Stochastic adaptive control for continuous-time linear systems with quadratic cost

An adaptive control problem is formulated and solved for a completely observed, continuous-time, linear stochastic system with an ergodic quadratic cost criterion. The linear transformationsA of the state,B of the control, andC of the noise are assumed to be unknown. Assuming only thatA is stable and that the pair (A, C) is controllable and using a diminishing excitation control that is asymptotically negligible for an ergodic, quadratic cost criterion it is shown that a family of least-squares estimates is strongly consistent. Furthermore, an adaptive control is given using switchings that is self-optimizing for an ergodic, quadratic cost criterion.This research was partially supported b y NSF Grants ECS-9102714, ECS-9113029, and DMS-9305936.     

Optimal NPID Stabilization of Linear Systems

We consider the problem of finding the optimal, robust stabilization of linear systems within a family of nonlinear feedback laws. Investigation of the efficiency of full-state based and partial-state based so-called NPID feedback schemes proposed for the stabilization of systems in robotic applications has provided the motivation for our work. We prove that, for a given quadratic Lyapunov function and a given family of nonlinear feedback laws, there exist optimal piecewise linear feedbacks that make the generalized Lyapunov derivative of the closed-loop system minimal. The result provides improved stabilization over the nonlinear stabilizing feedback law proposed in Ref. 1 as demonstrated in simulations of the Sarcos Dextrous Manipulator.     

Adaptive control of linear stochastic evolution systems

An adaptive control problem for some linear stochastic evolution systems in Hilbert spaces is formulated and solved in this paper. The solution includes showing the strong consistency of a family of least squares estimates of the unknown parameters and the convergence of the average quadratic costs with a control based on these estimates to the optimal average cost. The unknown parameters in the model appear affinely in the infinitesimal generator of the C ₀ semigroup that defines the evolution system. A recursive equation is given for a family of least squares estimates and the bounded linear operator solution of the stationary Riccati equation is shown to be a continuous function of the unknown parameters in the uniform operator topology     

Adaptive boundary control of stochastic linear distributed parameter systems described by analytic semigroups

A stochastic adaptive control problem is formulated and solved for some unknown linear, stochastic distributed parameter systems that are described by analytic semigroups. The control occurs on the boundary. The highest-order operator is assumed to be known but the lower-order operators contain unknown parameters. Furthermore, the linear operators of the state and the control on the boundary contain unknown parameters. The noise in the system is a cylindrical white Gaussian noise. The performance measure is an ergodic, quadratic cost functional. For the identification of the unknown parameters a diminishing excitation is used that has no effect on the ergodic cost functional but ensures sufficient excitation for strong consistency. The adaptive control is the certainty equivalence control for the ergodic, quadratic cost functional with switchings to the zero control.This research was partially supported by NSF Grants ECS-9102714, ECS-9113029, and DMS-9305936.     

Global asymptotical synchronization of chaotic neural networks by output feedback impulsive control: An LMI approach

In this paper, impulsive control for master–slave synchronization schemes consisting of identical chaotic neural networks is studied. Impulsive control laws are derived based on linear static output feedback. A sufficient condition for global asymptotic synchronization of master–slave chaotic neural networks via output feedback impulsive control is established, in which synchronization is proven in terms of the synchronization errors between the full state vectors. An LMI-based approach for designing linear static output feedback impulsive control laws to globally asymptotically synchronize chaotic neural networks is discussed. With the help of LMI solvers, linear output feedback impulsive controllers can be easily obtained along with the bounds of the impulsive intervals for global asymptotic synchronization. The method is finally illustrated by numerical simulations.     

Adaptive control of a continuous-time portfolio and consumption model

In this paper, an adaptive control problem is formulated and solved using Merton's stochastic differential equation for the wealth in a portfolio selection and consumption model. Since the asset prices are assumed to satisfy a log normal distribution, it suffices to consider two assets. It is assumed that the drift parameter for the price of the risky asset is unknown. A recursive family of estimators for this unknown parameter is defined and is shown to converge almost surely to the true value of the parameter. The controls in the equation for the wealth are obtained from the optimal controls where the estimates of the unknown parameter are substituted for the unknown parameter.This research was partially supported by NSF Grant No. ECS-84-03286-A01.The authors wish to thank P. Varaiya for some useful comments on this paper.     

