Search

In this article, we consider quantized learning control for linear networked systems with additive channel noise. Our objective is to achieve high tracking performance while reducing the communication burden on the communication network. To address this problem, we propose an integrated framework consisting of two modules: a probabilistic quantizer and a learning scheme. The employed probabilistic quantizer is developed by employing a Bernoulli distribution driven by the quantization errors. Three learning control schemes are studied, namely, a constant gain, a decreasing gain sequence satisfying certain conditions, and an optimal gain sequence that is recursively generated based on a performance index. We show that the control with a constant gain can only ensure the input error sequence to converge to a bounded sphere in a mean-square sense, where the radius of this sphere is proportional to the constant gain. On the contrary, we show that the control that employs any of the two proposed gain sequences drives the input error to zero in the mean-square sense. In addition, we show that the convergence rate associated with the constant gain is exponential, whereas the rate associated with the proposed gain sequences is not faster than a specific exponential trend. Illustrative simulations are provided to demonstrate the convergence rate properties and steady-state tracking performance associated with each gain, and their robustness against modeling uncertainties. [ABSTRACT FROM AUTHOR]

When discrete-time Markov jump linear systems are prone to the damaging effects of polytopic uncertainties, it is necessary to address all the vertices of each Markov mode in order to properly design robust controllers. To this end, we propose a robust recursive linear–quadratic regulator for this class of systems. We define a quadratic min–max optimization problem by combining least-squares and penalty functions in a unified framework. We design a one-step cost function to encompass the entire set of vertices of each mode altogether, while maintaining its quadratic structure and the convexity of the problem. The solution is then obtained recursively and does not require numerical optimization packages. We establish conditions for convergence and stability by extending the matrix structure of the recursive solution. In addition, we provide numerical and real-world application examples to validate our method and to emphasize recursiveness and diminished computational effort. [ABSTRACT FROM AUTHOR]

Two stochastic model predictive control algorithms, which are referred to as distributionally robust model predictive control algorithms, are proposed in this article for a class of discrete linear systems with unbounded noise. Participially, chance constraints are imposed on both of the state and the control, which makes the problem more challenging. Inspired by the ideas from distributionally robust optimization (DRO), two deterministic convex reformulations are proposed for tackling the chance constraints. Rigorous computational complexity analysis is carried out to compare the two proposed algorithms with the existing methods. Recursive feasibility and convergence are proven. Simulation results are provided to show the effectiveness of the proposed algorithms. [ABSTRACT FROM AUTHOR]

Linear complementarity system (LCS) consists of an ordinary differential equation (ODE) and a linear complementarity problem (LCP). In this article, we propose an exponential time-stepping method for solving LCS, which uses exponential integrator to discretize the ODE and solves the LCPs at the discrete time points. Numerical results are reported for illustrating its good performance. [ABSTRACT FROM AUTHOR]

The goal of this article is to investigate new and simple convergence analysis of dynamic programming for the linear–quadratic regulator problem of discrete-time linear time-invariant systems. In particular, bounds on errors are given in terms of both matrix inequalities and matrix norm. Under a mild assumption on the initial parameter, we prove that the $Q$ -value iteration exponentially converges to the optimal solution. Moreover, a global asymptotic convergence is also presented. These results are then extended to the policy iteration. We prove that in contrast to the $Q$ -value iteration, the policy iteration always converges exponentially fast. An example is given to illustrate the results. [ABSTRACT FROM AUTHOR]

This article proposes a method for enlarging the region of attraction of linear model predictive controllers (MPC) when tracking piecewise-constant references in the presence of pointwise-in-time constraints. It consists of an add-on unit, the feasibility governor (FG), that manipulates the reference command so as to ensure that the optimal control problem that underlies the MPC-feedback law remains feasible. Offline polyhedral projection algorithms based on multiobjective linear programming are employed to compute the set of feasible states and reference commands. Online, the action of the FG is computed by solving a convex quadratic program. The closed-loop system is shown to satisfy constraints, be asymptotically stable, exhibit zero-offset tracking, and display finite-time convergence of the reference. [ABSTRACT FROM AUTHOR]

