     

On the Solutions of the Matrix Equations in Optimal Stochastic Control

1.INTRoDUCTIoNoFTHEPRoBLEMConsiderthetime-invariantlinearIt6stochasticsystemwherexER,uCRm,FER""",BER""m,a=const.(i=1,2,'',k)arenaturalnumbers;W(t)=[W(t),W2(t),Wk(t)]"(tSo)isak-dimensionalstandardWienerprocesswithindependentcomponentsdefinedonthecompleteprobabilityspace(fl,FP),(A,B)iscontrollable.SupposethatRERm"m,QERmxmaregivenpositivedefiniteweightingmatrices,Ereferstothemathematicalexpectation.ConsidertheoptimalstochasticcontrolDroblemItiswell-knownthatifthereisapositive…     

基于模糊模型的非线性系统鲁棒控制方差

针对T S模糊模型描述的一类不确定非线性连续系统,研究使得闭环系统的稳态方差小于某个给定上界,同时闭环极点位于给定圆盘中的状态反馈鲁棒方差控制律设计问题。导出了以一组线性矩阵不等式表示控制律的存在条件。通过一个质量弹簧阻尼控制系统验证了这种方法的有效性。     

The Optimal Control for the Output Feedback Stochastic System at the Risk-Sensitive Cost

Abstract: The optimal control of the partially observable stochastic system at the risk-sensitive cost is considered in thispaper. The system dynamics has a general correlation between system and measurement noise. And the risk-sensitive costcontains a general quadratic term (with cross terms and extra linear terms). The explicit solution of such a problem ispresented here using the output feedback control method. This clean and direct derivation enables one to convert suchpartial observable problems into the equivalent complete observable control problems and use the routine ways to solvethem.     

Global Convergence of Adaptive Generalized Predictive Controller Based on Least Squares Algorithm

Some papers on stochastic adaptive control schemes have established convergence algorithm using a least-squares parameters. With the popular application of GPC, global convergence has become a key problem in automatic control theory. However, now global convergence of GPC has not been established for algorithms in computing a least squares iteration. A generalized model of adaptive generalized predictive control is presented. The global convergebce is also given on the basis of estimating the parameters of GPC by least squares algorithm.     

基于SVRGD的机载预警雷达自适应波束形成算法

自适应波束形成是机载预警雷达数字信号处理的一个关键环节。针对传统最小均方误差(least mean square, LMS)算法在短快拍数条件下的波束形成性能下降以及因迭代震荡易收敛于局部最优值的问题,提出了一种基于机器学习的随机方差减小梯度下降(stochastic variance reduction gradient descent, SVRGD)自适应波束形成方法。首先,建立面阵列接收信号数据模型。其次,基于随机梯度下降原理,引入方差缩减法通过内外循环迭代方式进行梯度修正,以减小随机梯度估计的方差,建立算法模型与实现流程。最后，通过设置平面阵列仿真场景,分析SVRGD自适应波束形成算法在波束形成、抗干扰、收敛速度等方面的性能,验证了该算法在低快拍数、强干扰和强噪声背景下具有的优良能力。     

Robust H∞ control for neutral stochastic uncertain systems with time-varying delay

The problem of robust H∞ control for uncertain neutral stochastic systems with time-varying delay is discussed.The parameter uncertaintie is assumed to be time varying norm-bounded.First,the stochastic robust stabilization of the stochastic system without disturbance input is investigated by nonlinear matrix inequality method.Then,a full-order stochastic dynamic output feedback controller is designed by solving a bilinear matrix inequality(BMI),which ensures a prescribed stochastic robust H∞ performance level for the resulting closed-loop system with nonzero disturbance input and for all admissible uncertainties.An illustrative example is provided to show the feasibility of the controller and the potential of the proposed technique.     

Observer-based H∞ control over packet dropping networks

A new controller design problem of networked control systems with packet dropping is proposed. Depending on the place that the observer is put in the system, the network control systems with packet dropping are modeled as stochastic systems with the random variables satisfying the Bernoulli random binary distribution. The observer-based controller is designed to stabilize the networked system in the sense of mean square, and the prescribed H∞ disturbance attenuation level is achieved. The controller design problem is formulated as the feasibility of the convex optimization problem, which can be solved by a linear matrix inequality (LMI) approach. Numerical examples illustrate the effectiveness of the proposed methods.     

Research on robust mean square stability of networked control systems with packet dropout

This paper is concerned with the robust stabilization problem of networked control systems with stochastic packet dropouts and uncertain parameters. Considering the stochastic packet dropout occuring in two channels between the sensor and the controller, and between the controller and the actuator, networked control systems are modeled as the Markovian jump linear system with four operation modes. Based on this model, the necessary and sufficient conditions for the mean square stability of the deterministic...