Classification of Local Conservation Laws of Maxwell's Equations

A complete and explicit classification of all independent local conservation laws of Maxwell's equations in four dimensional Minkowski space is given. Besides the elementary linear conservation laws, and the well-known quadratic conservation laws associated to the conserved stress-energy and zilch tensors, there are also chiral quadratic conservation laws which are associated to a new conserved tensor. The chiral conservation laws possess odd parity under the electric–magnetic duality transformation of Maxwell's equations, in contrast to the even parity of the stress-energy and zilch conservation laws. The main result of the classification establishes that every local conservation law of Maxwell's equations is equivalent to a linear combination of the elementary conservation laws, the stress-energy and zilch conservation laws, the chiral conservation laws, and their higher order extensions obtained by replacing the electromagnetic field tensor by its repeated Lie derivatives with respect to the conformal Killing vectors on Minkowski space. The classification is based on spinorial methods and provides a direct, unified characterization of the conservation laws in terms of Killing spinors.     

Feedback stabilization and adaptive stabilization of stochastic nonlinear systems by the control Lyapunov function

Our aims of this paper are twofold: On one hand, we study the asymptotic stability in probability of stochastic differential system, when both the drift and diffusion terms are affine in the control. We derive sufficient conditions for the existence of control Lyapunov functions (CLFs) leading to the existence of stabilizing feedback laws which are smooth, except possibly at the equilibrium state. On the other hand, we consider the previous systems with an unknown constant parameters in the drift and introduce the concept of an adaptive CLF for stochastic system and use the stochastic version of Florchinger's control law to design an adaptive controller. In this framework, the problem of adaptive stabilization of nonlinear stochastic system is reduced to the problem of non-adaptive stabilization of a modified system.     

Quadratic stabilization of a class of switched nonlinear systems via single Lyapunov function

This paper is concerned with the problem of globally quadratic stabilization for a class of switched cascade systems. The system under consideration is composed of two subsystems: a linear switched part and a nonlinear part, which are also switched systems. The feedback control law and the switching law are designed respectively when the first part is stabilized under some switching law and when both parts can be stabilized under some switching laws. We construct the single Lyapunov functions and design the switching laws based on the structure characteristics of the switched system. Also, the designed switching laws are of hysteresis switching form in order to avoid sliding models.     

Controlling the nonlinear dynamics of a beam system

Control based on linear error feedback is applied to reduce vibration amplitudes in a piecewise linear beam system. Hereto small amplitude 1-periodic solutions are stabilized wherever they coexist with two or more long-term solutions. In theory, no control effort is required to maintain the 1-periodic response once it has been stabilized. For the beam system, 1-periodic solutions are stabilized by feedback at one location along the beam. Feedback is represented by servo-stiffness or servo-damping which results from increasing two corresponding control parameters. At appropriate levels of these parameters local, or global, asymptotic stability (of the zero-equilibrium) of the error dynamics, i.e. stability of the underlying 1-periodic solutions, can be guaranteed. Local asymptotic stability can be guaranteed for a large range of actuator locations and excitation frequencies and is indicated by bifurcations. Global asymptotic stability can only be guaranteed for a limited range of actuator locations on the basis of the well-known circle criterion. The difference between local and global asymptotic stability in terms of the required values for the control parameters can be significant, and may result in large differences in control performance.     

Adaptive Linear-Quadratic Control for Stochastic Discrete-Time Systems

The optimality of certain adaptive control laws for partiallyobserved linear stochastic systems with unknown parameters andwith averaged quadratic-loss functions of the state and controlvalues is demonstrated. First, an explicit formula for the lossfunction in terms of system parameters, control actions, and the solutionof the Riccati equation is given; then the optimal value forthis loss function and an associated set of optimal controlsare derived using certain stability, observability, and controllabilityconditions on the system. Second, assuming strongly consistentparameter estimates are available and that the noise part ofthe system is minimum-phase, an adaptive control law is presentedthat with arbitrarily high probability achieves a loss arbitrarilyclose to the optimum.     

Feedback law design and the Liapunov approach

We study the problem of designing targeting-type feedback control laws for systems with state and control constraints. A systematic approach to this problem is described in terms of a number of Liapunov-like criteria. This approach is presented in tutorial form and is illustrated by means of two examples having some background in the literature.This work was supported by the National Science Foundation under Grant ECS-8210284.     