This article studied the stability and convergence of a robust iterative learning control (ILC) design for a class of nonlinear systems with unknown control input delay. First, the iterative integral sliding mode (IISM) design was proposed, which comprised iterative actions. The iterative action made the convergence of the tracking error under the ideal sliding mode. Then, a suitable iterative update law was provided for the IISM-based robust ILC controller. The controller had the capability of both minimizing the steady tracking error and suppressing the unrepeatable disturbance. Using the controller, the closed-loop system stability was analyzed, and the stability conditions were given. Consequently, the sliding mode convergence in the iteration domain was proved by a composite energy function (CEF). In addition, by analyzing the influence of affection on the tracking error, several measures were taken to solve the chattering problem of the sliding mode control. Finally, a one-link robotic manipulator and a vertical three-tank system were used to verify the control design. The application simulations validated the performance of the proposed sliding mode iterative learning control (SMILC) design, which achieved the stability of the nonlinear system and overcame the control input time delay. [ABSTRACT FROM AUTHOR]

In this article, a noisy-output-based direct learning tracking control is proposed for stochastic linear systems with nonuniform trial lengths. The iteration-varying trial length is modeled using a Markov chain for demonstration of the iteration dependence. The effect of the noisy output is asymptotically eliminated using a prior given decreasing gain sequence in the learning algorithm. Two alternative adaptive gains are presented for improving the tracking performance and the convergence speed. Both the mean-square and almost-sure convergence are provided. Numerical simulations on a four-degree-of-freedom robot arm are presented to illustrate the effectiveness of the proposed scheme. [ABSTRACT FROM AUTHOR]

The problem of solving linear equations is considered as one of the fundamental problems commonly encountered in science and engineering. In this article, the complex-valued time-varying linear matrix equation (CVTV-LME) problem is investigated. Then, by employing a complex-valued, time-varying QR (CVTVQR) decomposition, the zeroing neural network (ZNN) method, equivalent transformations, Kronecker product, and vectorization techniques, we propose and study a CVTVQR decomposition-based linear matrix equation (CVTVQR-LME) model. In addition to the usage of the QR decomposition, the further advantage of the CVTVQR-LME model is reflected in the fact that it can handle a linear system with square or rectangular coefficient matrix in both the matrix and vector cases. Its efficacy in solving the CVTV-LME problems have been tested in a variety of numerical simulations as well as in two applications, one in robotic motion tracking and the other in angle-of-arrival localization. [ABSTRACT FROM AUTHOR]

This article presents a computational method to test the stabilizability of discrete-time switched linear systems. The existence of a conic cover of the space on whose elements a convex condition holds is proved to be necessary and sufficient for stabilizability. An algorithm for computing a conic partition that satisfies the new necessary and sufficient condition is given. The algorithm, which allows us to determine bounds on the exponential convergence rate, is proved to overcome the conservatism of conditions equivalent to periodic stabilizability and is applied to a 4-D system. [ABSTRACT FROM AUTHOR]

Truncated Newton (TN) methods have been a useful technique for large-scale optimization. Instead of obtaining the full Newton direction, a truncated method approximately solves the Newton equation with an inner conjugate gradient (CG) procedure (TNCG for the whole method). These methods have been employed to efficiently solve linear classification problems. However, even in this deeply studied field, various theoretical and numerical aspects were not completely explored. The first contribution of this work is to comprehensively study the global and local convergence when TNCG is applied to linear classification. Because of the lack of twice differentiability under some losses, many past works cannot be applied here. We prove various missing pieces of theory from scratch and clarify many proper references. The second contribution is to study the termination of the CG method. For the first time when TNCG is applied to linear classification, we show that the inner stopping condition strongly affects the convergence speed. We propose using a quadratic stopping criterion to achieve both robustness and efficiency. The third contribution is that of combining the study on inner stopping criteria with that of preconditioning. We discuss how convergence theory is affected by preconditioning and finally propose an effective preconditioned TNCG. [ABSTRACT FROM AUTHOR]

A finite-time observer for a class of linear time-varying systems with bounded unknown inputs is presented in this note. The design of the observer exploits structural properties of the system and, through a linear operator, allows applying the robust exact differentiator in a block form without requiring additional stabilizing terms. The proposed observer provides an exact estimate of the states after a finite transient time despite the possible instability of the system and the effects of bounded unknown inputs. [ABSTRACT FROM AUTHOR]

