On convergence of active disturbance rejection control for a class of uncertain stochastic nonlinear systems

In this paper, the practical mean-square convergence of active disturbance rejection control for a class of uncertain stochastic nonlinear systems modelled by the Itô-type stochastic differential equations with vast stochastic uncertainties is developed. We first design an extended state observer (ESO) to estimate both the unmeasured states and the stochastic total disturbance which includes unknown internal system dynamics, external stochastic disturbance without known statistical characteristics, unknown stochastic inverse dynamics, and uncertainty caused by the deviation of control parameter from its nominal value. The stochastic total disturbance is then cancelled (compensated) in the feedback loop. An ESO-based output-feedback control is finally designed analogously as for the system without uncertainties. The practical mean-square reference tracking and practical mean-square stability of the resulting closed-loop system are achieved. The numerical experiments are carried out to illustrate the effectiveness of the proposed approach.     

On a discrete-time stochastic learning control algorithm

In an earlier paper by the author (2001), the learning gain for a D-type learning algorithm, is derived based on minimizing the trace of the input error covariance matrix for linear time-varying systems. It is shown that, if the product of the input/output coupling matrices is full-column rank, then the input error covariance matrix converges uniformly to zero in the presence of uncorrelated random disturbances, whereas, the state error covariance matrix converges uniformly to zero in the presence of measurement noise. However, in general, the proposed algorithm requires knowledge of the state matrix. In this note, it is shown that equivalent results can be achieved without the knowledge of the state matrix. Furthermore, the convergence rate of the input error covariance matrix is shown to be inversely proportional to the number of learning iterations     

Solution to a class of stochastic LQ problems with bounded control

A new approach for finding an exact analytical solution to the modified Hamilton-Jacobi-Bellman equation is proposed. Together with the recently developed hybrid solution method, the proposed strategy allows to find a solution to a whole class of stochastic optimal control problems with bounded in magnitude control force.     

Solutions to a class of non-standard non-linear H infinity control problems

This paper presents new solutions to certain non-standard non-linear H infinity control problems. We consider non-linear affine plants whose measurement output is of dimension larger than the dimension of the external input. This problem is, under proper assumptions, transformed to the problem of stabilization by means of output injection and solution of a Hamilton-Jacobi partial differential inequality arising in singular H infinity state-feedback control. General sufficient solvability conditions are given. Explicit solutions are available in the local and semilocal cases. The former concerns a certain neighbourhood of the origin in the closed loop state-space, whereas the latter assumes that the trajectories are restricted to a neighbourhood of an invariant manifold. The issue of the controller order is addressed and a reducedorder controller is obtained in the local case. A new generalization of the chain-scattering formalism provides a very useful framework for solving this problem.     

Optimal control of a class of nonlinear stochastic systems

Bellman's principle of optimality and dynamic programming are shown to be the basis for solution of a physically significant class of nonlinear stochastic control problems. Various previous results are integrated into a survey here, and a new result which extends the separation principle is presented. Certain bilinear and linear-in-control systems are included in the analysis.     

Adaptive stochastic control for a class of linear systems

The problem considered in this paper deals with the control of linear discrete-time stochastic systems with unknown (possibly time-varying and random) gain parameters. The philosophy of control is based on the use of an open-loop feedback optimal (OLFO) control using a quadratic index of performance. It is shown that the OLFO system consists of 1) an identifier that estimates the system state variables and gain parameters and 2) a controller described by an "adaptive" gain and correction term. Several qualitative properties and asymptotic properties of the OLFO adaptive system are discussed. Simulation results dealing with the control of stable and unstable third-order plants are presented. The key quantitative result is the precise variation of the control system adaptive gains as a function of the future expected uncertainty of the parameters; thus, in this problem the ordinary "separation theorem" does not hold.     

Solutions to a class of linear-quadratic-Gaussian (LQG) stochastic team problems with nonclassical information

We consider a stochastic dynamic team problem with two controllers and nonclassical information, which can be viewed as the transmission of a garbled version of a Gaussian message over a number of noisy channels under a fidelity criterion. We show that the optimum solution (under a quadratic loss functional) consists of linearly transforming the garbled message to a certain (optimum) power level P^* and then optimally decoding it by using a linear transformation at the receiving end. The power level P^* is determined by the solution of a fifth order algebraic equation. The paper also discusses an extension of this result to the case when the channel noise is correlated with the input random variable, and shows that for the single channel case the optimum solution is again linear.     

A discrete-time stochastic learning control algorithm

In this paper, the learning gain, for a selected learning algorithm, is derived based on minimizing the trace of the input error covariance matrix for linear time-varying systems. It is shown that, if the product of the input/output coupling matrices is a full-column rank, then the input error covariance matrix converges uniformly to zero in the presence of uncorrelated random disturbances. However, the state error covariance matrix converges uniformly to zero in presence of measurement noise. Moreover, it is shown that, if a certain condition is met, then the knowledge of the state coupling matrix is not needed to apply the proposed stochastic algorithm. The proposed algorithm is shown to suppress a class of nonlinear and repetitive state disturbance. The application of this algorithm to a class of nonlinear systems is also considered. A numerical example is included to illustrate the performance of the algorithm     

Output-feedback control of a class of high-order stochastic nonlinear systems

This article investigates the problem of output-feedback stabilisation for a class of high-order stochastic non-linear systems in which the diffusion terms depend on unmeasurable states besides the output. By introducing a new rescaling transformation, adopting an effective observer and choosing the appropriate Lyapunov function, an output-feedback controller is constructed to ensure that the equilibrium at the origin of the closed-loop system is globally asymptotically stable in probability, the output can be regulated to the origin almost surely, and the problem of inverse optimal stabilisation in probability is solved. The efficiency of the output-feedback controller is demonstrated by several simulation examples.     