Quadratic integral games and causal synthesis

The game problem for an input-output system governed by a Volterra integral equation with respect to a quadratic performance functional is an untouched open problem. In this paper, it is studied by a new approach called projection causality. The main result is the causal synthesis which provides a causal feedback implementation of the optimal strategies in the saddle point sense. The linear feedback operator is determined by the solution of a Fredholm integral operator equation, which is independent of data functions and control functions. Two application examples are included. The first one is quadratic differential games of a linear system with arbitrary finite delays in the state variable and control variables. The second is the standard linear-quadratic differential games, for which it is proved that the causal synthesis can be reduced to a known result where the feedback operator is determined by the solution of a differential Riccati operator equation.

     

随机线性系统保成本控制的线性矩阵不等式方法

考虑具有二次成本函数的随机线性系统,研究了状态反馈控制的保证成本控制问题.依据线性矩阵不等式得到了保证成本控制器存在的充分条件,最后得到了随机线性闭环系统保证成本最小的最优保证成本控制律的表达式.     

Quadratic stabilizability of uncertain linear systems containing both constant and time-varying uncertain parameters

This paper is concerned with the problem of stabilizing an uncertain linear system using state feedback control. The uncertain systems under consideration are described by state equations containing unknown but bounded uncertain parameters. The uncertain parameters are classified into two types: either constant or time-varying. Indeed, the main feature of this paper is that it allows one to exploit the fact that some of the uncertain parameters are constant. In order to investigate the question of stabilizability, quadratic Lyapunov functions are used. Hence, the paper deals with the notion of quadratic stabilizability. The main result of the paper is a necessary and sufficient condition for the quadratic stabilizability of the uncertain systems under consideration.     

On estimation of parameters for spatial autoregressive model

In the paper asymptotic properties of the estimated coefficients of multi-indexed autoregressive model are investigated. Considering Least Squares estimates we use some kind of self-normalization and obtain limit normal law independent of unknown parameters. In the proof of this result we use recent result on the Central Limit Theorem for dependent summands.      

Almost self-optimizing strategies for the adaptive control of diffusion processes

The ergodic control of a multidimensional diffusion process described by a stochastic differential equation that has some unknown parameters appearing in the drift is investigated. The invariant measure of the diffusion process is shown to be a continuous function of the unknown parameters. For the optimal ergodic cost for the known system, an almost optimal adaptive control is constructed for the unknown system.This research was partially supported by NSF Grants ECS-87-18026, ECS-91-02714, and ECS-91-13029.     

Robust anti-synchronization of uncertain chaotic systems based on multiple-kernel least squares support vector machine modeling

In this paper, we propose a robust anti-synchronization scheme based on multiple-kernel least squares support vector machine (MK-LSSVM) modeling for two uncertain chaotic systems. The multiple-kernel regression, which is a linear combination of basic kernels, is designed to approximate system uncertainties by constructing a multiple-kernel Lagrangian function and computing the corresponding regression parameters. Then, a robust feedback control based on MK-LSSVM modeling is presented and an improved update law is employed to estimate the unknown bound of the approximation error. The proposed control scheme can guarantee the asymptotic convergence of the anti-synchronization errors in the presence of system uncertainties and external disturbances. Numerical examples are provided to show the effectiveness of the proposed method.     

Contraction theory based adaptive synchronization of chaotic systems

Contraction theory based stability analysis exploits the incremental behavior of trajectories of a system with respect to each other. Application of contraction theory provides an alternative way for stability analysis of nonlinear systems. This paper considers the design of a control law for synchronization of certain class of chaotic systems based on backstepping technique. The controller is selected so as to make the error dynamics between the two systems contracting. Synchronization problem with and without uncertainty in system parameters is discussed and necessary stability proofs are worked out using contraction theory. Suitable adaptation laws for unknown parameters are proposed based on the contraction principle. The numerical simulations verify the synchronization of the chaotic systems. Also parameter estimates converge to their true values with the proposed adaptation laws.     

Impulsive synchronization of discrete-time chaotic systems under communication constraints

This paper investigates the problem of impulsive synchronization of discrete-time chaotic systems subject to limited communication capacity. Control laws with impulses are derived by using measurement feedback, where the effect of quantization errors is considered. Sufficient conditions for asymptotic stability of synchronization error systems are given in terms of linear matrix inequalities and algebraic inequalities. Some numerical simulations are given to demonstrate the effectiveness of the method.