Recent studies have shown how regularization may play an important role in linear system identification. An effective approach consists of searching for the impulse response in a high-dimensional space, e.g., a reproducing kernel Hilbert space (RKHS). Complexity is then controlled using a regularizer, e.g., the RKHS norm, able to encode smoothness and stability information. Examples are RKHSs induced by the so-called stable spline or tuned-correlated kernels, which contain a parameter that regulates impulse response exponential decay. In this article, we derive nonasymptotic upper bounds on the $\ell _2$ error of these regularized schemes and study their optimality in order (in the minimax sense). Under white noise inputs and Gaussian measurement noises, we obtain conditions which ensure the optimal convergence rate for all the class of stable spline estimators and several generalizations. Theoretical findings are then illustrated via a numerical experiment. [ABSTRACT FROM AUTHOR]

This article considers the problem of output feedback stabilization for a class of nonlinear planar systems with unknown structures and measurements, which prevent the construction of conventional state observers. By taking advantage of the stability-increasing capability of a lead compensator, we propose a dynamic output feedback controller to globally stabilize the uncertain planar systems. For the special case of linear planar systems with unknown coefficients, a finite-time output feedback stabilizer based on a nonlinear compensator is constructed for a faster convergence rate. [ABSTRACT FROM AUTHOR]

In this article, a robust nonlinear model reference adaptive control (MRAC) is proposed for disturbed linear systems, i.e., linear systems with parameter uncertainties, and external time-dependent perturbations or nonlinear unmodeled dynamics matched with the control input. The proposed nonlinear control law is composed of two nonlinear adaptive gains. Such adaptive gains allow the control to counteract the effects of some perturbations and nonlinear unmodeled dynamics ensuring asymptotic convergence of the tracking error to zero, and the boundedness of the adaptive gains. The nonlinear controller synthesis is given by a constructive method based on the solution of linear matrix inequalities. Besides, the simulation results show that, due to the nonlinearities, the rate of convergence of the proposed algorithm is faster than that provided by a classic MRAC. [ABSTRACT FROM AUTHOR]

This article studies fixed-time stabilization (FxTS) of a general controllable linear system with an input delay $\tau$. It is shown that such a problem is not solvable if the prescribed convergence time $T_{\tau }$ is smaller than $2\tau$. For $T_{\tau }\geq 3\tau$ , a solution based on linear periodic delayed feedback (PDF) without any distributed delay is established. For $T_{\tau }>2\tau$ , a solution based on linear predictor-based PDF containing a distributed delay is proposed. For both cases, the gains of the PDF can be chosen as continuous, continuously differentiable, and even smooth, in the sense of infinitely many times differentiable. If only an output signal is available for feedback, two classes of linear observers with periodic coefficients are designed, so that their states converge to the current and future states of the system at a prescribed finite time, respectively. With the observed current and future states, FxTS can also be achieved by using, respectively, the PDF and observer-based PDF. A linear periodic feedback (without delay) is also established to solve the FxTS problem of linear systems with both instantaneous and delayed controls, which cannot be stabilized by any constant instantaneous feedback in certain cases. Two numerical examples verify the effectiveness of the proposed approaches. [ABSTRACT FROM AUTHOR]

The linear quadratic regulator (LQR) problem has reemerged as an important theoretical benchmark for reinforcement learning-based control of complex dynamical systems with continuous state and action spaces. In contrast with nearly all recent work in this area, we consider multiplicative noise models, which are increasingly relevant because they explicitly incorporate inherent uncertainty and variation in the system dynamics and thereby improve robustness properties of the controller. Robustness is a critical and poorly understood issue in reinforcement learning; existing methods which do not account for uncertainty can converge to fragile policies or fail to converge at all. Additionally, intentional injection of multiplicative noise into learning algorithms can enhance robustness of policies, as observed in ad hoc work on domain randomization. Although policy gradient algorithms require optimization of a nonconvex cost function, we show that the multiplicative noise LQR cost has a special property called gradient domination, which is exploited to prove global convergence of policy gradient algorithms to the globally optimum control policy with polynomial dependence on problem parameters. Results are provided both in the model-known and model-unknown settings where samples of system trajectories are used to estimate policy gradients [ABSTRACT FROM AUTHOR]

A novel approach to solve the problem of distributed state estimation of linear time-invariant systems is proposed in this article. It relies on the application of parameter-estimation-based observers, where the state observation task is reformulated as a parameter estimation problem. In contrast with existing results, our solution achieves convergence in finite time, without injection of high gain, and imposes very weak assumptions on the communication graph—namely, the existence of an open Hamiltonian walk. It is shown that this assumption is strictly weaker than the usual strong connectivity requirement. The scheme is shown to be robust vis-á-vis external disturbances and communication delays. [ABSTRACT FROM AUTHOR]