Probabilistic-constrained optimal control of a class of stochastic hybrid systems

Stochastic hybrid systems have several applications such as biological systems and communication networks, but it is difficult to consider control of general stochastic hybrid systems. In this paper, a class of discrete-time stochastic hybrid systems, in which only discrete dynamics are stochastic, is considered. For this system, a solution method for the optimal control problem with probabilistic constraints is proposed. Probabilistic constraints guarantee that the probability that the continuous state reaches a given unsafe region is less than a given constant. In the propose method, first, continuous state regions, from which the state reaches a given unsafe region, are computed by a backward-reachability graph. Next, mixed integer quadratic programming problems with constraints derived from the backward-reachability graph are solved. The proposed method can be applied to model predictive control.     

一类高阶次随机非线性系统的输出反馈镇定

针对一类高阶次随机非线性系统,研究其输出反馈镇定问题．通过选择有效的观测器和李雅普诺夫函数,所设计的光滑输出反馈控制器保证了闭环系统的平衡点是依概率全局渐近稳定的,输出几乎处处调节到零．数值仿真验证了控制方案的有效性．     

Output feedback control of a class of stochastic hybrid systems

This paper deals with static output feedback control of a class of reconfigurable systems with Markovian Parameters and state-dependent noise. The main contribution is to formulate conditions for multi-performance design related to this class of stochastic hybrid systems. The specifications and objectives under consideration include stochastic stability, and performances. Another problem related to a more general class of stochastic hybrid systems, known as Markovian Jump Linear Systems (MJLS), is also addressed. This problem concerns the mode-independent output feedback control of MJLS. The obtained results are illustrated on a numerical example.     

Adaptive tracking control for a class of stochastic switched systems

The problem of adaptive tracking is considered for a class of stochastic switched systems, in this paper. As preliminaries, the criterion of global asymptotical practical stability in probability is first presented by the aid of common Lyapunov function method. Based on the Lyapunov stability criterion, adaptive backstepping controllers are designed to guarantee that the closed-loop system has a unique global solution, which is globally asymptotically practically stable in probability, and the tracking error in the fourth moment converges to an arbitrarily small neighbourhood of zero. Simulation examples are given to demonstrate the efficiency of the proposed schemes.     

Singular perturbation results for a class of stochastic control problems

Composite control originally proposed in a deterministic context is generalized to the problem with white-noise inputs. However, the approach used here is radically different from the deterministic approach. Presence of noise smoothes the system behavior and allows a more complete solution than in the deterministic case.     

Near optimal control for a class of stochastic hybrid systems

To address a computationally intractable optimal control problem for a class of stochastic hybrid systems, this paper proposes a near optimal state feedback control scheme, which is constructed by using a statistical prediction method based on approximate numerical solution that samples over the entire state space. A numerical example illustrates the potential of the approach.     

Infinite time optimal control for a class of stochastic systems

Optimization results developed by the author (Brandeberry and Wu 1970) for a class of stochastic regulator problems will be extended from the finite time interval to the infinite time interval. Conditions are given for the existence and stability of the infinite constant feedback of the system state; necessary conditions will also be obtained for the inverse problem (i.e. conditions for a linear constant control law to be optimal for some cost functional).     

Adaptive observer-based control for a class of nonlinear stochastic systems

In this paper, the adaptive state estimation and state-feedback stabilization problems for a class of nonlinear stochastic systems with unknown constant parameters are studied. The sequential design methods are proposed to construct the adaptive controllers. Adaptive state and parameter estimators are designed by using a stochastic Lyapunov method and the separation theory of the design for the state-feedback gain and observer gain, which guarantees that the closed-loop system is asymptotically stable in the mean-square sense. Sufficient conditions for the existence of parameters estimator are given in terms of linear matrix inequalities. Finally, the numerical examples are provided to illustrate the feasibility of the proposed theoretical results.     

Efficiency of rolling schedules for a class of stochastic control problems

Analytic results are derived to compare the expected cost of using a rolling schedule (sequential open-loop optimal controller) to the minimum expected cost of employing a closed-loop optimal feedback controller for LQG control problems with varying nominal state and control trajectories and nonzero-mean disturbances. Two numerical examples illustrate the results.     

Limiting optimal discounted-cost control of a class of time-varying stochastic systems

We consider a class of time-varying -valued control models, and with possibly unbounded costs. The processes evolve according to the system equation x_n+1=G_n(x_n,a_n)+ξ_n ( ), where {ξ_n} are i.i.d. random vectors and {G_n} a sequence of known functions converging to some function G_∞. Under suitable hypotheses, we show the existence of an α-discount optimal policy for the limiting system x_n+1=G_∞(x_n,a_n)+ξ_n.