In this article, we deal with the consensus problem of multiple partial-state coupled linear systems, which are neutrally stable. These systems communicate over dynamic undirected networks, which change continuously and can be disconnected at any time. We develop an analysis framework from uniform complete observability theory to work out the necessary and sufficient conditions for exponential consensus. It turns out that, with a suitably designed feedback matrix, exponential consensus can be realized globally and uniformly if and only if a jointly $(\delta,T)$ -connected condition and an observability condition relying only on system and input matrices are satisfied. A lower bound of the convergence rate is also provided. We figure out the proof by applying matrix analysis and linear functional analysis. A simulation example is presented to illustrate our theoretical findings. [ABSTRACT FROM AUTHOR]

In the article, we study the distributed model predictive control (DMPC) problem for a network of linear discrete-time systems, where the system dynamics are decoupled, the system constraints are coupled, and the communication networks are described by time-varying directed graphs. A novel distributed optimization algorithm called the push-sum dual gradient (PSDG) algorithm is proposed to solve the dual problem of the DMPC optimization problem in a fully distributed way. We prove that the sequences of the primal, and dual variables converge to their optimal values. Furthermore, to solve the implementation issues, stopping criteria are designed to allow early termination of the PSDG Algorithm, and the gossip-based push-sum algorithm is proposed to check the stopping criteria in a distributed manner. It is shown that the optimization problem is iteratively feasible, and the closed-loop system is exponentially stable. Finally, the effectiveness of the proposed DMPC approach is verified via an example. [ABSTRACT FROM AUTHOR]

This article studies the performance evaluation of distributed linear regression Kalman filtering fusion for nonlinear systems. Sufficient conditions are established for the convergence of the centralized fusion (CF) under the assumption of bounded estimation error covariance, and a measure of performance is derived from the convergence conditions. By the performance analysis, it can be found that the CF has a better performance than the distributed fusion with feedback, especially at the beginning of the estimation. Moreover, the performance of the local estimator can be improved by receiving the fused estimate from the fusion center, which is different from the fusion estimation in linear systems. Finally, by simulations of a target tacking example, the comparisons of the centralized fusion and the distributed fusion with and without feedback are presented to show the accuracy of the performance analysis. [ABSTRACT FROM AUTHOR]

We present some new results on the dynamic regressor extension and mixing parameter estimators for linear regression models recently proposed in the literature. This technique has proven instrumental in the solution of several open problems in system identification and adaptive control. The new results include the following, first, a unified treatment of the continuous and the discrete-time cases; second, the proposal of two new extended regressor matrices, one which guarantees a quantifiable transient performance improvement, and the other exponential convergence under conditions that are strictly weaker than regressor persistence of excitation; and, third, an alternative estimator ensuring convergence in finite-time whose adaptation gain, in contrast with the existing one, does not converge to zero. Simulations that illustrate our results are also presented. [ABSTRACT FROM AUTHOR]

Due to their relevance in controller design, we consider the problem of determining the L2-gain, passivity properties, and conic relations of an input–output system. While, in practice, the input–output relation is often undisclosed, input–output data tuples can be sampled by performing (numerical) experiments. Hence, we present sampling strategies for discrete time and continuous time linear time-invariant systems to iteratively determine the L2-gain, the shortage of passivity and the cone with minimal radius that the input–output relation is confined to. These sampling strategies are based on gradient dynamical systems and saddle point flows to solve the reformulated optimization problems, where the gradients can be evaluated from only input–output data samples. This leads us to evolution equations, whose convergence properties are then discussed in continuous time and discrete time. [ABSTRACT FROM AUTHOR]

This article addresses the problem of robust observer design for nonlinear systems with transformed output. We first assume that there exists an observer design for a given nonlinear system. The structure of the system output is often crucial for estimating the system states. Therefore, we consider the problem that arises when the output is only available under a nonlinear transformation (affected by noise). This formulation is especially useful in engineering applications, where nonlinear sensors are frequent. We propose a new interconnected observer that estimates both the unavailable output and the system states, and we establish small-gain type conditions ensuring the asymptotic convergence of the overall error dynamics. To explicit our design parameters, we study the particular cases of state-affine systems up to output injection and systems with additive triangular nonlinearities. Finally, we provide corresponding numerical examples and comparisons with known observers. [ABSTRACT FROM AUTHOR]

This article is concerned with the moving horizon estimation (MHE) problem for networked linear systems (NLSs) with unknown inputs under dynamic quantization effects. For the NLSs with unknown input signals, the conventional MHE strategy is incapable of guaranteeing the satisfactory performance as the estimation error is dependent on the external disturbances. In this work, a novel MHE strategy is developed to cope with the underlying NLS with unknown inputs by dedicatedly introducing certain temporary estimates of unknown inputs, where the desired estimator parameters are designed to decouple the estimation error dynamics from the unknown inputs. A two-step design strategy (namely, decoupling step and convergence step) is proposed to obtain the estimator parameters. In the decoupling step, the decoupling parameter of the moving horizon estimator is designed based on certain assumptions on system parameters and quantization parameters. In the convergence step, by employing a special observability decomposition scheme, the convergence parameters of the moving horizon estimator are achieved such that the estimation error dynamics is ultimately bounded. Moreover, the developed MHE strategy is extended to the scenario with direct feedthrough of unknown inputs. Two simulation examples are given to demonstrate the correctness and effectiveness of the proposed MHE strategies. [ABSTRACT FROM AUTHOR]

This article is a comprehensive study of a long-observed phenomenon of increase in the stability margin and so the rate of convergence of a class of linear systems due to time delay. We use the Lambert W function to determine: 1) in what systems the delay can lead to increase in the rate of convergence; 2) the exact range of time delay, for which the rate of convergence is greater than that of the delay-free system; and 3) an estimate on the value of the delay that leads to the maximum rate of convergence. For the special case when the system matrix eigenvalues are all negative real numbers, we expand our results to show that the rate of convergence in the presence of delay depends only on the eigenvalues with minimum and maximum real parts. Moreover, we determine the exact value of the maximum rate of convergence and the corresponding maximizing time delay. We demonstrate our results through a numerical example on the practical application in accelerating an agreement algorithm for networked systems by using a delayed feedback. [ABSTRACT FROM AUTHOR]

To achieve control objectives for extremely large-scale complex networks using standard methods is essentially intractable. In this article, a theory of the approximate control of complex network systems is proposed and developed by the use of graphon theory and the theory of infinite dimensional systems. First, graphon dynamical system models are formulated in an appropriate infinite dimensional space in order to represent arbitrary-size networks of linear dynamical systems, and to define the convergence of sequences of network systems with limits in the space. Exact controllability and approximate controllability of graphon dynamical systems are then investigated. Second, the minimum energy state-to-state control problem and the linear quadratic regulator problem for systems on complex networks are considered. The control problem for graphon limit systems is solved in each case and approximations are defined which yield control laws for the original control problems. Furthermore, convergence properties of the approximation schemes are established. A systematic control design methodology is developed within this framework. Finally, numerical examples of networks with randomly sampled weightings are presented to illustrate the effectiveness of the graphon control methodology. [ABSTRACT FROM AUTHOR]

This article considers the problem of designing physical watermark signals in order to optimally detect possible replay attack in a linear time-invariant system, under the assumption that the system parameters are unknown and need to be identified online. We first provide a replay attack model, where an adversary replays the previous sensor data in order to fool the system. A physical watermarking scheme, which leverages a random input as a watermark to detect the replay attack, is then introduced. The optimal watermark signal design problem is cast as an optimization problem, which aims to achieve the optimal trade-off between control performance and intrusion detection. An online watermarking design and system identification algorithm is provided to deal with systems with unknown parameters. We prove that the proposed algorithm converges to the optimal one and characterize the almost sure convergence rate. An industrial process example is provided to illustrate the effectiveness of the proposed strategy. [ABSTRACT FROM AUTHOR]

In this technical note, the tradeoff between the attack detectability and the performance degradation in stochastic cyber-physical systems is investigated. We consider a linear time-invariant system in which the attack detector performs a hypothesis test on the innovation of the Kalman filter to detect malicious tampering with the actuator signals. We adopt a notion of attack stealthiness to quantify the degree of stealth by limiting the maximum achievable exponents of both false alarm probability and detection probability below certain thresholds. And the conditions for any actuator attack to have a specific level of stealthiness are derived. Additionally, we characterize the upper bound of the performance degradation induced by attacks with a given extent of stealthiness that produces independent and identically distributed Gaussian innovations, and design the attack, which achieves the stated upper bound for right-invertible systems. Finally, our results are illustrated via numerical examples. [ABSTRACT FROM AUTHOR]